I figured you might come back with an, "I told you so". Interesting that this guy said:
for numbers that were 310,000 for this year for HD DVD through the 18th of March, and yet you were claiming more than 2 per player per month and 500,000 just for January (and 925,000 through February). If he cross checked with studios to ensure accuracy and got 310k, why would his numbers be so far from your claims? Maybe you are the one who needs to show he is a crackpot. :)

Just to be sure, are you still claiming 500,000 for HD DVD for January and more than 2 per player per month for this year, so far?

--Darin


Like I said previously, the 925k disc sales for Jan/Feb included all sales outlets including retail, rental, possibly for freebies (I don't know what kind of deal exists to supply Toshiba with their freebies), etc.

Since they are projecting based on Nielsen numbers, and Nielsen only tracks retail sales, then the 310k number they verified must also pertain to retail sales, or their statements are inaccurate. I see nothing there to dispute anything I've said. If anything, it supports it.

It also points out that the numbers Icemange arrived at, were lucky to be as close as they are. In all past estimates, the numbers arrived at on this forum were presented as discs sold, which then had to be increased by X% to estimate actual volumes (basically X% was not built into the calculations). However, now the same numbers appear to correspond with projected values, despite the 40% not being part of the equation.

Or basically, had Icemage been performing the estimations correctly from the beginning, his table would show something like 124k HD-DVD disc sales YTD. He would then need to divide that by .4 to arrive at the projected 310k. The same for Blu-ray figures. So basically, if you believe that HMM has better access and is a better source of data, then you also must believe that all the numbers calculated and presented to date have been wrong, which means the process at arriving at those numbers must also have a flaw.

There is no reason the sample should be biased ....

To me this is the key. I don't see why Nielsen's 40% would be biases one way or the other (w.r.t. two formats).

It doesn't really matter whether the sample is really random or not. The question is - is it biased in favor of one or the other format.

For eg. let us assume Nielsen basically gets data from only B&M. Then further assume that HD DVD users are known to buy mostly online. Then we would say Nielsen sample is biased.

Unless a rational argument is made showing Nielsen is biased, I'd have to say they are a very good representative of actual sales.

.... <sigh>... pls don't ressurect the "free" - "$10 off" "etc." PS3 vouchers debate.... go back 3 to 5 pages and read all the point made by each side....

We can talk all we want about statistical accuracy from Nielsen, but the fact that they cover even 40% of the industry for standard DVD is a powerful statement (and I'll go with nataraj and say their representation may actually be better than that for high definition since there are far fewer significant sales venues for high def discs than for SD DVD).

The point here is that a 40% sample of any sort is most likely to be representative unless you have a good reason to believe that the remaining 60% are behaving in a different way.

In this case, you have consumers buying two competing formats which have the same properties pretty much whereever you go in the statistical geographic region (North America). While Kosty makes good points about it being possible that a 40%+ sample can give an incorrect projection, what this does not address is that what we are dealing with are consumers who are buying from largely unaligned retailers. They have the same general conditions for purchase; they want the same things, they get the same general pricing, etc.

Statistical inaccuracy from 40% from a non-random sampling matters little if retailers as a whole show similar sales trends. Given what information we've heard over the past several months, all of the evidence points to most retailers seeing similar changes in sales volumes. The effect might be more pronounced in one venue than another (particularly in the occasional cases where the retailer is biased), but the overall trend should remain the same.


I think conditions for bias definately exist in the Nielsen data collection, but as time goes by, the conditions (and bias) stabilize. For example, Let's say you live in a town that only has a CC. Up until a month or so ago, CC only carried BD, so it was not possible to by an HD-DVD from them. Now let's say you did by an HD-DVD player. You would not be able to buy discs locally, so you turn to the net. Let's say you buy your discs at Deep Discount or any other online retailer that doesn't report to Videoscan. This puts a potential bias into the Videoscan numbers in favor of BD. However, as time has went by, CC carries both, and the bias caused by lack of neutrality with CC has been removed. There are other online and offline retailers that are not neutral, and therefore present a potential bias for consumers that shop in those areas.

Even a retailers lack of adequate inventory to bias the consumers true purchasing wishes. For example, let's say you really want to get Happy Feet in hi-def (a dual format release). Now lets say you are also dual format in your household. Let's say you want the disc on HD-DVD because the same review source gave it 5/5 and the BD version 4.5/5. You go to your local retailer (or a website you use often; ecost for example, because it has the BD, but not HD-DVD), and they don't have a HD-DVD copy in stock because of inadequte inventory purchase. But since you really want the film, and can play both, you buy the BD copy. BD just got a biased +1 (the true desire of the consumer was for HD-DVD) because of a retailers poor stock desisions in a niche product. This is a problem for niche products, which exists to a much lesser extent in mainstream products (most retailers probably got a bulk display with a 100 copies of the DVD).

There are plenty of potential sources of bias in a niche market, when using a model developed for a mainstream market. Further, it is potentially a false assumption to assume that every retailer tracked by Nielsen sales more BD than HD-DVD. BD sales more in total, but which format sales better retailer to retailer can differ. Further, it is a poor assumption to assume that because this 40% population of retailers is comprised primarily of retailers that sell more BD, that it must mean the same holds true for the other 60%. Its the same with elections. When you elect a Govenor, just because 40% of precincts have reported and show candidate X in the lead, it doesn't mean they other 60% of precincts will vote the same way. Candidate Y could actually turn out to be ahead and the winner, once the other 60% of precincts are accounted for.

Nielson's Official Site

http://www.videoscan.com/about.html


Again here is kdragon's pdf's

PDF on Methodology - -this is basically a match for a copy I had received at a trade show several years ago
http://www.avsforum.com/avs-vb/attachment.php?attachmentid=75417


PDF on Included Retailers, including first alert POS providers
http://www.avsforum.com/avs-vb/attachment.php?attachmentid=75416


I have been told by several people with access to the real numbers that both documents are reliable, just not current.

From these and the data on the Videoscan site, it doesn't seem like there are any projections for the total market implied in the Nielson Videoscan numbers.

If anyone ever discredited these documents, please post here or PM me. I have confirmed to my satisfaction that they are basically accurate but not absolutely current.
=============================


earlier discussion in this thread
http://www.avsforum.com/avs-vb/showthread.php?p=9767608&&#post9767608


====================
I have now been told those documents are basically right. BTW Nielson/Videoscan wants a lot for the detailed information and their NDA is nasty.


I think we did more than enough to prove at least the list is inaccurate, and potentially falsified. Surely you remember that discussion earlier in this thread. There are retailers on the list that have been out of business since 2000, yet the list is dated Nov and Feb 2006! If the list is merely out of date, I would say 6 years out of date is huge (assuming only the more recent 2006 dates were falsified)! Further, I contacted ACNielsen and they would not supply a list of retailers, and in a follow-up, would not say that this list was one of theirs. Further, they directed me to contact Videoscan with any further inquiry, and gave me completely different contact info than what is posted on Videoscan's site.

I also attempted to contact Videoscan using the information on their website, and using the information provided to me by ACNielsen. I never received any response to my inquiries from the contact information on the web. Using the contact info given me by ACNielsen, I was told that Videoscan contributors are protected by a NDA, and that a list of contributors could not be given out. The explained that only a few parties had access to a list of contributing members, and that each contributing member had to consent to allowing their name to be released to those select few (meaning you could submit data, but not consent to a release, which means your would contribute anonymous to everyone but Nielsen).

In light of everything uncovered about this "list", and the apparent difficulty to obtain such a list from Nielsen, I give it a credibility of zero...

Candidate Y could actually turn out to be ahead and the winner, once the other 60% of precincts are accounted for.

I don't know your background, but I guess it is not statistical analysis.

From what we know, is there an inherent bias in the Nielsen sample ? If you think so let us hear your logic.

I see nothing there to dispute anything I've said. If anything, it supports it.So, why not answer the question about whether your position is still that HD DVD has been selling more than 2 discs per player per month this year? If you think that the 310k supports that claim you made, then you must be using some pretty screwed up math. So, please show how it supports 2 discs per player per month. I'm sure others can see that it doesn't support the position you took and to now claim that it doesn't dispute anything you've said seems pretty disingenuous, at best.

--Darin

I don't know your background, but I guess it is not statistical analysis.

From what we know, is there an inherent bias in the Nielsen sample ? If you think so let us hear your logic.

From what we know, how can we verify that their sample is not biased? From what I know BB has been buying 4X as many BD titles than HD DVD - at least in my area. As a result some of the newer titles like Lucky #Slevin, The Departed and Clerks etc. were sold out in the week of their release, but they had plenty of BD release titles. So actions such as this will force people to buy from other sources....so IF BB is a significant part of the Neilson data, I would say it is biased. No doubt in my mind!

BB's desire to have one format is biasing any (defacto) sampling Videoscan may be doing. Example: If you try to track the avg. speed on the fast lane by using a police car to track the speed, you will have a bias since people will slow down when they see a police car and the avg. will be low. So here too, BB by actively stocking and promoting BD over HD DVD is creating a bias greater than the avg. difference in sales. I am not saying that BD isn't outselling HD DVD, but unless we know how much BD is outselling HD DVD by at stores like BB (which are biased) and then compare it to the actual results Videoscan presents and also compare BB contribution to the 40% of the data they claim to have....we cannot ascertain how reliable Videoscan data is. You have to eliminate bias for it to be reliable.

From what we know, how can we verify that their sample is not biased? From what I know BB has been buying 4X as many BD titles than HD DVD - at least in my area. As a result some of the newer titles like Lucky #Slevin, The Departed and Clerks etc. were sold out in the week of their release, but they had plenty of BD release titles. So actions such as this will force people to buy from other sources....so IF BB is a significant part of the Neilson data, I would say it is biased. No doubt in my mind!

Using your example, this not bias by Nielson, but rather bias by Best Buy. And this "bias" by Best Buy is probably a result of "bias" of the local buying public. In general, there is no bias at the public level (as a whole), because of that point, its called consumer will.

I don't know your background, but I guess it is not statistical analysis.

From what we know, is there an inherent bias in the Nielsen sample ? If you think so let us hear your logic.


Actually, I aced probability and statistics and my college senior capstone design project relied on statistics (the design of a pseudo-random noise generator whose gausian distribution could be biased with seed values). I also use probability to some extent in my current employment is determining the repeatability of measurements (simple statistics, but still statistics).

As I stated above, there is an inherent bias because the product is niche. I gave a few examples of sources for the bias, and there are more beyond what is listed. No one can say with any certainty how prevelent they are, or how much they might impact the final analysis.

Do you find fault with my comparing the Nielsen data and election precinct reporting? The analogy holds true. Each disc bought counts as a vote towards one format or the other (the difference being individuals can vote as many times as their wallet affords, not just once). Nielsen is asking stores (ala precincts) how many BD discs (votes) and how many HD-DVD discs (votes) they logged. They then compile all these stores (precincts) data to determine which format has the most sales (votes). However, in this case, Nielsen only acquires 40% of disc sales (votes), so they cannot determine definatively who is winning. Statistically, they might be able to say with a high probability, one format beats the other, if enough sales (votes) are reported. However, 40% reporting is not a large enough sample size to say with a high probability that one format is beating the other. You might be able to sell it to me if the number was 70-80%, but 40% is still way too low.

There is no established relation that says because Best Buy sales 2:1 for BD, that an untracked store sales 2:1 in the same ratio. Just because a Super Wal-Mart sells 3k jugs of milk a week (2k 1% and 1k skim), doesn't mean the Kroger next door also sales the same volume in the same ratio. The product would be the same both places, but pricing, availability, size of customer base, customer demographics, etc. all effect volume and ratio.

In a mainstream market, like DVD, less bias exists, and tracking methods like Nielsen, can produce results representative of the market. In a mainstream product, the prices are more regulated, the items are available in almost every outlet (especially new releases)(you can get DVD's at my local gas station, Rite Aid, Kroger, etc.), a performance history exists to help order adequate inventory, tacking services catch a higher percentage of the total sales, and so forth.

With a new and niche product like either HD format, the prices vary wildly, the discs are harder to find (the majority of outlets don't even carry them yet), some outlets carry only one format or the other, retailers lack a performance history to adequately order new releases (and stock previous releases), the total percentage of sales is lower, and so forth.

A model that works well for one type of product (mainstream) will not necessarily work well for another type of product (niche). Given the low volume, variations and potential sources of bias, a 40% sample size is not nearly large enough, to say with any high probability, that one format is performing better than the other.

Using your example, this not bias by Nielson, but rather bias by Best Buy. And this "bias" by Best Buy is probably a result of "bias" of the local buying public. In general, there is no bias at the public level (as a whole), because of that point, its called consumer will.


The bias would be introduced by Best Buy, but it would be tracked and reported on by Nielsen.

Using your example, this not bias by Nielson, but rather bias by Best Buy. And this "bias" by Best Buy is probably a result of "bias" of the local buying public. In general, there is no bias at the public level (as a whole), because of that point, its called consumer will.


Not true!

1. IF you have bias, then the sample is invalid

2. IF BB does not carry stock, then it is underreporting demand, so it does NOT reflect the consumers will, but rather they wish to enforce their will on the consumer. Not the same thing...

3. The consumer will simply buy from a source that is not BB. The question is, are those alternate sources covered by Videscan as well or not!

4. BB may be trying to project a view of the market that they wish to see and if they have a significant share of the 40% can do that...

josh,

Using your example, this not bias by Nielson, but rather bias by Best Buy.

I think you are confused. No one is saying that Nielson is consciously biased towards one format or another. The argument is whether the sample that Nielsen collects is biased for whatever reason. Plazman says that the BB sample may be biased by the purchasing decisions they make. WNorris correctly points out that until recently CC didnt even carry HD DVDs, making their sample biased, unless Nielsen differentially weights that.

And this "bias" by Best Buy is probably a result of "bias" of the local buying public.

You have no evidence that this is necessarily the case. It could be that BB gets a bigger discount on BD discs, or some other incentive. One doesnt know.

yes it did. Out of the >100 HD DVDs that have been produced that is just 1. If>99% of films on HD DVD have AACS you can find the needle. I never doissagreed that AACS is necessary on HD DVD nor did I ever say all HD DVD titles hyave it. what I did is disagree with Noris who pretended none of them had AACS. The AACS fees are AACS fees. if someone adds it to their HD DVD then the cost is there. If not it is not there. Saying AACS fees only apply to BD is just wrong and BS and FUD.


Sorry, wrong. The arguement about replication costs stem from porn producers not wanting to pay more to replicate in BD. If they do not want AACS and it is forced on them, then they are paying more for BD than they would for HD-DVD.

I never pretened anything regarding AACS. In this case, a small studio (relatively) wanted the cheapest alternative to get their product on the market. Because of BD forced fees and differences in manufacturing processes, BD was more expensive compared to HD-DVD.

Actually, I aced probability and statistics and my college senior capstone design project relied on statistics (the design of a pseudo-random noise generator whose gausian distribution could be biased with seed values). I also use probability to some extent in my current employment is determining the repeatability of measurements (simple statistics, but still statistics).

As I stated above, there is an inherent bias because the product is niche. I gave a few examples of sources for the bias, and there are more beyond what is listed. No one can say with any certainty how prevelent they are, or how much they might impact the final analysis.

Do you find fault with my comparing the Nielsen data and election precinct reporting? The analogy holds true. Each disc bought counts as a vote towards one format or the other (the difference being individuals can vote as many times as their wallet affords, not just once). Nielsen is asking stores (ala precincts) how many BD discs (votes) and how many HD-DVD discs (votes) they logged. They then compile all these stores (precincts) data to determine which format has the most sales (votes). However, in this case, Nielsen only acquires 40% of disc sales (votes), so they cannot determine definatively who is winning. Statistically, they might be able to say with a high probability, one format beats the other, if enough sales (votes) are reported. However, 40% reporting is not a large enough sample size to say with a high probability that one format is beating the other. You might be able to sell it to me if the number was 70-80%, but 40% is still way too low.

There is no established relation that says because Best Buy sales 2:1 for BD, that an untracked store sales 2:1 in the same ratio. Just because a Super Wal-Mart sells 3k jugs of milk a week (2k 1% and 1k skim), doesn't mean the Kroger next door also sales the same volume in the same ratio. The product would be the same both places, but pricing, availability, size of customer base, customer demographics, etc. all effect volume and ratio.

In a mainstream market, like DVD, less bias exists, and tracking methods like Nielsen, can produce results representative of the market. In a mainstream product, the prices are more regulated, the items are available in almost every outlet (especially new releases)(you can get DVD's at my local gas station, Rite Aid, Kroger, etc.), a performance history exists to help order adequate inventory, tacking services catch a higher percentage of the total sales, and so forth.

With a new and niche product like either HD format, the prices vary wildly, the discs are harder to find (the majority of outlets don't even carry them yet), some outlets carry only one format or the other, retailers lack a performance history to adequately order new releases (and stock previous releases), the total percentage of sales is lower, and so forth.

A model that works well for one type of product (mainstream) will not necessarily work well for another type of product (niche). Given the low volume, variations and potential sources of bias, a 40% sample size is not nearly large enough, to say with any high probability, that one format is performing better than the other.

cool! My masters thesis was on studying currency volatility. On the technical side I relied upon game theory and econometrics for my analysis :)

Icemage,

The point here is that a 40% sample of any sort is most likely to be representative unless you have a good reason to believe that the remaining 60% are behaving in a different way.

I will have to pick a nit here. The statement simply isnt true. As an extreme example, let us say I am interested in cognitive performance in the elderly. If I sample a bunch of college students, they wont be representative - even if I sample every college student in the world, the sample isnt representative.

plazman,

IF you have bias, then the sample is invalid

actually if you know the source of the bias, and now what you are doing, one can minimize the effects of the bias.

game theory

I like game theory. It is a lot of fun.

So, why not answer the question about whether your position is still that HD DVD has been selling more than 2 discs per player per month this year? If you think that the 310k supports that claim you made, then you must be using some pretty screwed up math. So, please show how it supports 2 discs per player per month. I'm sure others can see that it doesn't support the position you took and to now claim that it doesn't dispute anything you've said seems pretty disingenuous, at best.

--Darin

Did I not answer it already. My position hasn't changed that 925k discs were sold by studios in Jan/Feb. I don't know how many players are on the market, but if you guess at 250k, then that would roughly be 2:1.

However, in light of finding out that number also includes sales to rental outlets, then the attach rate may be less than 2:1 as rentals are not attached sales per se (actually, they could become that if the rental disc is sold at the end of it's rental period, so the water is muddy there).

I'm sorry if you are unable to comprehend my position. I don't know how to make it any clearer for you.

We can talk all we want about statistical accuracy from Nielsen, but the fact that they cover even 40% of the industry for standard DVD is a powerful statement (and I'll go with nataraj and say their representation may actually be better than that for high definition since there are far fewer significant sales venues for high def discs than for SD DVD).

The point here is that a 40% sample of any sort is most likely to be representative unless you have a good reason to believe that the remaining 60% are behaving in a different way.

In this case, you have consumers buying two competing formats which have the same properties pretty much whereever you go in the statistical geographic region (North America). While Kosty makes good points about it being possible that a 40%+ sample can give an incorrect projection, what this does not address is that what we are dealing with are consumers who are buying from largely unaligned retailers. They have the same general conditions for purchase; they want the same things, they get the same general pricing, etc.

Statistical inaccuracy from 40% from a non-random sampling matters little if retailers as a whole show similar sales trends. Given what information we've heard over the past several months, all of the evidence points to most retailers seeing similar changes in sales volumes. The effect might be more pronounced in one venue than another (particularly in the occasional cases where the retailer is biased), but the overall trend should remain the same.

I thought Nielsen captured 65+% of sales for DVD, and are capturing signifigantly less of the HD market.

From what we know, how can we verify that their sample is not biased? From what I know BB has been buying 4X as many BD titles than HD DVD - at least in my area. As a result some of the newer titles like Lucky #Slevin, The Departed and Clerks etc. were sold out in the week of their release, but they had plenty of BD release titles. So actions such as this will force people to buy from other sources....so IF BB is a significant part of the Neilson data, I would say it is biased. No doubt in my mind!
Again, we're back to what if scenarios with no proof.

I don't know what the BB's in your region are doing, but the ones near me all have the same shelf space for Blu-ray and HD DVD discs.

We can play what if games all day long, but accusing one retailer of preferential treatment doesn't get us anywhere since we have quite a lot of circumstantial evidence that Best Buy, by and large, remains neutral with respect to what's on the shelf. If there's bias there, it seems small, and does not really impact the totals unless the customers that would have bought from Best Buy would then take their business elsewhere that isn't covered by Nielsen.

Think about it: You can't buy it at Best Buy because they don't stock what you want. What do you do? Refuse to buy the title from inconvenience? No. You go to the next most likely place, or order it online. Either way, consumers will buy what they want, particularly since (as we know) the people who own a high def player of any type are largely early adopters, and have a reasonably good idea where to find what they need.

If you want to accuse Best Buy of preferential treatment, you might as well accuse Wal-Mart as well. There are a lot of Wal-Marts out there (including the ones near me) that seem to only stock a handful of high definition titles, mostly Blu-ray, and mostly keep them in the case with the PlayStation 3's.

If there's some sort of preferential treatment going on at a certain percentage of retailers, that preference is likely to show up at other retailers, for much the same reasons. These retailers are mostly neutral; most of them are going where they feel the market is shifting to. If HD DVD is selling particularly well in one area, I would expect the stores in that area to treat it preferentially, and vice versa. It all should come out in the wash when looking at the big picture; that's the idea behind statistical analysis.

I like game theory. It is a lot of fun.

Yup. I use it all the time. For once something I learnt in school has real world applicability :)

Sure. You can reweigh the model to correct for bias. But I can guarantee you that Videoscan has no way of doing that since they are probably simply reporting what goes through the cash register. So, IF I carry one copy of The Departed on HD DVD and 5 of the same on BD. One day later the HD DVD copy sells out and 3 days later the 5 BD copies sell out we end up with 5:1 sales ratio for BD. However, we have no idea how many HD DVD titles would have sold IF copies were available for that week....I've seen this so many times at my local BB that I've decided not to buy anything from them.

cool! My masters thesis was on studying currency volatility. On the technical side I relied upon game theory and econometrics for my analysis :)

Unfortunately my Capstone turned out to be rather useless (which wasn't my decision). It ended up used in a novelty product (a rather stupid one at that) that played noises (animal sounds, electronic tones, songs, etc.) at random intervals and in random durations. :(

Your studies at least seemed to have some relevance.

Again, we're back to what if scenarios with no proof.

I don't know what the BB's in your region are doing, but the ones near me all have the same shelf space for Blu-ray and HD DVD discs.

We can play what if games all day long, but accusing one retailer of preferential treatment doesn't get us anywhere since we have quite a lot of circumstantial evidence that Best Buy, by and large, remains neutral with respect to what's on the shelf. If there's bias there, it seems small, and does not really impact the totals unless the customers that would have bought from Best Buy would then take their business elsewhere that isn't covered by Nielsen.

Think about it: You can't buy it at Best Buy because they don't stock what you want. What do you do? Refuse to buy the title from inconvenience? No. You go to the next most likely place, or order it online. Either way, consumers will buy what they want, particularly since (as we know) the people who own a high def player of any type are largely early adopters, and have a reasonably good idea where to find what they need.

If you want to accuse Best Buy of preferential treatment, you might as well accuse Wal-Mart as well. There are a lot of Wal-Marts out there (including the ones near me) that seem to only stock a handful of high definition titles, mostly Blu-ray, and mostly keep them in the case with the PlayStation 3's.

If there's some sort of preferential treatment going on at a certain percentage of retailers, that preference is likely to show up at other retailers, for much the same reasons. These retailers are mostly neutral; most of them are going where they feel the market is shifting to. If HD DVD is selling particularly well in one area, I would expect the stores in that area to treat it preferentially, and vice versa. It all should come out in the wash when looking at the big picture; that's the idea behind statistical analysis.

I am really interested in knowing what sort of statistical methods Videoscan uses to eliminate bias.

Second, what you are saying is this - I can walk into a store and see what is going on there and extrapolate to the rest of the world, because unless I have proof that this one store is statistically different from the rest of the stores in the world that it reflects the behavior for all stores!!!!!

IF BB corporate buying habits are designed to ensure that we end up with one format (BD more likely since Sony is a key supplier and Toshiba is not) it would not introduce bias into the results? sure it would. Also, if they are capturing only 40% or so of the market, we have no way of knowing if people are substituting BB by going to one of those 60% sales not captured by Videoscan.

When gallup makes a prediction based on a 1,000 sample size it's not the same thing as me interviewing 1,000 colleagues at work and coming up with a prediction. Sampling just does not work that way. You can't have a biased sample and use that to extrapolate for the rest of the market. garbage in - garbage out.

Regarding the tables of projected Nielsen sales data that everyone likes to post here:

Shouldn't that table be corrected to show 124k YTD sales for HD-DVD and 280k YTD for BD. Basically, HMM is saying that that projected sales figures are 700k this year for BD and 310 k for HD-DVD. They say that Nielsen only captures ~40% of sales, and that they projected off of that, and then confirmed their projections.

So that would mean the Nielsen data told them 124k and 280k sales YTD. Since this is a thread to analyze Nielsen data (and not projections of Nielsen data), shouldn't the tables posted on this forum be corrected to reflect what Nielsen reports, and then users can use whatever percentage they justify to project estimated total sales. Seems like the charts before this article were showing numbers much high than what Nielsen tracked.

Again, we're back to what if scenarios with no proof.

I don't know what the BB's in your region are doing, but the ones near me all have the same shelf space for Blu-ray and HD DVD discs.

We can play what if games all day long, but accusing one retailer of preferential treatment doesn't get us anywhere since we have quite a lot of circumstantial evidence that Best Buy, by and large, remains neutral with respect to what's on the shelf. If there's bias there, it seems small, and does not really impact the totals unless the customers that would have bought from Best Buy would then take their business elsewhere that isn't covered by Nielsen.

Think about it: You can't buy it at Best Buy because they don't stock what you want. What do you do? Refuse to buy the title from inconvenience? No. You go to the next most likely place, or order it online. Either way, consumers will buy what they want, particularly since (as we know) the people who own a high def player of any type are largely early adopters, and have a reasonably good idea where to find what they need.

If you want to accuse Best Buy of preferential treatment, you might as well accuse Wal-Mart as well. There are a lot of Wal-Marts out there (including the ones near me) that seem to only stock a handful of high definition titles, mostly Blu-ray, and mostly keep them in the case with the PlayStation 3's.

If there's some sort of preferential treatment going on at a certain percentage of retailers, that preference is likely to show up at other retailers, for much the same reasons. These retailers are mostly neutral; most of them are going where they feel the market is shifting to. If HD DVD is selling particularly well in one area, I would expect the stores in that area to treat it preferentially, and vice versa. It all should come out in the wash when looking at the big picture; that's the idea behind statistical analysis.

You also have no proof to your claim that stores see an equal ratio of sales, because a disc is a disc.

Unfortunately my Capstone turned out to be rather useless (which wasn't my decision). It ended up used in a novelty product (a rather stupid one at that) that played noises (animal sounds, electronic tones, songs, etc.) at random intervals and in random durations. :(

Your studies at least seemed to have some relevance.

Interesting the most important thing that I found was that data from the IMF was highly suspect in many cases. Especially on measures on GDP growth and money supply and government spending...basically almost all of it. Hence one reason why I went from econometrics heavy to game theory for a lot of my work ;)

Counting accurately it appears is much harder than most of us expect it to be :)

Regarding the tables of projected Nielsen sales data that everyone likes to post here:

Shouldn't that table be corrected to show 124k YTD sales for HD-DVD and 280k YTD for BD. Basically, HMM is saying that that projected sales figures are 700k this year for BD and 310 k for HD-DVD. They say that Nielsen only captures ~40% of sales, and that they projected off of that, and then confirmed their projections.

So that would mean the Nielsen data told them 124k and 280k sales YTD. Since this is a thread to analyze Nielsen data (and not projections of Nielsen data), shouldn't the tables posted on this forum be corrected to reflect what Nielsen reports, and then users can use whatever percentage they justify to project estimated total sales. Seems like the charts before this article were showing numbers much high than what Nielsen tracked.


Not if it messes up those cute charts.

You know what seems really strange to me from those charts? That HD DVD and BD sales increases and decreases happen at the same time every week. Both go up and down like synchronized swimmers? Possible? yes. Likely? NO!!!!

Actually the lack of protests from any one at HD DVD may not mean much. There have been some comments about the numbers in interviews but there has not been very much official press or attempt at push back.

But that may be strategy and not entirely acknowledgment. First off, the studios will have better feelings of numbers later on. Secondly, it may be a better PR campaign to wait until HD DVD numbers have advanced so they can say HD DVD has climbed back and is now in control, and finally no one is really caring much about the 1st quarter numbers so why even attempt to quash them and either validate them or make them an issue of controversy.

Even if they were in error, a basic principle of PR is to hold you tongue when an attack is failing against you and is burning itself out. Even if you are being misrepresented, it is better to be silent and bite your tongue than be accurate and reignite a controversy.

Right now with HD DVD sales increasing and more HD DVD releases bending, it is better to be silent about the subject even if the talked about points are currently misrepresented ( not saying they are) or unfavorable to your side.

HD DVD PRG is just practicing good PR discipline and damage control. :)

You're on the right track.

Toshiba has not historically been a company to sling mud, or get into PR battles. On the other side, Sony excels at this.

Toshiba nows that if they fire back at Sony, Sony will fire back at them. They are also smart enough to realize their PR doesn't respond as well, or as quickly in mud smearing campaigns (as they have less experiece in that sort of thing). So why would they jump into one? I was actually shocked they called Sony's claims "propoganda". Even that is going out on a limb for Toshiba.

Also, Toshiba's philosophy is a bit different than Sony's, at the corporate level. My impression is that Toshiba believes it is to their advantage to let an opponent believe they have the lead, even if they do not. It can cause the oppoent to spend more to maintain the "lead", or on the other side of the coin, can cause them to slack off a bit when they should be trying even harder. What better way to beat your opponent in a race than to be in the lead and your opponent not even realize it?

I personally buy into their line of thinking, but I do think there is a fine line where that strategy can become damaging to Toshiba. Toshiba's success or failure may come from being able to recognize this line and react accordingly whenever it is approached.

3. The consumer will simply buy from a source that is not BB. The question is, are those alternate sources covered by Videscan as well or not!


Exactly. So, what other significant retailers are there that are not tracked ?

We need someone to show

- There are significant retailers missed by Nielsen
- These significant retailers somehow sell more of one format than the other

All other examples - including those of speed guns are of little use.

From what we know, how can we verify that their sample is not biased?

No, we can't. We take Nielsen numbers in good faith, like rest of the world.

It is incumbent on people who are claiming the Nielsen numbers are not representative to provide proof.

Top HD DVD Titles For Week Ended 3/25/2007

RANK TITLE (LABEL/DISTRIBUTOR, SRP)
1 The Departed (WB, $39.99)
2 Superman Returns (WB, $39.99)
3 Batman Begins (WB, $28.99)
4 Casino (UNI, $29.98)
5 Babel (PAR, $39.99)
6 Fear And Loathing In Las Vegas (UNI, $29.98)
7 The Mummy Returns (UNI, $29.98)
8 V For Vendetta (WB, $28.99)
9 The Mummy (UNI, $29.98)
10 The last Samurai (WB, $28.99)


Top Blu-Ray Titles For Week Ended 3/25/2007

RANK TITLE (LABEL/DISTRIBUTOR, SRP)
1 Casino Royale (SONY, $38.96)
2 Rocky Balboa (SONY, $38.96)
3 The Departed (WB, $34.99)
4 Eragon (FOX, $39.98)
5 The Prestige (BV, $34.99)
6 Black Hawk Down (SONY, $28.95)
7 The Fifth Element (SONY, $28.95)
8 House of flying Daggers (SONY, $28.95)
9 Layer Cake (SONY, $28.95)
10 X-Men: The last Stand (FOX, $39.98)

Source: Rentrak’s Retail Essentials ™.
Sales estimations are based on preliminary data provided through an exclusive arrangement with Rentrak Corp.’s Retail Essentials service. Point-of-Sale data is collected weekly and projected nationally for the U.S. bricks-and-mortar sales channel.

I think conditions for bias definately exist in the Nielsen data collection, but as time goes by, the conditions (and bias) stabilize. For example, Let's say you live in a town that only has a CC. Up until a month or so ago, CC only carried BD, so it was not possible to by an HD-DVD from them. Now let's say you did by an HD-DVD player. You would not be able to buy discs locally, so you turn to the net. Let's say you buy your discs at Deep Discount or any other online retailer that doesn't report to Videoscan. This puts a potential bias into the Videoscan numbers in favor of BD. However, as time has went by, CC carries both, and the bias caused by lack of neutrality with CC has been removed. There are other online and offline retailers that are not neutral, and therefore present a potential bias for consumers that shop in those areas.

Even a retailers lack of adequate inventory to bias the consumers true purchasing wishes. For example, let's say you really want to get Happy Feet in hi-def (a dual format release). Now lets say you are also dual format in your household. Let's say you want the disc on HD-DVD because the same review source gave it 5/5 and the BD version 4.5/5. You go to your local retailer (or a website you use often; ecost for example, because it has the BD, but not HD-DVD), and they don't have a HD-DVD copy in stock because of inadequte inventory purchase. But since you really want the film, and can play both, you buy the BD copy. BD just got a biased +1 (the true desire of the consumer was for HD-DVD) because of a retailers poor stock desisions in a niche product. This is a problem for niche products, which exists to a much lesser extent in mainstream products (most retailers probably got a bulk display with a 100 copies of the DVD).

There are plenty of potential sources of bias in a niche market, when using a model developed for a mainstream market. Further, it is potentially a false assumption to assume that every retailer tracked by Nielsen sales more BD than HD-DVD. BD sales more in total, but which format sales better retailer to retailer can differ. Further, it is a poor assumption to assume that because this 40% population of retailers is comprised primarily of retailers that sell more BD, that it must mean the same holds true for the other 60%. Its the same with elections. When you elect a Govenor, just because 40% of precincts have reported and show candidate X in the lead, it doesn't mean they other 60% of precincts will vote the same way. Candidate Y could actually turn out to be ahead and the winner, once the other 60% of precincts are accounted for. I agree with you here. :)

I think we did more than enough to prove at least the list is inaccurate, and potentially falsified. Surely you remember that discussion earlier in this thread. There are retailers on the list that have been out of business since 2000, yet the list is dated Nov and Feb 2006! If the list is merely out of date, I would say 6 years out of date is huge (assuming only the more recent 2006 dates were falsified)! Further, I contacted ACNielsen and they would not supply a list of retailers, and in a follow-up, would not say that this list was one of theirs. Further, they directed me to contact Videoscan with any further inquiry, and gave me completely different contact info than what is posted on Videoscan's site.

I also attempted to contact Videoscan using the information on their website, and using the information provided to me by ACNielsen. I never received any response to my inquiries from the contact information on the web. Using the contact info given me by ACNielsen, I was told that Videoscan contributors are protected by a NDA, and that a list of contributors could not be given out. The explained that only a few parties had access to a list of contributing members, and that each contributing member had to consent to allowing their name to be released to those select few (meaning you could submit data, but not consent to a release, which means your would contribute anonymous to everyone but Nielsen).

In light of everything uncovered about this "list", and the apparent difficulty to obtain such a list from Nielsen, I give it a credibility of zero... I don't agree with you on this one. :(

The list is given as a sample example. The current list maybe not available and the itemized discrete data from the submitting retailers is certainly under NDA.

Actually, I aced probability and statistics and my college senior capstone design project relied on statistics (the design of a pseudo-random noise generator whose gausian distribution could be biased with seed values). I also use probability to some extent in my current employment is determining the repeatability of measurements (simple statistics, but still statistics).

As I stated above, there is an inherent bias because the product is niche. I gave a few examples of sources for the bias, and there are more beyond what is listed. No one can say with any certainty how prevelent they are, or how much they might impact the final analysis.

Do you find fault with my comparing the Nielsen data and election precinct reporting? The analogy holds true. Each disc bought counts as a vote towards one format or the other (the difference being individuals can vote as many times as their wallet affords, not just once). Nielsen is asking stores (ala precincts) how many BD discs (votes) and how many HD-DVD discs (votes) they logged. They then compile all these stores (precincts) data to determine which format has the most sales (votes). However, in this case, Nielsen only acquires 40% of disc sales (votes), so they cannot determine definatively who is winning. Statistically, they might be able to say with a high probability, one format beats the other, if enough sales (votes) are reported. However, 40% reporting is not a large enough sample size to say with a high probability that one format is beating the other. You might be able to sell it to me if the number was 70-80%, but 40% is still way too low.

There is no established relation that says because Best Buy sales 2:1 for BD, that an untracked store sales 2:1 in the same ratio. Just because a Super Wal-Mart sells 3k jugs of milk a week (2k 1% and 1k skim), doesn't mean the Kroger next door also sales the same volume in the same ratio. The product would be the same both places, but pricing, availability, size of customer base, customer demographics, etc. all effect volume and ratio.

In a mainstream market, like DVD, less bias exists, and tracking methods like Nielsen, can produce results representative of the market. In a mainstream product, the prices are more regulated, the items are available in almost every outlet (especially new releases)(you can get DVD's at my local gas station, Rite Aid, Kroger, etc.), a performance history exists to help order adequate inventory, tacking services catch a higher percentage of the total sales, and so forth.

With a new and niche product like either HD format, the prices vary wildly, the discs are harder to find (the majority of outlets don't even carry them yet), some outlets carry only one format or the other, retailers lack a performance history to adequately order new releases (and stock previous releases), the total percentage of sales is lower, and so forth.

A model that works well for one type of product (mainstream) will not necessarily work well for another type of product (niche). Given the low volume, variations and potential sources of bias, a 40% sample size is not nearly large enough, to say with any high probability, that one format is performing better than the other. I agree with this one also.

But the accuracy of the Nielson/Videoscan model will approve as the product category goes from niche to mainstream.

Regarding the tables of projected Nielsen sales data that everyone likes to post here:

Shouldn't that table be corrected to show 124k YTD sales for HD-DVD and 280k YTD for BD. Basically, HMM is saying that that projected sales figures are 700k this year for BD and 310 k for HD-DVD. They say that Nielsen only captures ~40% of sales, and that they projected off of that, and then confirmed their projections.

So that would mean the Nielsen data told them 124k and 280k sales YTD. Since this is a thread to analyze Nielsen data (and not projections of Nielsen data), shouldn't the tables posted on this forum be corrected to reflect what Nielsen reports, and then users can use whatever percentage they justify to project estimated total sales. Seems like the charts before this article were showing numbers much high than what Nielsen tracked. I'm of the opinion that the HMM ratios that are being published are being based on the data available in the Nielson/Videoscan tracking data base.

That database is meant as a consistent device, like iSupplu evaluations in a way, but more limited to compare titles and retail channels over a more limited time period.

But it is not purported to be the entire market. It is not alleged to be the entire amount of sales for the entire US market. Its only what Nielson/Videoscan tracks.

Until I see a single piece of Nielson/Videoscan explanation that they adjust their collected figures to reflect the sales they do not collect by census or sampling (Wal-Mart and other niche retailers) I think that the numbers are representative but not complete for US sales.

How that affects Icemage's calculations here, since we are using various starting points for the data in EOY 2006 sales , I do not know. If we take Nielson EOY data as the start its one thing, but if we take someone else numbers, and then use Nielson after that for the number sales per week, its possible there is a short term bias as the numbers are small and the 40% data collection net may have some unknown quirks that may not project to the rest of the market.

A quirk like a niche retailer VE, sells a ton of HD DVD movies, cause he sells a ton of HD DVD players, but he sells no DVDs, and is completely un captured by Nielson. At the current volume of sales his thousand or more HD DVD sales per week are significant. As time goes on it becomes noise.

I am not sure if the rest of the world takes Videscan data as the gospel. We've all seen official folks quote data that does not line up with Videoscan. And it's not a secret that Videoscan does not account for a significant portion of sales and it is not a secret that BB is doing what it can to push BD. Todays USA Today has a full page HD DVD ad. BB is not mentioned as a retailer...

IF BB is doing 5:1 BD and overall is 3:1 BD and BB represents 30% of Videoscan data and we don't know the order and % sell thru for BB, I'm afraid we are being made a goat - and your pretty graph is a good attempt at SWAG.

IF BB is doing 5:1 BD and overall is 3:1 BD and BB represents 30% of Videoscan data and we don't know the order and % sell thru for BB, I'm afraid we are being made a goat - and your pretty graph is a good attempt at SWAG.

I'm not trying to convince anyone that our estimates are great. All I'm saying is nobody is giving rational reasons why the estimates are significantly off.

I'd take Nielsen data against wnorris "heard from someone" - any day of the week.

Sometime back when npd reported equal player sales data for HD DVD and BD I was one fo the first to challenge it. Giving examples of significant retailers (like VE and OnePass store) that were (probably) not being counted. Tosh exec has said something similar (200K vs 30K).

But I see no challenge to the ratios from anyone at all - that gives me confidence that these ratios are good.

I am really interested in knowing what sort of statistical methods Videoscan uses to eliminate bias.
Well if we knew that we'd be in their line of work. :)

Second, what you are saying is this - I can walk into a store and see what is going on there and extrapolate to the rest of the world, because unless I have proof that this one store is statistically different from the rest of the stores in the world that it reflects the behavior for all stores!!!!!
Not at all. What I am saying is that, in the absence of proof there is no way to prove retailer bias based on anecdotal observation such as you are suggesting. I am also saying that retailer bias such as you are suggesting does not reduce overall sales in the manner you suggest; it is only true when there are no other alternatives for sales venue.

If, as a consumer, I cannot buy what I want from Best Buy, I can buy it from another retailer, and I will shift my buying habits in the future to those other outlets.

IF BB corporate buying habits are designed to ensure that we end up with one format (BD more likely since Sony is a key supplier and Toshiba is not) it would not introduce bias into the results? sure it would. Also, if they are capturing only 40% or so of the market, we have no way of knowing if people are substituting BB by going to one of those 60% sales not captured by Videoscan.
Possible, but there's no proof of retailer bias on the part of Best Buy or any other retailer thus far. If the bias is slight, then the effect is statistically ignorable. If the bias is heavy, it should be obvious.

It should also be noted that the point of this thread is to extrapolate from Nielsen/VideoScan data. I think a lot of the arguments over the last few pages are from people who are questioning Nielsen's accuracy, which is fine as far as that goes, but if you're going to question what is effectively the gold standard in the industry, you need more than just "I think" or "what if?"

Put more simply: If you have better data, then present it. If we start questioning our fundamental sources, then we'd best put away our calculators, stop reading press releases, and just wait for the market to do whatever it wants in continued ignorance.

I've already stated this once in this thread but it bears repeating:

We're not saying that Nielsen is infallible; far from it, as we have proof that they can make mistakes from early this year. But saying that they "might" be inaccurate does not mean we can't draw meaningful conclusions from their published data.

That's all we're really trying to do. We are demystifying the numbers that Nielsen publishes, from their nebulous ratios into something more concrete that everyone can sink their teeth into as far as discussing where the current market is shifting to. We're not holding it forth as "this is definitely what is happening". We're presenting it as "this is our best guess at what Nielsen says is happening".

If you truly believe Nielsen is radically wrong, then frankly... stop reading this thread completely. Seriously. We're not going to come to a point of agreement on this, because those of us who believe they are the best available market indicator are evidently never going to convince the doubters amongst the fold who distrust everything.

If you truly believe Nielsen is radically wrong, then frankly... stop reading this thread completely. Seriously. We're not going to come to a point of agreement on this, because those of us who believe they are the best available market indicator are evidently never going to convince the doubters amongst the fold who distrust everything.

There needs to be two seperate threads regarding the Nielsen figures. One that posts the data and another for discussion of the numbers.

~Alan

Icemage,

If you truly believe Nielsen is radically wrong, then frankly... stop reading this thread completely. Seriously. We're not going to come to a point of agreement on this, because those of us who believe they are the best available market indicator are evidently never going to convince the doubters amongst the fold who distrust everything.

Good point and well said.

Questions that may have alreday been anwered - don't see them

1) Are there a significant number of HD-DVD/DVD combos that may be reported as DVD
2) Have the Neilson numbers (at least in scale) been verified by any independent, objective source
3) Do we have any idea what % of sales occur outside of Neilson purview
4) Could it be market is so small that Neilson techniques don't apply?

1) Are there a significant number of HD-DVD/DVD combos that may be reported as DVD

Sales figures include combos, as indicated on the graphs.

During the first two weeks of the year, they didn't include combos, but as of Jan 21, they do. As the figures are not week-by-week, but year-to-date and since-inception, the only effect of that omission is that the data for the first two weeks were not usable.

2) Have the Neilson numbers (at least in scale) been verified by any independent, objective source

Nielsen itself is fairly independent and objective. But the figures have been recently verified by the Home Media Magazine staff through the studios.

3) Do we have any idea what % of sales occur outside of Neilson purview

We know Nielsen covers 40 percent of the market, so it leaves out 60 per cent.

4) Could it be market is so small that Neilson techniques don't apply?

It is similar to small-run DVDs, which have been tracked by Nielsen for a long time.

A quirk like a niche retailer VE, sells a ton of HD DVD movies, cause he sells a ton of HD DVD players, but he sells no DVDs, and is completely un captured by Nielson. At the current volume of sales his thousand or more HD DVD sales per week are significant.Where did you get that number? I'm pretty sure he told me 200 discs per week. I would have to search for the post, but am wondering why you are say a thousand or more per week.

--Darin

darin,

you are right. I remember Robert say he shipped about 200 or so a week.

I am not sure if the rest of the world takes Videscan data as the gospel.


Just like nobody takes soundscan as gospel right? :rolleyes:

I'm not trying to convince anyone that our estimates are great. All I'm saying is nobody is giving rational reasons why the estimates are significantly off.

I'd take Nielsen data against wnorris "heard from someone" - any day of the week.

Sometime back when npd reported equal player sales data for HD DVD and BD I was one fo the first to challenge it. Giving examples of significant retailers (like VE and OnePass store) that were (probably) not being counted. Tosh exec has said something similar (200K vs 30K).

But I see no challenge to the ratios from anyone at all - that gives me confidence that these ratios are good.

Good and valid points :)

Just like nobody takes soundscan as gospel right? :rolleyes:

Not sure about soundscan....but does it cover more than 40% of the market?

Where did you get that number? I'm pretty sure he told me 200 discs per week. I would have to search for the post, but am wondering why you are say a thousand or more per week.

--Darin You are right, I meant 1000 per month as a rounded figure. I understand he does above 200 HD DVD sales per week by other communication, I did not see his post here that said 200.

My point was that if one guy was around 1% of the reported HD DVD sales, and he is not captured, if there are others like him or dozens at half his volume, that adds up to a lot at this level of sales, but becomes insignificant as volumes rises.

Also his sales is almost entirely HD DVD so not capturing him and other niche retailers like him is a bias.

My point was that if one guy was around 1% of the reported HD DVD sales, and he is not captured, if there are others like him or dozens at half his volume, that adds up to a lot at this level of sales, but becomes insignificant as volumes rises.

But VE is kind of unique. I don't think there are many who sell at his levels. I don't think they can really make up much considering 25K sales of HD DVD per week. Even if you had a dozen stores like his (which we don't) it is still just 2K and the overall BD to HD DVD ratio remains unperturbed.

It should also be noted that the point of this thread is to extrapolate from Nielsen/VideoScan data. I think a lot of the arguments over the last few pages are from people who are questioning Nielsen's accuracy, which is fine as far as that goes, but if you're going to question what is effectively the gold standard in the industry, you need more than just "I think" or "what if?"

Put more simply: If you have better data, then present it. If we start questioning our fundamental sources, then we'd best put away our calculators, stop reading press releases, and just wait for the market to do whatever it wants in continued ignorance.

I've already stated this once in this thread but it bears repeating:

We're not saying that Nielsen is infallible; far from it, as we have proof that they can make mistakes from early this year. But saying that they "might" be inaccurate does not mean we can't draw meaningful conclusions from their published data.

That's all we're really trying to do. We are demystifying the numbers that Nielsen publishes, from their nebulous ratios into something more concrete that everyone can sink their teeth into as far as discussing where the current market is shifting to. We're not holding it forth as "this is definitely what is happening". We're presenting it as "this is our best guess at what Nielsen says is happening".

If you truly believe Nielsen is radically wrong, then frankly... stop reading this thread completely. Seriously. We're not going to come to a point of agreement on this, because those of us who believe they are the best available market indicator are evidently never going to convince the doubters amongst the fold who distrust everything. I also think its perfectly valid here to talk about the significance of the Nielson/Videoscan numbers and that point counterpoint of that naturally leads to a discussion of the methodology and assumptions used by you to interpret them and of their own coverage and applicability to the rest of the sales that are not captured.

I think the Nielson numbers are a valid, indeed the best weekly sales indicator we have and are much more comprehensive that any of the Amazon or other minor internet sites we have. But our Nielson/Videoscan info we have through HMM and not subject to NDA is deliberately vague.

I look forward to see the methodology applied here followed throughout the year.

But I think discussion of why the numbers are this way and possible errors in the assumptions made are just as important as what the numbers are calculated out to be. ALso how we take those calculations and scale them up to the entire market is also important.

Just one example, if the SI number, which was beaten to death a Blu-ray milestone was only of the numbers that Nielson/Videoscan had captured, and did not reflect the entire market, then that figure becomes less significant in some ways as a nice pretty PR benchmark.

The gross sales benchmark has been used for Blu-ray PR because of it decisive 1st quarter sales ratios as published by HMM. If those ratios have been misleading because of systemic bias or volume quirks or have been misrepresented as total market sales when Nielson/Videoscan admits they are not, then that is a topic worthy of discussion.

I think the discussion overall in this thread has been useful and interesting, among the best currently on AVS on this broad subject and has a good balance.

Anyone want to come over to my place to have a six pack?

Exactly. So, what other significant retailers are there that are not tracked ?

We need someone to show

- There are significant retailers missed by Nielsen
- These significant retailers somehow sell more of one format than the other

All other examples - including those of speed guns are of little use.


We've already shown at least one example of that, DeepDiscount. They claim to sell more DVD's than DVD Empire (don't know about hi-def). They say they don't report to Nielsen. I haven't checked lately, but their Top 20 list usually is 12-15 HD-DVD's, seeming to indicate that format is favored there. DeepDiscount will also price match any online vendor to give you a cheaper total cost (sometimes it may only be a $.25 cheaper, but their service gets my business), which is why I buy most of my HD-DVD's there.

I'm not trying to convince anyone that our estimates are great. All I'm saying is nobody is giving rational reasons why the estimates are significantly off.

I'd take Nielsen data against wnorris "heard from someone" - any day of the week.

Sometime back when npd reported equal player sales data for HD DVD and BD I was one fo the first to challenge it. Giving examples of significant retailers (like VE and OnePass store) that were (probably) not being counted. Tosh exec has said something similar (200K vs 30K).

But I see no challenge to the ratios from anyone at all - that gives me confidence that these ratios are good.

Then why are you not using data points based on Nielsen and reducing the YTD projected volumes to 124k and 280k?

We've already shown at least one example of that, DeepDiscount.

What is their volume vis-a-vis total BL DVD volume ?

Then why are you not using data points based on Nielsen and reducing the YTD projected volumes to 124k and 280k?

Because it doesn't matter ! Finally we are looking at ratios and trends. Might as well go with our immediate source of data HMM.

I also think its perfectly valid here to talk about the significance of the Nielson/Videoscan numbers and that point counterpoint of that naturally leads to a discussion of the methodology and assumptions used by you to interpret them and of their own coverage and applicability to the rest of the sales that are not captured.

I think the Nielson numbers are a valid, indeed the best weekly sales indicator we have and are much more comprehensive that any of the Amazon or other minor internet sites we have. But our Nielson/Videoscan info we have through HMM and not subject to NDA is deliberately vague.

I look forward to see the methodology applied here followed throughout the year.

But I think discussion of why the numbers are this way and possible errors in the assumptions made are just as important as what the numbers are calculated out to be. ALso how we take those calculations and scale them up to the entire market is also important.

Just one example, if the SI number, which was beaten to death a Blu-ray milestone was only of the numbers that Nielson/Videoscan had captured, and did not reflect the entire market, then that figure becomes less significant in some ways as a nice pretty PR benchmark.

The gross sales benchmark has been used for Blu-ray PR because of it decisive 1st quarter sales ratios as published by HMM. If those ratios have been misleading because of systemic bias or volume quirks or have been misrepresented as total market sales when Nielson/Videoscan admits they are not, then that is a topic worthy of discussion.

I think the discussion overall in this thread has been useful and interesting, among the best currently on AVS on this broad subject and has a good balance.
Agreed on all points.

I'm not telling people not to question the data; that's normal. I have questioned the data myself at several points in this discussion and illustrate my points with possible scenarios when possible.

To keep the discussion relavent, though, we really need to cut down on the wild speculation. Yes, Nielsen isn't perfect. Yes, of course there are sales missing because they simply aren't aware of some sales channels. Yes, there might be bias in some of the sampling done. Does it matter?

Right now, what we're most interested in is whether:

(a) One format is selling more than another.
and
(b) Whether both formats are experiencing growth.

Answer to (a):
Even if there is a huge amount of missing data out there, I think it is a safe statement to say that Blu-ray is still outselling HD DVD in North America. There's no other way to interpret what we're seeing; even if the rest of the unseen ~60% of the market were tilted toward HD DVD (which is glaringly unlikely), it would not erase the 2:1 or higher sales ratios we are seeing.

Answer to (b):
This is a tougher question to answer. The honest truth is that I think both formats are treading water. Blu-ray has a large player base, but much of it is uncommitted. HD DVD has a stronger and more enthusiastic base, but much smaller (you could say one is a function of the other). Neither one is doing much to spur growth. Blu-ray players are still too expensive. HD DVD players are still too expensive. Software is still too expensive. The only inroads really being made on either front are by the consoles: The PS3 and the Xbox 360 HD DVD add-on.

Honestly the only difference between the performance of the two formats is that the larger install base of Blu-ray is overshadowing the more enthusiastic buying of HD DVD when there is a compelling reason for the Blu-ray consumers to take note (Casino Royale). Otherwise both formats experience lackluster sales on off weeks with no releases as you'd expect (since the content on both formats thus far is of limited quality), with perhaps a slight nod to Blu-ray since it takes longer for many of the Blu-ray buyers to key into new releases, as we see with the performance on Amazon of titles. Blu-ray titles rise, then sink slowly down. HD DVD spikes quickly on release day, but since the enthusiasm is higher, the residual sales are lower, and titles seem to drop faster.

Is this enough to keep the formats alive? I wonder. Certainly I think we are starting to see enough hard evidence that there is something impeding further growth. We can theorize that it is the uncertainty of the format war, overpriced software, overpriced hardware, or any combination thereof, but the truth seems inescapable: neither format is "healthy" for this stage development. Blu-ray has shown low net growth week-on-week, on a vector matching the sales of the PS3. HD DVD shows virtually no growth at all since the beginning of the year. Neither one is doing "well" IMO.

Maybe I'm wrong. Maybe it's post-holiday/tax season blues. Maybe people are just waiting for tax returns to buy in. All the promises of low priced players won't matter if retailer confidence drops to zero. All the promises of blockbuster releases won't matter if stores start pulling their inventory to free up floor space.

Is it all doom and gloom? No, of course not. I do think there is potential for both formats to come out swinging. HD DVD needs those cheaper players now (and no, $399 MSRP on the HD-A2 is a step in the right direction, but it still doesn't get anywhere near close enough to what I'd call a magic number). Blu-ray needs cheaper players now, too, and a price cut on the PS3 wouldn't be out of order at the first possible opportunity to open the console effect up. Software needs to come down in pricing. $40 MSRP is jus way too high for mass market. I don't care how cheap your players are if the brick and mortars are selling titles at those sorts of prices. Look at what happened when Fox unbent and dropped their title price for Amazon; huge influx of demand on titles that had only minimal interest previously.

I really think these things have to happen if either side wants their numbers to ramp up the way they've been predicting.

^^ Well said.

Because it doesn't matter ! Finally we are looking at ratios and trends. Might as well go with our immediate source of data HMM.

Of course it matters. If your volumes are off from previous weeks, and you apply the current ratios to previously incorrect data, you continue to get more incorrect data. Your methodology has a systemtic problem that results in compounded error week after week.

The article that stated 310k and 700k projections for YTD was HMM. Further, they said they used the number 40% to project from the Nielsen data to the numbers published in the article. So what they are saying is that roughly, Nielsen is reporting 124k and 280k YTD for their disc volumes, projected to 310k and 700k.

However, this contradicts everything you guys have done, so you want to throw it out the window and look the other way. The bias is starting to show...

If you were really just looking at ratios and trends, there would be no tables that try to tie those ratios to volumes of discs sold. So I don't agree that is all that is being done, and apparantly, being done incorrectly.

(a) One format is selling more than another.
and
(b) Whether both formats are experiencing growth.

Answer to (a):
Even if there is a huge amount of missing data out there, I think it is a safe statement to say that Blu-ray is still outselling HD DVD in North America. There's no other way to interpret what we're seeing; even if the rest of the unseen ~60% of the market were tilted toward HD DVD (which is glaringly unlikely), it would not erase the 2:1 or higher sales ratios we are seeing.

This is a silly statement. If the remaining 60% of the market was was tilted towards HD-DVD, it could/would certainly erase t he 2:1 sales ratio. We are talking roughly 400k discs counted by Nielsen to make up 40% (according to the HMM article) for 2007 YTD. This means 600k discs unaccounted for. If the ratio of that 600k was 1.5:1 in favor of HD-DVD, it would mean an actual market of 484k HD-DVD and 516 k BD, which results in nearly a 1:1 ratio.

Answer to (b):
This is a tougher question to answer. The honest truth is that I think both formats are treading water. Blu-ray has a large player base, but much of it is uncommitted. HD DVD has a stronger and more enthusiastic base, but much smaller (you could say one is a function of the other). Neither one is doing much to spur growth. Blu-ray players are still too expensive. HD DVD players are still too expensive. Software is still too expensive. The only inroads really being made on either front are by the consoles: The PS3 and the Xbox 360 HD DVD add-on.

Again, it is not possible to answer this question with any certainty while only 40% of sales are accounted for. A disc that is $35 at BB and $25 at Deep Discount will result in people not buying at BB and purchasing at DD. If the more expensive retailer is tracked by Nielsen and the less exensive not, then there is certainly more possibility for growth at the lower cost outlet, than the overprice one. Therefore growth can be easily masked buy pushing sales away from Nielsen outlets, to potential non-Nielsen outlets.

Of course it matters. If your volumes are off from previous weeks, and you apply the current ratios to previously incorrect data, you continue to get more incorrect data. Your methodology has a systemtic problem that results in compounded error week after week.

They are all ratios. Once the first two numbers are chosen rest follow by solving simple simulateous equations. They will all be equally "wrong".

In anycase, I'm sure you don't have an open mind. And you want to beleive only what you are told ...

This is a silly statement. If the remaining 60% of the market was was tilted towards HD-DVD, it could/would certainly erase t he 2:1 sales ratio. We are talking roughly 400k discs counted by Nielsen to make up 40% (according to the HMM article) for 2007 YTD. This means 600k discs unaccounted for. If the ratio of that 600k was 1.5:1 in favor of HD-DVD, it would mean an actual market of 484k HD-DVD and 516 k BD, which results in nearly a 1:1 ratio.



Again, it is not possible to answer this question with any certainty while only 40% of sales are accounted for. A disc that is $35 at BB and $25 at Deep Discount will result in people not buying at BB and purchasing at DD. If the more expensive retailer is tracked by Nielsen and the less exensive not, then there is certainly more possibility for growth at the lower cost outlet, than the overprice one. Therefore growth can be easily masked buy pushing sales away from Nielsen outlets, to potential non-Nielsen outlets.

You must wear a foil hat before going to sleep. These guys have been doing a steller job trying to keep track of sales. You are obviously hell bent on trying to discredit them without much basis for what you say.

If you don't like it, leave the thread.

These guys have been doing a steller job trying to keep track of sales. You are obviously hell bent on trying to discredit them without much basis for what you say.
I have to agree here. I don't post a lot but I lurk in this thread quite often. It just seems like wnorris's main goal here is to constantly discount anything that Icemage, nataraj, and the other's come up with. wnorris, if you simply want to state that the Nielsen numbers may not be accurate then you've done that several times over. It has been acknowledged that Nielsen probably isn't completely accurate and nobody with any common sense is taking these numbers as gospel. I think people constantly arguing about just how inaccurate the ratios are when nobody can prove anything doesn't help this thread. If anything, this thread is becoming more and more like the rest of the Blu-ray/HD DVD section, which I really hate to see happen. Can't we just let this thread be about the analysis of the only numbers we're given with the caveat that we know Nielsen's numbers aren't totally accurate?

I used to really love this thread. Thanks to all of those that put in the work to analyse the numbers that we have.

Not true!

1. IF you have bias, then the sample is invalid

2. IF BB does not carry stock, then it is underreporting demand, so it does NOT reflect the consumers will, but rather they wish to enforce their will on the consumer. Not the same thing...

3. The consumer will simply buy from a source that is not BB. The question is, are those alternate sources covered by Videscan as well or not!

4. BB may be trying to project a view of the market that they wish to see and if they have a significant share of the 40% can do that...

1. If the people reporting the data have bias, then the sample is invalid. If one of the data points is biased, then that bias should be normalized by increasing sample size.

2. If Best Buy doesn't stock something, it is probably because there is no demand for it. They are a business, not a movement. They will stock what sells.

3. Wouldn't this alternate source also sell Blu-ray? And therefore, Blu-ray sales also be underrepresented?

4. Conspiracy theories are not valid to this thread.

I hope we will get soon weekly data,
so it will be again nice discussion about ratios :) YTD and SI and month :)

Marek

1. If the people reporting the data have bias, then the sample is invalid. If one of the data points is biased, then that bias should be normalized by increasing sample size.

2. If Best Buy doesn't stock something, it is probably because there is no demand for it. They are a business, not a movement. They will stock what sells.

3. Wouldn't this alternate source also sell Blu-ray? And therefore, Blu-ray sales also be underrepresented?

4. Conspiracy theories are not valid to this thread.

You either didn't read what I wrote or understood it. Bu to your points:

1. Yes. But we know that Videoscan uses a fixed sample - the same sample every week. So any bias cannot be removed.

2. Best Buy is a business just like Disney and Universal are businesses. Yet, they are exclusive in their support. Are you saying that no one will buy a Disney movie on HD DVD and hence why they don't release on HD DVD? Same for Universal? Best Buy is the same way, they too are bound by politics, lobbying and other factors that go beyond simple customer demand for a single product. They could be looking at their relationship with Sony as a whole v just BD v. HD DVD.

3. True. And that is the crux of my point. The 60% that Videoscan does not cover. How does the ordering activities of one retailer that IS covered by Neilson affect retailers not covered, if that retailer does not stock a particular format or stocks less of one v. the other.

4. This is AV Science forum and that includes challenging conventional wisdom. The issues I raise have nothing to do with a conspiracy theory but rather with understanding how the business practices of one retailer who is reported can have an impact on retailers who are not reported and hence the 40% that is reported can be different from the 60% that is not.

Nataraj made some good and valid points. Both Nat and Ice have done a great job with the Neilson data and I am sure both recognie as well that these are not comprehensive statements about state of the format war....that's all!

New data in for the week ended March 25:

YTD BD 100.00 HD DVD 42.13
SI BD 100.00 HD DVD 79.77

Top 5 BD
1 Casino Royale 100.00
2. Rocky Balboa 62.66
3. Eragon 57.72
4. The Departed 19.73
5. The Prestige 9.48

Top 5 HD DVD
1. The Departed 100.00
2. Batman Begins 69.69
3. Babel 54.55
4. Troy 50.26
5. Goodfellas 38.45

New data in:

YTD BD 100.00 HD DVD 42.13
SI BD 100.00 HD DVD 79.77


For what week and what's the source?

For what week and what's the source?

Last week. Source Nielsen VideoScan (unofficial as of now ;) )


Day Week YTD SI
01/07 63.3/36.7 n/a n/a
01/14 68.2/31.8 n/a n/a
01/21 67.8/32.2 66.4/33.6 45.1/54.9
01/28 68.8/31.2 67.0/33.0 46.7/53.3
02/04 69.0/31.0 67.4/32.6 48.1/51.9
02/11 69.6/30.4 67.7/32.3 49.3/50.7
02/18 65.0/35.0 67.4/32.6 50.3/49.7
02/25 68.5/31.5 67.4/32.6 51.5/48.5
03/04 65.7/34.3 67.2/32.8 52.2/47.8
03/11 n/a 67.9/32.1 52.8/47.2
03/18 n/a 69.2/30.8 54.3/45.7
03/25 n/a 70.4/29.6 55.6/44.4

For what week and what's the source?

I think it's for week ending March 25

New data in for the week ended March 25:

YTD BD 100.00 HD DVD 42.13
SI BD 100.00 HD DVD 79.77

YTD BD 70.35% HD DVD 29.65%
SI BD 55.62% HD DVD 44.38%

I have to agree here. I don't post a lot but I lurk in this thread quite often. It just seems like wnorris's main goal here is to constantly discount anything that Icemage, nataraj, and the other's come up with. wnorris, if you simply want to state that the Nielsen numbers may not be accurate then you've done that several times over. It has been acknowledged that Nielsen probably isn't completely accurate and nobody with any common sense is taking these numbers as gospel. I think people constantly arguing about just how inaccurate the ratios are when nobody can prove anything doesn't help this thread. If anything, this thread is becoming more and more like the rest of the Blu-ray/HD DVD section, which I really hate to see happen. Can't we just let this thread be about the analysis of the only numbers we're given with the caveat that we know Nielsen's numbers aren't totally accurate?

I used to really love this thread. Thanks to all of those that put in the work to analyse the numbers that we have.

My goal is to try to get them to not misrepresent the Nielsen numbers and misinform the folks reading this thread.

If, as nat said, this forum was just to look at Nielsen sales ratios and trend their performance, then all that would be required is the weekly published sales ratios, and an analysis of only the sales ratios.

However, they want to go one step farther and attempt to speculate on the volume of discs sold, buy guessing at key figures. Now HMM, the source of Nat and Ice's Nielsen sales ratios, has printed an article that contradicts their speculation on volumes that have been done in this thread. They clearly indicate ~124k and ~280k discs sold YTD, according to Nielsen. However, Nat and Ice's figures for YTD are ~280k and ~630k. They are off by a factor 2.25! This obviously shows a serious error in their method of calculating these projections.

I don't have to discredit them, HMM has already done that. Now the question remains, does their method work if they lower the volumes to levels reported by HMM. If it doesn't, then there is a flaw in the method somewhere that needs to be corrected.

I have no problem if everyone wants this thread to strictly be an analysis of Nielsen numbers. But if this type of analysis is going to be presented to the public, then it must at least be done as accurately as possible, not with an error of 2.25X. Otherwise it is just misinformation...

Day Week YTD SI
01/07 63.3/36.7 n/a n/a
01/14 68.2/31.8 n/a n/a
01/21 67.8/32.2 66.4/33.6 45.1/54.9
01/28 68.8/31.2 67.0/33.0 46.7/53.3
02/04 69.0/31.0 67.4/32.6 48.1/51.9
02/11 69.6/30.4 67.7/32.3 49.3/50.7
02/18 65.0/35.0 67.4/32.6 50.3/49.7
02/25 68.5/31.5 67.4/32.6 51.5/48.5
03/04 65.7/34.3 67.2/32.8 52.2/47.8
03/11 n/a 67.9/32.1 52.8/47.2
03/18 n/a 69.2/30.8 54.3/45.7
03/25 n/a 70.4/29.6 55.6/44.4

New data in for the week ended March 25:

YTD BD 100.00 HD DVD 42.13
SI BD 100.00 HD DVD 79.77

Using ice's initial numbers. Looks like BR continues to benefit from CR. Next week will be interesting considering Amazon rankings.

http://img402.imageshack.us/img402/8662/nielsenanalysiscw6.gif (http://imageshack.us)

What is their volume vis-a-vis total BL DVD volume ?


What is the volume of Amazon, or DVD Empire, or any retailer claimed to be tracked by Nielsen?

All I know is their claim that they sell more non-porn DVD's than DVD Empire. So if we want to say DD volume of hi-def material is too small to matter, then likewise, we would have to conclude that DVD Empire's is too small as well, which means they BD vs. HD-DVD numbers are meaningless.

Using ice's initial numbers. Looks like BR continues to benefit from CR. Next week will be interesting considering Amazon rankings.

http://img402.imageshack.us/img402/8662/nielsenanalysiscw6.gif (http://imageshack.us)


Someone else pointed out how odd it was that every week, if BD projected volume goes up, HD-DVD goes up. If BD goes down, HD-DVD goes down.

Does this seem reasonable over 10 weeks of data? Shouldn't we see at least one week where BD goes up and HD-DVD goes down, or vice versa?

...which means they BD vs. HD-DVD numbers are meaningless.

If you think because nielsen doesn't capture deepdiscount their numbers are meaningless, good luck with that.

Someone else pointed out how odd it was that every week, if BD projected volume goes up, HD-DVD goes up. If BD goes down, HD-DVD goes down.

Does this seem reasonable over 10 weeks of data? Shouldn't we see at least one week where BD goes up and HD-DVD goes down, or vice versa?
That is odd actually . .. .. .

Someone else pointed out how odd it was that every week, if BD projected volume goes up, HD-DVD goes up. If BD goes down, HD-DVD goes down.

Does this seem reasonable over 10 weeks of data? Shouldn't we see at least one week where BD goes up and HD-DVD goes down, or vice versa?

That is a good question to talk about.

Using ice's initial numbers. Looks like BR continues to benefit from CR. Next week will be interesting considering Amazon rankings.

http://img402.imageshack.us/img402/8662/nielsenanalysiscw6.gif (http://imageshack.us)

I think it will be interesting to see what the sales ratio will be. For all 3 weeks without new HD DVD releases, BD has maintained a 4:1 sales ratio. It will be interesting to see if this goes back to 2:1 or maybe split the difference.

I think it will be interesting to see what the sales ratio will be. For all 3 weeks without new HD DVD releases, BD has maintained a 4:1 sales ratio. It will be interesting to see if this goes back to 2:1 or maybe split the difference.

there's about a ZERO chance of them going even, since (1) the sales rankings in amazon are being heavily influenced by pre-orders of The Matrix (HD-DVD), and (2) it seems sales rankings that are not very high (say, less than #15 or #10) do not reflect actual sales very well, probably because the actual sales numbers for lower rankings are relatively low in amazon.com. Remember how the Departed HD-DVD was also slightly ahead of The departed Blu-ray for several days. then it turns out the blu-ray version actually sold significanty more than the Hd-dvd version.

Agreed on all points.

I'm not telling people not to question the data; that's normal. I have questioned the data myself at several points in this discussion and illustrate my points with possible scenarios when possible.

To keep the discussion relavent, though, we really need to cut down on the wild speculation. Yes, Nielsen isn't perfect. Yes, of course there are sales missing because they simply aren't aware of some sales channels. Yes, there might be bias in some of the sampling done. Does it matter?

Right now, what we're most interested in is whether:

(a) One format is selling more than another.
and
(b) Whether both formats are experiencing growth.

Answer to (a):
Even if there is a huge amount of missing data out there, I think it is a safe statement to say that Blu-ray is still outselling HD DVD in North America. There's no other way to interpret what we're seeing; even if the rest of the unseen ~60% of the market were tilted toward HD DVD (which is glaringly unlikely), it would not erase the 2:1 or higher sales ratios we are seeing.

Answer to (b):
This is a tougher question to answer. The honest truth is that I think both formats are treading water. Blu-ray has a large player base, but much of it is uncommitted. HD DVD has a stronger and more enthusiastic base, but much smaller (you could say one is a function of the other). Neither one is doing much to spur growth. Blu-ray players are still too expensive. HD DVD players are still too expensive. Software is still too expensive. The only inroads really being made on either front are by the consoles: The PS3 and the Xbox 360 HD DVD add-on.

Honestly the only difference between the performance of the two formats is that the larger install base of Blu-ray is overshadowing the more enthusiastic buying of HD DVD when there is a compelling reason for the Blu-ray consumers to take note (Casino Royale). Otherwise both formats experience lackluster sales on off weeks with no releases as you'd expect (since the content on both formats thus far is of limited quality), with perhaps a slight nod to Blu-ray since it takes longer for many of the Blu-ray buyers to key into new releases, as we see with the performance on Amazon of titles. Blu-ray titles rise, then sink slowly down. HD DVD spikes quickly on release day, but since the enthusiasm is higher, the residual sales are lower, and titles seem to drop faster.

Is this enough to keep the formats alive? I wonder. Certainly I think we are starting to see enough hard evidence that there is something impeding further growth. We can theorize that it is the uncertainty of the format war, overpriced software, overpriced hardware, or any combination thereof, but the truth seems inescapable: neither format is "healthy" for this stage development. Blu-ray has shown low net growth week-on-week, on a vector matching the sales of the PS3. HD DVD shows virtually no growth at all since the beginning of the year. Neither one is doing "well" IMO.

Maybe I'm wrong. Maybe it's post-holiday/tax season blues. Maybe people are just waiting for tax returns to buy in. All the promises of low priced players won't matter if retailer confidence drops to zero. All the promises of blockbuster releases won't matter if stores start pulling their inventory to free up floor space.

Is it all doom and gloom? No, of course not. I do think there is potential for both formats to come out swinging. HD DVD needs those cheaper players now (and no, $399 MSRP on the HD-A2 is a step in the right direction, but it still doesn't get anywhere near close enough to what I'd call a magic number). Blu-ray needs cheaper players now, too, and a price cut on the PS3 wouldn't be out of order at the first possible opportunity to open the console effect up. Software needs to come down in pricing. $40 MSRP is jus way too high for mass market. I don't care how cheap your players are if the brick and mortars are selling titles at those sorts of prices. Look at what happened when Fox unbent and dropped their title price for Amazon; huge influx of demand on titles that had only minimal interest previously.

I really think these things have to happen if either side wants their numbers to ramp up the way they've been predicting. I think this assessment is accurate for what we have seen in the first quarter, but overly pessimistic for the rest of the year.

In particular, for HD DVD ( which I more about) the April 1st price reductions and increased marketing and advertising effort along with more inventory will dramatically increase HD A2 sales, and the normal seasonal increase in CE should dramatically boost them as well. In short Toshiba and the HD DVD PRG has keep their powder dry and has not wasted resources during the 1st quarter. Universal in a way has done so as well. Despite some pressure IMHO they have been sticking to their plan. Now that springtime is here and the birds are a chirping, its time to start spending their war chest.

In the past couple days, we've see the results of what a couple ne titles releases can do for the HD DVD Amazon charting. We are just now starting to see HD DVD releases after months of drought.

In other words, HD DVD is just starting to move. The HD DVD strategy is also about getting mass sales of players into the market sooner by dropping player price points. We are just starting to see those effects.

On the Blu-ray side, those PS3 sales will start to accumulate and we have a potential large numbers of hits to be released. Plenty of room for growth there too.

I just think the assessment above is too pessimistic as it is trying to project the 1st quarter over the entire year and HD players are accumulating in peoples hands and content is yet to be released.

I look forward to what the next months bring. :)

Of course it matters. If your volumes are off from previous weeks, and you apply the current ratios to previously incorrect data, you continue to get more incorrect data. Your methodology has a systemtic problem that results in compounded error week after week.

The article that stated 310k and 700k projections for YTD was HMM. Further, they said they used the number 40% to project from the Nielsen data to the numbers published in the article. So what they are saying is that roughly, Nielsen is reporting 124k and 280k YTD for their disc volumes, projected to 310k and 700k.

However, this contradicts everything you guys have done, so you want to throw it out the window and look the other way. The bias is starting to show...

If you were really just looking at ratios and trends, there would be no tables that try to tie those ratios to volumes of discs sold. So I don't agree that is all that is being done, and apparantly, being done incorrectly.

The article that stated 310k and 700k projections for YTD was HMM. Further, they said they used the number 40% to project from the Nielsen data to the numbers published in the article. So what they are saying is that roughly, Nielsen is reporting 124k and 280k YTD for their disc volumes, projected to 310k and 700k. I may agree with this point. I am not sure. But if that is accurate and the YTD numbers are actually a hard 124K and 280K in the Nielson database YTD, thats a 156K delta between HD DVD and Blu-ray over a three month period with their census and sampling. Thats only around 50K per month or 12K per week delta.

IF ( and I am not saying there is) there was a systemic bias ( no small niche retailers like VE, overemphasis on BB etc) a smaller delta between the formats makes the effects of these possible systemic issues magnified.

Besides the fact that the BDA and Sony have tried hard to increase and brag about those 1st quarter stats for PR purposes.

As volumes increase in the 2nd quarter those kinds of issues work become less significant and the analysis here will become more accurate and representative of the total sales picture.

Again, I think the analysis here is sound, its a quaestion of whether the volumes involved and projectioning these ratios to the rest of the unseen sales that I think might have a short term sensitivity problem.

In particular, for HD DVD ( which I more about) the April 1st price reductions and increased marketing and advertising effort along with more inventory will dramatically increase HD A2 sales, and the normal seasonal increase in CE should dramatically boost them as well.

To me the question is - what is the demand for BL DVD players. What does the demand-price curve for that look like. Is there a early-adopter / videophile only demand ... or is it deeper ?

For eg. SACD/DVD-A even when they dropped heavily in price failed to take off - because there never was much demand. We know HD demand is better than HiFi - but how much exactly - we don't know.

Here's an updated chart based on the new numbers. I base my chart on slightly lower start numbers for both formats (BD: 355,698, HD: 555,778).

http://75.126.103.40/images/guy03/nielsenmarch25.jpg

My goal is to try to get them to not misrepresent the Nielsen numbers and misinform the folks reading this thread.

If, as nat said, this forum was just to look at Nielsen sales ratios and trend their performance, then all that would be required is the weekly published sales ratios, and an analysis of only the sales ratios.

However, they want to go one step farther and attempt to speculate on the volume of discs sold, buy guessing at key figures. Now HMM, the source of Nat and Ice's Nielsen sales ratios, has printed an article that contradicts their speculation on volumes that have been done in this thread. They clearly indicate ~124k and ~280k discs sold YTD, according to Nielsen. However, Nat and Ice's figures for YTD are ~280k and ~630k. They are off by a factor 2.25! This obviously shows a serious error in their method of calculating these projections.

I don't have to discredit them, HMM has already done that. Now the question remains, does their method work if they lower the volumes to levels reported by HMM. If it doesn't, then there is a flaw in the method somewhere that needs to be corrected.

I have no problem if everyone wants this thread to strictly be an analysis of Nielsen numbers. But if this type of analysis is going to be presented to the public, then it must at least be done as accurately as possible, not with an error of 2.25X. Otherwise it is just misinformation... I would like to see a response to this comment also.

Where is the link to the HMM YTD figures of 124k and 280K?

Someone else pointed out how odd it was that every week, if BD projected volume goes up, HD-DVD goes up. If BD goes down, HD-DVD goes down.

Does this seem reasonable over 10 weeks of data? Shouldn't we see at least one week where BD goes up and HD-DVD goes down, or vice versa? that usually indicates a modeling issue and a sensitivity to an external events because of volume ( 5 week months , holidays, weather) or some issue in timely reporting that causes data from one week to bunch into the next week.

It means those factors are a larger consideration in the sales volume that is actually affecting the sales together more than raw differences between the objects being measured. Those external factors are overwhelming and affecting both volumes at the same time.

Which he evidently was confused about, because to the best of my knowledge there were no free movie vouchers...just $10 rebate slips requiring you to cut the sleeves to send in to redeem...
Yeah, sure sounds like Toshiba is immune from propaganda. :rolleyes:

If you think because nielsen doesn't capture deepdiscount their numbers are meaningless, good luck with that.

I never said that, but nice try...

I believe I was one (I am sure there are others) that both formats are lock step synchonized week after week in going up and down. For me this was one red flag that something fishy is going on with the data collection. No. not a conspiracy but it seems odd to be sure....

That is odd actually . .. .. .

Why is it odd that the rises and falls of BD and HD DVD are in synch?

I think we should tie in the rise and fall of DVD and CD sales to see if they are out of sync with the HD market. I wouldn't be surprised to see them move in almost exactly the same way.

I think this assessment is accurate for what we have seen in the first quarter, but overly pessimistic for the rest of the year.

I just think the assessment above is too pessimistic as it is trying to project the 1st quarter over the entire year and HD players are accumulating in peoples hands and content is yet to be released.

I look forward to what the next months bring. :)
Agreed, and I do think I am being too pessimistic, but if a $399 price point was going to convince people to buy into HD DVD in noticeably accelerated numbers, the disc sales are not showing it (this might, however, be a function of the X number of free discs offered with each player purchase).

For the PS3, it continues to have woes in the sales arena against its primary competitor, the Xbox 360. Sony needs to let Europe settle down and get a lower price tag out there ASAP if they want to stay in the console game IMO, regardless of new titles, and the standalone prices need to continue to fall to get within grasping distance of HD DVD.

Using ice's initial numbers. Looks like BR continues to benefit from CR. Next week will be interesting considering Amazon rankings.

http://img402.imageshack.us/img402/8662/nielsenanalysiscw6.gif (http://imageshack.us)

If wnorris is correct and Nielson is collecting only 40% of the sales, should not the assumed in yellow (assumed sales as of 1/1/07) be actually on the "Nielson chart" be based off of what we should assume Nielson capture of those 1 M sales?

ie, that would give us 40% of the numbers on the rest of the table, but that would then show the actual Nielson sampling and census deltas. So that would put into perspective that Nielson is only showing say a 40% of 25,000 = 10,000 unit delta each week on average.

Of course, we could then scale that chart up to what we have now to show a total market sales projection identical to the above chart. Or we could add columns showing that smaller Nielson collection scaled up, but by showing actual smaller numbers we could see that the volumes are actually so small that any quirks are expanded in the larger market projection.

Again, if the collected numbers were only 40% of the 2006 start numbers, then the chart is implying that those are hard collected Nielson/Videoscan numbers instead of the (appropriately) adjusted assumed total market figure.

Its like the chart columns should not be named Nielson YTD or SI, but Total Market projections based on the scaling up of the Nielson numbers.

Its like the calculated exact numbers are based on 40% hard census and survey and 60% projection based on scaling up the reported numbers.

Am I making sense to anyone here with this post?

Looking at monthly totals (2nd week of data for the month through first week of the following month) proves interesting as well:


Month BD HD
January 306,638 148,351
February 268,692 132,007
March* 195,646 44,454


We still have one more week for data, but it appears that sales are actually decreasing on a monthly basis.

I would like to see a response to this comment also.

Where is the link to the HMM YTD figures of 124k and 280K?


No direct link, but this article says 700k BD and 310k HD-DVD.

http://www.homemediamagazine.com/news/html/breaking_article.cfm?article_id=10441

Grubert did a follow-up to confirm this number and was told by HMM:


Since Nielsen only represents about 40% of the business the numbers have been bumped up accordingly and then cross checked with studios to ensure accuracy. And yes the numbers are thru the 18th.

This confirms the 700k and 310k number were projections to cover the entire market, not just the outlets that report to Nielsen. The Nielsen numbers were bumped up to 700k and 310k, using "about 40%" as the estimate for Nielsen market coverage.

http://www.avsforum.com/avs-vb/showthread.php?t=798080&page=115&pp=30

This clearly indicates HMM is seeing Nielsen numbers around 124k and 280k (maybe give or take a few % for the "about" part of 40%). This is a far cry though from the sales volumes projected from Nielsen data on these forums, and seems to indicate a problem with what is being published here.

Also, as I recall, the model being used by Ice and Nat did not work with lower starting values. It generated negative volumes. So if the 2006 starting value cannot be lowered to produce 124k and 280k YTD, as indicated by HMM, without generating negative weekly volumes, then it would indicate there is an error in their method.

This issue needs to be addressed, or the "science" portion of this thread goes out the window...

Am I making sense to anyone here with this post?

Nielson projects their data, so they are already making up for the fact that they don't capture 100% of the market.

Agreed, and I do think I am being too pessimistic, but if a $399 price point was going to convince people to buy into HD DVD in noticeably accelerated numbers, the disc sales are not showing it (this might, however, be a function of the X number of free discs offered with each player purchase).

For the PS3, it continues to have woes in the sales arena against its primary competitor, the Xbox 360. Sony needs to let Europe settle down and get a lower price tag out there ASAP if they want to stay in the console game IMO, regardless of new titles, and the standalone prices need to continue to fall to get within grasping distance of HD DVD. The affect of the amount of HD DVD players sold may have been muted in the past because of the number of titles being released.

The $399 MSRP of HD DVD players with street prices below that will sell more players. In combination with 2nd quarter seasonal buying patterns and more HD DVDs being released may very well lead to a sustained increase in HD DVD disc sales. ;)

Here is the link:

http://www.nxtbook.com/nxtbooks/questex/hom040107/index.php?startpage=2

If wnorris is correct and Nielson is collecting only 40% of the sales, should not the assumed in yellow (assumed sales as of 1/1/07) be actually on the "Nielson chart" be based off of what we should assume Nielson capture of those 1 M sales?

ie, that would give us 40% of the numbers on the rest of the table, but that would then show the actual Nielson sampling and census deltas. So that would put into perspective that Nielson is only showing say a 40% of 25,000 = 10,000 unit delta each week on average.

Of course, we could then scale that chart up to what we have now to show a total market sales projection identical to the above chart. Or we could add columns showing that smaller Nielson collection scaled up, but by showing actual smaller numbers we could see that the volumes are actually so small that any quirks are expanded in the larger market projection.

Again, if the collected numbers were only 40% of the 2006 start numbers, then the chart is implying that those are hard collected Nielson/Videoscan numbers instead of the (appropriately) adjusted assumed total market figure.

Its like the chart columns should not be named Nielson YTD or SI, but Total Market projections based on the scaling up of the Nielson numbers.

Its like the calculated exact numbers are based on 40% hard census and survey and 60% projection based on scaling up the reported numbers.

Am I making sense to anyone here with this post?

Yes. I understand what you are saying. My understanding is that the Nielsen SI and YTD volumes on that table, are derived out only using Nielsen values. Therefore, based on this table, if you wanted to project for the whole market, you would take 771k/.4 (for BD in this example), which would give you nearly two millions projected disc sales for the BD total market YTD. This number is wildly different than what HMM would indicate. So you either must say that HMM is wrong, or the tables in this forum are wrong. In which case, I would say HMM has more credibility, as they have access to actual sales numbers from Nielsen, not just ratios.

Nielson projects their data, so they are already making up for the fact that they don't capture 100% of the market.


Everything I've seen says Nielsen does not project their data.

Nielson projects their data, so they are already making up for the fact that they don't capture 100% of the market. I thought so at first , but now I think that is NOT the case. There is nothing from Nielson that says that.

I think people have assumed that they are a total market capture, but I don't think that is the case. I think the raw numbers and detailed NDA data includes only there census and sampling items.

Can anybody find anything that says different?

If wnorris is correct and Nielson is collecting only 40% of the sales, should not the assumed in yellow (assumed sales as of 1/1/07) be actually on the "Nielson chart" be based off of what we should assume Nielson capture of those 1 M sales?

What we have been doing all along is to use the latest "official" figures and use the "official" ratios to get meaningful numbers. The current numbers take projected HMM numbers.

For me this was one red flag that something fishy is going on with the data collection.

I'm not sure what could happen with data collection that would cause that. it's not just nielsen, both HD DVD and Blu-ray have had similar paths on amazon's rankings too.

I initially felt something wasn't right but it occurs to me that the factors affecting one affect the other too. Sales of HD media are rising and falling as they would with only one format. The same trends are happening now but of course we still see the split in sales.

As pointed out, I bet DVD's follow a similar pattern, except their numbers will simply be higher. There are little things that would make some difference, like certain titles being released. But my guess is, if someone wanted to buy a film on disc they would even if title x isn't on sale. All title x does is take sales from title y and z but doesn't necessarily change the market

Back to this week. Very impressive numbers again from Blu-ray. My feeling is things will plateuax very soon. I can't see them going above 72% (2.5:1) because the weekly numbers don't seem to and if BD isn't getting 72+ in a week, they can't change the YTD.

I wonder when/if we'll see worldwide figures soon. Now that the PS3 has launched everywhere (I think) they must be getting closer to 80% of the world market.

I thought so at first , but now I think that is NOT the case. There is nothing from Nielson that says that.

I think people have assumed that they are a total market capture, but I don't think that is the case. I think the raw numbers and detailed NDA data includes only there census and sampling items.

Can anybody find anything that says different?

Nielson data wouldn't be useful if they did not project. Does anyone truly believe that when they say 6M people watched the Superbowl, that they had boxes in 6M+ homes? Of course not. They project all their data. Its what they (and any other survey company) do.

Using ice's initial numbers. Looks like BR continues to benefit from CR. Next week will be interesting considering Amazon rankings.

http://img402.imageshack.us/img402/8662/nielsenanalysiscw6.gif (http://imageshack.us)

I think this chart should have lower starting values for 2006. Low enough to yield 124k for HD-DVD and 280k for BD, YTD through 3-18. This will allow it to match HMM's statements on Nielsen numbers. If the method works and doesn't yield negative volumes, then it is probably a sound method to use.

Then people can use whatever market coverage % they want (with about 40% being indicated by HMM) on the projected volumes. So the estimated volumes will be projected volumes based on Nielsen monitored volumes.

Basically, a column for estimated volumes (ie Nielsen, 124k and 280 k on 3/18), and then a 2nd "projected" column can be added that uses the 40% market coverage to project the total market volume.

Also, what is the purpose of a 64:100 market share on the starting number. This seems way out of wack with the very next SI market share. Is this just another fudge factor to make the results turn out a certain way? I never noticed it before. Did it really jump from 64 to 82 in two weeks, while during those same two weeks it was discovered that not all HD-DVD sales were being counted. The best jump for SI was about 5 points in a week, but this would indicate 9 points per week, which seems off.

What we have been doing all along is to use the latest "official" figures and use the "official" ratios to get meaningful numbers. The current numbers take projected HMM numbers.

No, that isn't the case. You guys were just lucky that your projected numbers somewhat correponded to what Nielsen said. If your method truely works, you should be able to enter a lower 2006 starting value, to yield disc sales with YTD volumes of 124k and 280k through 3/18. No negative volumes should be generated by your method.

This would be the litmus test of your conversion of market share ratio to disc volume.

Nielson data wouldn't be useful if they did not project. Does anyone truly believe that when they say 6M people watched the Superbowl, that they had boxes in 6M+ homes? Of course not. They project all their data. Its what they (and any other survey company) do.


HMM, this threads source of Nielsen data, disagrees with your statement. They claim Nielsen only captures 40% of the market, reports on that 40%, and then outside parties must perform and validate their own projections.

I've always been told that Nielsen only compiles barcode scans and reports the actual number of bar code scans compiled. No projecting or extroplating done. Just data gathering.

IF ( and I am not saying there is) there was a systemic bias ( no small niche retailers like VE, overemphasis on BB etc) a smaller delta between the formats makes the effects of these possible systemic issues magnified.

If there is problem - then ofcourse there is a problem ;)

I'm basically looking for collaborative evidence since I belong to what we call "reality based" community. Right now there is good collaborative evidence that Nielsen data is a good representative.

We all know Amazon rankings favored HD DVD heabily in '06. So does Nielsen data. In '07 Amazon ratings favor BD. So does Nielsen. In that last week in Amazon we see better rankings for HD with overall gaps norrowing. Let us see if Nielsen figures show that too.

This is like global warming. Is there 100% confidence - no. But is there mounting evidence - yes.

Since HD DVD and BD had such different release schedules their synchronized movement seems unlikely - to me at least. But it's possible for sure.

That is a good question to talk about.

Indeed. I noticed a similar trend with the Amazon numbers though - most of the time the two formats rise & fall in unison, but at different rates.

With the Amazon numbers, you could explain that by assuming there was a big new DVD release that pushed the relative rankings of both formats up/down, but that doesn't make as much sense with the Neilson numbers.

My guess is that studios tend to pick certain weeks as "good" weeks for big releases and multiple studios release stuff in those weeks, and other weeks are "bad" weeks (for god knows what reasons, I'm sure they take a lot into account when picking dates) and release fewer good titles that week.

I've definitely seen a pattern where certain weeks that have big events (holidays, academy awards, sports championships, TV season finales) have fewer DVD offerings.

I think this chart should have lower starting values for 2006. Low enough to yield 124k for HD-DVD and 280k for BD, YTD through 3-18.

The spreadsheet is available in this thread for downloading - thats how I got it too. Nobody is stopping you from coming up with your numbers/analysis !

With the Amazon numbers, you could explain that by assuming there was a big new DVD release that pushed the relative rankings of both formats up/down, but that doesn't make as much sense with the Neilson numbers.

To me, Neilson numbers should kind of tie in with Amazon. That would make sense.

The other part is that the ratios we get are inaccurate. They are based on first alert - and we don't get updated ratios from HMM (unless Grubert can somehow arrange for those ;) ;) ). That could alter the actual weekly sales figure we get. I'd look at 2 or 3 week moving average as a better sales indicator.

Anyway, coming back to the original question - someone could analyze the releases and comeup with reasons for the fall and raise and see if that makes sense. We know Warner releases movies in both formats at once and that can make the numbers go up and down in unison. Just need to see what happens in other weeks - after all HD DVD hasn't had new releases on some weeks.

HMM, this threads source of Nielsen data, disagrees with your statement. They claim Nielsen only captures 40% of the market, reports on that 40%, and then outside parties must perform and validate their own projections.

I've always been told that Nielsen only compiles barcode scans and reports the actual number of bar code scans compiled. No projecting or extroplating done. Just data gathering.

I have seen no email from HMM stating as such. Link?

Using ice's initial numbers. Looks like BR continues to benefit from CR. Next week will be interesting considering Amazon rankings.

http://img402.imageshack.us/img402/8662/nielsenanalysiscw6.gif (http://imageshack.us)
How about the idea for someone to attach this image to an email for someone at nielsen/videoscan/HMM and see if anyone will comment.

Even the numbers for both are unfortunately low, the idea of having a good model that accurately tracks unit sales is still a good thing.

I have seen no email from HMM stating as such. Link?

Read the thread. It's only been talked about for the past week...

Also, what is the purpose of a 64:100 market share on the starting number. This seems way out of wack with the very next SI market share. Is this just another fudge factor to make the results turn out a certain way? I never noticed it before. Did it really jump from 64 to 82 in two weeks, while during those same two weeks it was discovered that not all HD-DVD sales were being counted. The best jump for SI was about 5 points in a week, but this would indicate 9 points per week, which seems off.

From January 1 to January 21 are three weeks, not two...

Well, it does pass the litmus test of no negative volumes.

I propose the table reported on in this thread should look like the attached.

Read the thread. It's only been talked about for the past week...

The only thing I've read was:

Since Nielsen only represents about 40% of the business the numbers have been bumped up accordingly and then cross checked with studios to ensure accuracy.

'and' links two separate thoughts, so let's remove the unnecessary cross checking statement.

Since Nielsen only represents about 40% of the business the numbers have been bumped up accordingly

So, numbers have been "bumped" up because Nielson only represents 40% of the business. You don't see how that is saying the numbers were adjusted to compensate to represent 100% of the market - and are not directly reported numbers?

The only thing I've read was:



'and' links two separate thoughts, so let's remove the unnecessary cross checking statement.



So, numbers have been "bumped" up because Nielson only represents 40% of the business. You don't see how that is saying the numbers were adjusted to compensate to represent 100% of the market - and are not directly reported numbers?

This "bump" was done by HMM and not Nielsen, which is exactly what I said. Nielsen reported the number with no "bump". HMM took the number and "bumped" it because the Nielsen number only represented 40% of the market. HMM then confirmed their "bumped" number with studio sources.

Exactly what you just quoted confirms that Nielsen does not "bump" what they report (or it confirms HMM made a HUGE error).

This "bump" was done by HMM and not Nielsen, which is exactly what I said. Nielsen reported the number with no "bump". HMM took the number and "bumped" it because the Nielsen number only represented 40% of the market. HMM then confirmed their "bumped" number with studio sources.

Exactly what you just quoted confirms that Nielsen does not "bump" what they report (or it confirms HMM made a HUGE error).

Where does it say HMM made the adjustment? Does it say "we bumped"? No. It simply says the numbers were bumped. Also, do you think Home Media Magazine personally cross-checked the data with the studios? Or that Nielson did?

You assumed that HMM made the adjustment, but nothing in that statement says as such.

WRT top 5

BD: Casino Royale continues to be #1, outselling the two new entries (Rocky Balboa at #2 and Eragon at #3). The Holiday and Layer Cake exit the Top 5.

HD: Not a lot of change. Clerks II leaves the table, replaced by Goodfellas.

...

Back to this week. Very impressive numbers again from Blu-ray. My feeling is things will plateuax very soon. I can't see them going above 72% (2.5:1) because the weekly numbers don't seem to and if BD isn't getting 72+ in a week, they can't change the YTD.

...

Not totally sure what you are trying to point out here, but if you look at the weekly sales columns (far right of table) they list as a % of weekly sales. BD has been 81+% already for the last few weeks...

Using ice's initial numbers. Looks like BR continues to benefit from CR. Next week will be interesting considering Amazon rankings.

http://img402.imageshack.us/img402/8662/nielsenanalysiscw6.gif (http://imageshack.us)

Remember they sold through 50,000 copies of CR very quickly and sent 30,000 more. So it could account for nearly half of the BD sales for last 2 weeks.

Where does it say HMM made the adjustment? Does it say "we bumped"? No. It simply says the numbers were bumped. Also, do you think Home Media Magazine personally cross-checked the data with the studios? Or that Nielson did?

You assumed that HMM made the adjustment, but nothing in that statement says as such.

The author of the article, who works for HMM stated that he checked with the studios to confirm the "bump".

The author of the article, who works for HMM stated that he checked with the studios to confirm the "bump".

Again, I didn't see anywhere that posted in this thread. If you could simply point it out, I would really appreciate it.

What we have been doing all along is to use the latest "official" figures and use the "official" ratios to get meaningful numbers. The current numbers take projected HMM numbers. But if the "official " 2006 YTD sales data is based on a total market projection, and the 'official" weekly ratios YTD and SI are based on Nielsons 40% assumed capture of the market rate, aren't we possibly multiplying two error rates or margins of error together?

My error analysis thinking is rusty, but what happens if you multiply say a 7% margin of error of one sample with a 7% margin of error with another data point?

I mean if we are basing the weekly sale stats off something that was a projection to begin with, can we apply a 40% sample rate of that every week and get consistent results? Or should we stick to a calculation that is based on what we think the 2006 EOY Nielson capture figure was? Or does it matter? My head hurts...... ;)

What is the latest "official numbers we are using for the yellow blocks in the excel chart(2006 EOY)?

No, that isn't the case. You guys were just lucky that your projected numbers somewhat correponded to what Nielsen said. If your method truely works, you should be able to enter a lower 2006 starting value, to yield disc sales with YTD volumes of 124k and 280k through 3/18. No negative volumes should be generated by your method.

This would be the litmus test of your conversion of market share ratio to disc volume.

That is why asked sometime back what your math background is. If you have some math background you shouldn't have to ask this question - we get two ratios every week (so two new equations to solve) and two unknowns. And I solve those equations. Simple. This is not rocket science - just simple highschool level algebra. Does this make sense to you ?

Ice does it totally differently and comes up with the same numbers.

Now, what exactly is your question about our "methodology" and what methodology are you referring to anyway - mine or Ice's ?

But if the "official " 2006 YTD sales data is based on a total market projection, and the 'official" weekly ratios YTD and SI are based on Nielsons 40% assumed capture of the market rate, aren't we possibly multiplying two error rates or margins of error together?

The assumption I'm making is that when HMM projected, they stuck with the ratios that Nielsen reported. They better do.

Or in other words - they multiplied all the numbers with 2.5 (100/40).

Agreed, and I do think I am being too pessimistic, but if a $399 price point was going to convince people to buy into HD DVD in noticeably accelerated numbers, the disc sales are not showing it (this might, however, be a function of the X number of free discs offered with each player purchase).

Icemage, unless something has changed, to get the "free" HD DVDs one has to mail in the UPC code from the box + receipt.... then you get the titles. I have yet to do so myself but according to the thread in the HD DVD software forum this can take up to 2 mo.

New buyers of A2 and XA2 are buying new titles outright, my guesstimate somewhere in the 5 to 10 withing weeks of their player purchase.

Definitely not showing up in the data.....

Nielson data wouldn't be useful if they did not project. Does anyone truly believe that when they say 6M people watched the Superbowl, that they had boxes in 6M+ homes? Of course not. They project all their data. Its what they (and any other survey company) do. We are are not talking broadcast media figures here.

There it is clearly stated for Nielson broadcast numbers that they are a projection based on a market sample and they are trying to show what is in that market. I use them all the time. You can get the sample size information and the error margin.

Retail stats are a different kettle of fish than broadcast TV. There you have different classes of retailers not just people demographics and classifications.

Definitely not showing up in the data.....Do you mean the purchased discs are not being tracked, or the assumed rise in sales due to lower price is not happening?

But if the "official " 2006 YTD sales data is based on a total market projection, and the 'official" weekly ratios YTD and SI are based on Nielsons 40% assumed capture of the market rate, aren't we possibly multiplying two error rates or margins of error together?

My error analysis thinking is rusty, but what happens if you multiply say a 7% margin of error of one sample with a 7% margin of error with another data point?
Not exactly.

Realistically, we have two points where error can occur. Either our assumptions for starting values are wrong (64% BD to 100% HD ratio), or Nielsen's ratios are wrong.

Aside from that, the actual magnitude of the numbers does not impact anything except the finalized totals; the overall ratios remain the same as long as you do the multiplication correctly.

I mean if we are basing the weekly sale stats off something that was a projection to begin with, can we apply a 40% sample rate of that every week and get consistent results? Or should we stick to a calculation that is based on what we think the 2006 EOY Nielson capture figure was? Or does it matter? My head hurts...... ;)
Actually, that's exactly what we're doing.

Each week of data is analyzed independently of the others.

What my spreadsheet does is look at the assumed start of 2007 SI totals, and then finds the most accurate total of year-to-date sales in the proportions indicated that will satisfy the reported weekly SI ratios.

The only way they are dependent on one another is when we are projecting weekly sales, where we subtract the total from the current week from the previous week to arrive at weekly totals.

What is the latest "official numbers we are using for the yellow blocks in the excel chart(2006 EOY)?

The yellow squares are the only wild card assumed values we use in the entire process. Our assumption is that Blu-ray and HD DVD were progressing at a roughly same vector for SI in the weeks between 1/1/07 and 1/21/07, when we received our first "clean" data point. I explained it a while back, but it's basically a projection from the ~5+% rising SI differential per week from 1/21 to 1/28, walked backwards 3 weeks for a net total of ~18%.

Note that tweaking the ratio between the two numbers in the yellow squares impacts the weekly percentages somewhat, but altering them radically causes all kinds of problems; our current 64% assumption brings our weekly %s almost 100% in line with what PC Magazine has been reporting, and deviating too heavily from that makes no sense since the numbers still have to be plausible for the first three weeks of January (in other words, the SI difference can't realistically be more than maybe +/-10% from the 64% figure of the total of sales between 1/1 and 1/21 stop making any sense for those 3 weeks)

Icemage, unless something has changed, to get the "free" HD DVDs one has to mail in the UPC code from the box + receipt.... then you get the titles. I have yet to do so myself but according to the thread in the HD DVD software forum this can take up to 2 mo.

New buyers of A2 and XA2 are buying new titles outright, my guesstimate somewhere in the 5 to 10 withing weeks of their player purchase.

Definitely not showing up in the data.....

Agreed. My point though, was that because new HD DVD owners are receiving titles for free, they're not as willing to run out and just buy everything in sight, which depresses the actual sales numbers down since those discs aren't being reported as sales (they're really not). Nonetheless the free discs affect sales in the sense that those owners will never purchase those titles.

I think this chart should have lower starting values for 2006. Low enough to yield 124k for HD-DVD and 280k for BD, YTD through 3-18. This will allow it to match HMM's statements on Nielsen numbers. If the method works and doesn't yield negative volumes, then it is probably a sound method to use.

Then people can use whatever market coverage % they want (with about 40% being indicated by HMM) on the projected volumes. So the estimated volumes will be projected volumes based on Nielsen monitored volumes.

Basically, a column for estimated volumes (ie Nielsen, 124k and 280 k on 3/18), and then a 2nd "projected" column can be added that uses the 40% market coverage to project the total market volume..... This is a more articulate way of saying what I was trying to get at. It seems to me to place projected data in the EOY yellow box start figures is asking for issues. Or if we leave it as is, the column headings are inaccurate because the weekly numbers are projections off a projected or estimated EOY number that already had some margin of error and assumptions inherent in it.

To me, Neilson numbers should kind of tie in with Amazon. That would make sense.

The other part is that the ratios we get are inaccurate. They are based on first alert - and we don't get updated ratios from HMM (unless Grubert can somehow arrange for those ;) ;) ). That could alter the actual weekly sales figure we get. I'd look at 2 or 3 week moving average as a better sales indicator.

Anyway, coming back to the original question - someone could analyze the releases and comeup with reasons for the fall and raise and see if that makes sense. We know Warner releases movies in both formats at once and that can make the numbers go up and down in unison. Just need to see what happens in other weeks - after all HD DVD hasn't had new releases on some weeks.

.....ahhh would not the first alert data probably have a bias with larger retailer outlets? That would probably lead to a bias for more blockbuster mainstream releases and for the newer released titles not catalog or the older released titles.

Since Blu-ray leads in both of those categories, as Blu-ray studio releases have been in the first quarter both recent and plentiful, wouldn't the first alert data typically favor Blu-ray more?

Maybe some Blu-ray PR flak, whispered to HMM that maybe they should start using first alert data when the SI equality date was getting close, and they snookered HMM to go along as they did not know the consequences?

That sounds like X Files territory, but it may have been a subtle point HMM missed.

But if that was true, doesn't the SI and YTD numbers have a self correcting mechanism in them as the are based on final numbers. The problem in that instance if there is a big difference in First alert and final numbers is that the weekly stats would show a systemic HD DVD undercount and Blu-ray overcount and then the SI and YTD figures would not match the total of the weekly calculations.

We know there is some error in that we are using the first alert numbers now, the issue is how much.

That's reaching.

Figures for each week also include the data for the previous that didn't make the First Alert. ie if sales at a given retailer don't get into First Alert for that week, they go into the following one.

That's reaching.

Figures for each week also include the data for the previous that didn't make the First Alert. ie if sales at a given retailer don't get into First Alert for that week, they go into the following one. You sure? I think the first alert weekly figures are independent. The weekly FINAL numbers are adjusted as well as the YTD and SI figures, but each first alert number is never updated. And we are now using first alert data in the chart.

First alert providers are only a small subset of the data base, your sources said sometimes the first alert numbers have been noticably off.

Correct me if I am wrong, but we are not ever going back and getting corrected weekly numbers once HMM publishes those based on the first alert data.




Just let me say right now though, I think these will all wash out as sales volumes increase so I think any issues will be smaller in significance as time goes on.

And I also can't really think of any better way to do it than what Grubert Icemage Natarj and others have done here. I'm not bitching, I'm just trying to see the total picture. :D

You sure? I think the first alert weekly figures are independent. The weekly FINAL numbers are adjusted as well as the YTD and SI figures, but each first alert number is never updated. And we are now using first alert data in the chart.
The First Alert data we are seeing is cumulative YTD and SI.

At most the error would be in the last week of data, which, now that we're now 3 full months into the year, would be of minimal impact.

Home Media Magazine has confirmed that they have seen the cumulative numbers adjusted. There would be no reason to have First Alert data that doesn't make use of the best available cumulative totals.

First alert providers are only a small subset of the data base, your sources said sometimes the first alert numbers have been noticably off.


On the contrary. They have said that "to date the adjustments have not been measurable".

But if that was true, doesn't the SI and YTD numbers have a self correcting mechanism in them as the are based on final numbers. The problem in that instance if there is a big difference in First alert and final numbers is that the weekly stats would show a systemic HD DVD undercount and Blu-ray overcount and then the SI and YTD figures would not match the total of the weekly calculations.

To expand on what Grubert said, this part is only useful in explaining vagaries for a particular week (like we got 3 weeks back) - not week after week. The YTD and SI numbers include First Alert only for the last week.

Also we were told that the difference between First Alert and Complete were not significant by HMM - which to me means no more than 3-5% difference in YTD and SI ratios.

Well, it does pass the litmus test of no negative volumes.

I propose the table reported on in this thread should look like the attached.

Problem with your numbers is that for the first 3 weeks of Jan your figures yield a total which is similar to the weekly figures after that. 37K vs 29K for BD and 19K vs 14K for HD. How realistic is that ?

Look at Ice's explanation for initial numbers above.

I did a lot of simulations to come up with the initial SI ratio between BD & HD DVD (to 64% - that looked better than 65% or 63% - let alone others). As we adjusted our initial numbers to match various figures we got over the last few weeks, I tried several times to change that number. But each time 64% looked better than others ...

http://www.avsforum.com/avs-vb/showthread.php?p=9843212&&#post9843212

I did a lot of simulations to come up with the initial SI ratio between BD & HD DVD (to 64% - that looked better than 65% or 63% - let alone others). As we adjusted our initial numbers to match various figures we got over the last few weeks, I tried several times to change that number. But each time 64% looked better than others ...

http://www.avsforum.com/avs-vb/showthread.php?p=9843212&&#post9843212
I agree. I've done the same scenario testing and come to the same conclusion. For the numbers we see posted from PC Magazine, 64% produces a better weekly ratio than any other number you can pick without going into fractional percentages.

I think this chart should have lower starting values for 2006. Low enough to yield 124k for HD-DVD and 280k for BD, YTD through 3-18.

You can get this by changing HD initial number to 240K. You get 279,754 for BD and 124,322 for HD DVD YTD.

Here are the charts for HMM and (derived from there) Videoscan numbers.

http://img141.imageshack.us/img141/8801/nielsenhidefuf8.gif (http://imageshack.us)

BD has been 81+% already for the last few weeks...

True, but they are projected figures. They've forced the YTD up but if the weekly figure drops to under 70% then the YTD will stay where it now.

I wonder, does videoscan get data from Warner Home Video ?

http://www.avsforum.com/avs-vb/showthread.php?t=827214

You must enter each of these coupon codes sparately: WBNEWS, USD AND USFS at the WHV DVD Shop to get 30% off your entire purchase WITH FREE SHIPPING.

I wonder, does videoscan get data from Warner Home Video ?

http://www.avsforum.com/avs-vb/showthread.php?t=827214
I don't know if these etailers move that many disks. I used to think that the volumes should be huge, but following casino royale at amazon recently, it appears that when it got around rank#007 (and above) after street date, it only needed <400 disks / day to get that ranking.

This means that that sale in amazon that moved so many titles into the top 50 basically did not result in as many sales/title as one would expect.

Meaning that for some strange reason, there are still a lot of people walking in B&Ms and paying more for the same movies. Maybe it's the impulse buy thing with all luxury goods.

I don't know if these etailers move that many disks. I used to think that the volumes should be huge, but following casino royale at amazon recently, it appears that when it got around rank#007 (and above) after street date, it only needed <400 disks / day to get that ranking.

Interesting...but how did you know it was only 400 discs per day sold? In order to derive that number you need to know rate of input and rate of output. the inventory numbers by themselves mean nothing.

Interesting...but how did you know it was only 400 discs per day sold? In order to derive that number you need to know rate of input and rate of output. the inventory numbers by themselves mean nothing.
You're right, I did not factor in new stock, but I looked at their stock in hand every day at roughly the same time for a couple of days when the sales rank stayed there, and I recalled that they sold less than 400 each of those two days. Not scientific I know, but I didn't really follow eproduct #s anymore ever since nielsen videoscan numbers.

If it wasn't for the psychological milestones of those silly rankings during the sale and that CR crossed #007 in rank, I wouldn't even try to follow how many amazon sells.

Anyway, my point is that, it seems the etailers don't move as many disks as we think. Perhaps eproduct.com could try to track actual # sold for each title on a weekly/daily basis since they sample every 15 minutes and have a better chance of capturing the stock increase and factor those in.

Problem with your numbers is that for the first 3 weeks of Jan your figures yield a total which is similar to the weekly figures after that. 37K vs 29K for BD and 19K vs 14K for HD. How realistic is that ?

Look at Ice's explanation for initial numbers above.

I did a lot of simulations to come up with the initial SI ratio between BD & HD DVD (to 64% - that looked better than 65% or 63% - let alone others). As we adjusted our initial numbers to match various figures we got over the last few weeks, I tried several times to change that number. But each time 64% looked better than others ...

http://www.avsforum.com/avs-vb/showthread.php?p=9843212&&#post9843212

Yes, but your method totally ignores, that in a roundabout way, HMM has told us the actual YTD numbers should be 124k and 280k through March 18th. To me, saying the number is 2.25X as larger, is a much more glaring error than the missing two weeks of data.

There are multiple starting points that can get you in the ballpark of the 124k and 280k numbers though, so maybe there are two better starting numbers.

I personally think a 100:64 ratio for jan 1 is a bit low. If you look at that change in SI all through Jan and Feb, it's around 5 points each week. So it would project back to be about 77:100 for Jan 14, 72:100 for Jan 7, and 67:100 for Dec 31. I'll play around to see how the numbers look while maintaining a 67:100 ratio.

I personally think a 100:64 ratio for jan 1 is a bit low. If you look at that change in SI all through Jan and Feb, it's around 5 points each week. So it would project back to be about 77:100 for Jan 14, 72:100 for Jan 7, and 67:100 for Dec 31. I'll play around to see how the numbers look while maintaining a 67:100 ratio.The changes from week to week in SI through March 4th were:

Jan 28: 5.46
Feb 4: 4.8
Feb 11: 4.52
Feb 18: 4.21
Feb 25: 4.46
Mar 4: 2.79

Why would you expect the SI to change by the same absolute amount each week, when the leader for runrate is different than the leader in SI? Absolute changes tend to slow down the closer SI is to the weekly numbers and be the largest when the SI and weekly numbers are the farthest apart. Of course it depends on the volumes and the gaps, but assuming the same absolute change in SI for each week doesn't make any sense to me. Especially using a number smaller than the 5.46 for Jan 28 for each of the 3 weeks. Going from 82.3 to 64.0 backwards could be in steps of 5.7, 6.1, and 6.5.

--Darin

You're right, I did not factor in new stock, but I looked at their stock in hand every day at roughly the same time for a couple of days when the sales rank stayed there, and I recalled that they sold less than 400 each of those two days. Not scientific I know, but I didn't really follow eproduct #s anymore ever since nielsen videoscan numbers.I looked at the eproductwars.com/dvd links for "Casino Royale" around this time where you can click on the sales chart and see the stock numbers and times. Sometimes the stock drops or rises a lot in a very short period and I didn't count those. I think your number was pretty close for the Tuesday it shipped and maybe the next day. My best guestimate from looking at the rankings leading up to ship day and then the graph for the week after is that Amazon may have had about 4k preorders and then sold about 2k more in the first week.

--Darin

If anyone was wondering why I've been so quiet over the last day since the release of the latest Nielsen data, part of it is because I've been awfully busy... and the other part is because I've been working on a little something special. :)

Here's a new spreadsheet.

New features:
- Reorganized columns that make a bit more sense
- Allowed for a second set of projections to be calculated if the purple block is filled in with a percentage value for an estimate of Nielsen/VideoScan's market coverage.
- Greatly optimized code which performs the calculations much faster

Observe this projection, derived from a 600K assumption for HD DVD SI on 1/1/07 and 40% Nielsen coverage (which I find matches Home Media Magazine's internal estimates pretty well):

http://endrop.com/album/photos/edh3amjmm121timrwmmt.jpg (http://endrop.com/viewer.php?file=edh3amjmm121timrwmmt.jpg)

If anyone was wondering why I've been so quiet over the last day since the release of the latest Nielsen data, part of it is because I've been awfully busy... and the other part is because I've been working on a little something special. :)

Here's a new spreadsheet.

New features:
- Reorganized columns that make a bit more sense
- Allowed for a second set of projections to be calculated if the purple block is filled in with a percentage value for an estimate of Nielsen/VideoScan's market coverage.
- Greatly optimized code which performs the calculations much faster

Observe this projection, derived from a 600K assumption for HD DVD SI on 1/1/07 and 40% Nielsen coverage (which I find matches Home Media Magazine's internal estimates pretty well):

http://endrop.com/album/photos/edh3amjmm121timrwmmt.jpg (http://endrop.com/viewer.php?file=edh3amjmm121timrwmmt.jpg)


Cool. Good work. :)

I'm a happy little Kosty now. :D

Ok, you math gurus.

Since we (1) now see the new weekly estimates from the first alert data, and we (2) have the final corrections for that last week in the new final YTD and SI numbers (that include the estimated new first alert numbers) is there any way at all possible to back up and tell (3) how much the previous weeks first alert numbers were different than the final completed numbers for the week?

Any way we can get the "final" weekly numbers instead of HMM's published "first alert" numbers to put in the spreadsheet?


That is besides having Grubert send flowers and chocolates to the HMM editor and staff?

....then again Grubert, thats a thought, since they are publishing first alert numbers why can't they footnote the final numbers if that ""to date the adjustments have not been measurable". ? Or at least e-mail them to you that you could share with your buddies here? ;)

If anyone was wondering why I've been so quiet over the last day since the release of the latest Nielsen data, part of it is because I've been awfully busy... and the other part is because I've been working on a little something special. :)

Here's a new spreadsheet.

New features:
- Reorganized columns that make a bit more sense
- Allowed for a second set of projections to be calculated if the purple block is filled in with a percentage value for an estimate of Nielsen/VideoScan's market coverage.
- Greatly optimized code which performs the calculations much faster

Observe this projection, derived from a 600K assumption for HD DVD SI on 1/1/07 and 40% Nielsen coverage (which I find matches Home Media Magazine's internal estimates pretty well):

http://endrop.com/album/photos/edh3amjmm121timrwmmt.jpg (http://endrop.com/viewer.php?file=edh3amjmm121timrwmmt.jpg) If we know the first alert data is only a subset of the final weeks data (at least from a quantity standpoint) and the final weekly database numbers for that week will be higher, are we not undercounting the total volume for that week, and hence for the entire YTD and SI number's sold figures by using that ""first alert" data for the weekly sold number?

I do like the new chart though. It clearly addresses my concerns over how the relatively small Nielson sample can affect the ratios. I love this stuff, and you guys have done a bang up job. :cool:

I wish we could get the final weekly numbers. I would have gladly waited an extra week for them. ;)

Since we (1) now see the new weekly estimates from the first alert data, and we (2) have the final corrections for that last week in the new final YTD and SI numbers (that include the estimated new first alert numbers) is there any way at all possible to back up and tell (3) how much the previous weeks first alert numbers were different than the final completed numbers for the week?

Any way we can get the "final" weekly numbers instead of HMM's published "first alert" numbers to put in the spreadsheet?

That is besides having Grubert send flowers and chocolates to the HMM editor and staff?
Unfortunately not. If this were mathematically possible I'd already be doing it. :) Because we don't know the methodology of how the First Alert data is constructed, we have no way of knowing just how much the current week's data is influenced (and thus no way of knowing what last week's data was).

....then again Grubert, thats a thought, since they are publishing first alert numbers why can't they footnote the final numbers if that ""to date the adjustments have not been measurable". ? Or at least e-mail them to you that you could share with your buddies here? ;)
That would be helpful. What they consider "not measurable" may in fact affect our calculations quite a lot (as we can see from the weird numbers on 3/4 and 3/11).

I'm a renter as I have said many times in the past.


Nataraj: I know, that is why I also asked the second set. Buying is more important to a studio, they don't make the same $ from renting. Also I was curious if you bought any movies.

About 50% of my rentals are English. Rest are Indian. Of the English ones may be 60% are from the major studios - and now-a-days almost 100% of these are HD DVD.

edit : I didn't notice the "HD" in your question ... however these figures should give you an idea why I want the authoring and other costs to be low so that I can get the 70% of what I watch in HD in future.

when I said large studios I did not mean Hollywood. All the majore studios (international or US) that have released on HD DVD have used AACS. I am also relatively certain that if Bollywood studios will release on HD DVD they will also use AACS. That is why I was asking about the movies you saw on HD DVD, it is hard to know what studiops that have not released will do. The thing is except for possibly some extremely niche titles I don't think that studios will skip AACS on HD DVD

I am also relatively certain that if Bollywood studios will release on HD DVD they will also use AACS.

Not sure about that. I have seen many DVDs without CSS and if it costs quite a bit for AACS, they will skip it.

Sorry, wrong. The arguement about replication costs stem from porn producers not wanting to pay more to replicate in BD.


Yous ure like to posrt garbedge. Can you link to one porn producer that stated they wanted BD, they found a replicator, got a qupote and the AACS mastering fee that was forced on them that made them change their mind?

My best guestimate from looking at the rankings leading up to ship day and then the graph for the week after is that Amazon may have had about 4k preorders and then sold about 2k more in the first week.

That looks like 10 to 15% of all sales of BD CR. Sounds reasonable.

My guess is Amazon is selling more BL DVDs than B&M (even BB or Frys). That is the reason we will see them in top 50 at amazon and probably not overall (if someone had a list like that).

On that 40% - I still don't beleive it. I think videoscan captures a lot more than that.

Ofcourse we know HMM contacted studios for the numbers after they multiplied Videoscan numbers by 2.5. And the studios confirmed them.

Let us think about that for a second. HMM gets say 10K for a title from videoscan. They ask the studios - did you sell 25K of this title ? Even if the studios think the actual number is 15K - are they going to say that ? I'm willing to bet they will agree with the higher number.

Esp. if even majors like Amazon is only doing 15% of total sales - how big is Deepdiscount going to be ?

Im I the only one that think its more interesting to judge/compare sales on movies released on both format the same time?

Of course will a movie like CR sell alot of movies, and with numbers this smal, a single blockbuster title can really push the numbers. That dont really say som much about the potential of either format. CR would also sell very well on HD DVD if it were released today.

Im I the only one that think its more interesting to judge/compare sales on movies released on both format the same time?

Of course will a movie like CR sell alot of movies, and with numbers this smal, a single blockbuster title can really push the numbers. That dont really say som much about the potential of either format. CR would also sell very well on HD DVD if it were released today.

WB sales their movie sales are about equal on the two formats.

WB sales their movie sales are about equal on the two formats.

WB is very careful to stay "neutral" so i would take anything they say with a BIG grain of salt.

the reality is, The Departed Blu-ray outsold the HD-DVD version by 20,000+ copies versus 13,000 copies on the first week (based on third party numbers), and based on the very different amazon.com rankings has probably outsold it going forward every week since then by a ton as well.

the reality is, The Departed Blu-ray outsold the HD-DVD version by 20,000+ copies versus 13,000 copies on the first week (based on third party numbers), and based on the very different amazon.com rankings has probably outsold it going forward every week since then by a ton as well.

And HappyFeet has generally outsold on HD. The WB comment was from a few weeks back and could have been true at that time.

And HappyFeet has generally outsold on HD. The WB comment was from a few weeks back and could have been true at that time.

Actually, we don't know the actual sales numbers for Happy feet....

The departed (HD-DVD) outranked the Departed (Blu-ray) on amazon.com for part of the first week as well, and we all know the actual sales did not reflect that RANKING.

WB sales their movie sales are about equal on the two formats.
Seemed to be carefully worded. Just look at the sales charts you do and you can see that as of February 18th it would be accurate for somebody to say that sales were about equal between the two formats, carefully ignoring the runrates and YTD factor.

--Darin

I remember there was a week when BD had several titles in the top 50. I was wondering how that showed up on Neilson? Did we see a sharp spike on overall sales or not? To me it looked like BD sales really took off for a week or two, I'm looking at the charts and it looks like Neilson shows a sharp drop! Is that true? A massive spike on Amazon but a massive sales drop on Neilson?

I remember there was a week when BD had several titles in the top 50. I was wondering how that showed up on Neilson? Did we see a sharp spike on overall sales or not? To me it looked like BD sales really took off for a week or two, I'm looking at the charts and it looks like Neilson shows a sharp drop! Is that true? A massive spike on Amazon but a massive sales drop on Neilson?There was a week with no new releases, but a sale on Amazon. I think some people estimated maybe 10k sold there, but I'm not sure how long a period that was. In any case, it would seem a little strange if Nielsen only counted less than 13k total that week and Amazon is included.

--Darin

Nataraj: don't know what you would consider much. But RBfilms gave us the fees
http://www.avsforum.com/avs-vb/showthread.php?p=10058863#post10058863
3k-10k set-up fee for a studio (looks like either a one time charge or an annual fee) and then 1700 for each AACS key (at least once a movie, but possibly more)

if you are thinking 1000 movies then not cheap (in his example it added up to 10k) but if you are thinking 100k (I am sure Indian movies can do that in numbers even relatively early on in HD) it is not much

Actually, we don't know the actual sales numbers for Happy feet....

The departed (HD-DVD) outranked the Departed (Blu-ray) on amazon.com for part of the first week as well, and we all know the actual sales did not reflect that RANKING.

it is possible that more HD DVD owners ordering movies online, and Blu-ray owners going shoping outside of online shopping...., if Happy Feat HD DVD version will at the end outsold by Blu-ray version first week same as Departed, we will know more :)

Marek

On that 40% - I still don't beleive it. I think videoscan captures a lot more than that.

Of course we know HMM contacted studios for the numbers after they multiplied Videoscan numbers by 2.5. And the studios confirmed them.

Let us think about that for a second. HMM gets say 10K for a title from videoscan. They ask the studios - did you sell 25K of this title ? Even if the studios think the actual number is 15K - are they going to say that ? I'm willing to bet they will agree with the higher number.

Esp. if even majors like Amazon is only doing 15% of total sales - how big is Deepdiscount going to be ?

I don't know if it is 40% or 60% or 80%, not even sure if it matters. The question should be is VS representative. If it is then it is a bit less important to know if it is 300k 255k or 350k

as for studios, I don't think they would lie, but on the other hand a sold to a studio is not the same as a sold to us.

PS if I had the contacts and the info, what I would do is on the VS (or any data publisher) side is ask POS and shipped from studio for companies they get data and then ask the studios how much they shipped.

i.e. lets say a title from Warner sold b and h in POS and B and H were shipped to those stores if Warner shipped 2B and 2Y then it is extremely safe to assume the b:h ratio is good. If on the other hand they shipped 28,H or B,2H then we know the ratio should be off.

WB sales their movie sales are about equal on the two formats.

Nataraj they also released more movies on HD DVD. For example BB is one of the top HD DVDs.

I remember there was a week when BD had several titles in the top 50. I was wondering how that showed up on Neilson? Did we see a sharp spike on overall sales or not? To me it looked like BD sales really took off for a week or two, I'm looking at the charts and it looks like Neilson shows a sharp drop! Is that true? A massive spike on Amazon but a massive sales drop on Neilson?

no spike, actually the opposite (for both formats, it is the dip in the graph). Part of the issue is that ranking is order base while the VS numbers are POS. Since many of them ended up out of stock they count as rank but not POS

Nataraj they also released more movies on HD DVD. For example BB is one of the top HD DVDs.

It sounded like they were talking about movies released on both formats - they were comparing individual titles.

Unfortunately not. If this were mathematically possible I'd already be doing it. :) Because we don't know the methodology of how the First Alert data is constructed, we have no way of knowing just how much the current week's data is influenced (and thus no way of knowing what last week's data was).


That would be helpful. What they consider "not measurable" may in fact affect our calculations quite a lot (as we can see from the weird numbers on 3/4 and 3/11). The first alert data is the data they collect from the electronically collected POS system data from fully automated retailers, typically the larger chain retailers. It is 100% sales data from scanners by SKU. It is a census not a sample and is categorically somewhat different than the sample based data that comes from the other tracked retailers and is incorporated into the final data.

On the first alert methodology:

I have received the following information recently from Nielson and I have it from some other documents I had from a trade show form several years ago. It is identical to the information that was included on kdragon's alledged emails from Nielson/Videoscan. Its one of the reasons I think his documents were legitimate.

http://www.avsforum.com/avs-vb/attachment.php?attachmentid=75417



Nielsen VideoScan’s web site is available to subscribers only and contains weekly unit sales for over 60,000 VHS and DVD items. Our data is collected in two stages. The first wave of data is made available three days after the close of the business week, and the second wave of data is available ten days after the close of the business week.

FIRST ALERT

We call the system that delivers the first wave of data the First Alert system. First Alert is designed to provide our clients with a preview of things to come. It can be equated to “early box office.” It should be noted that chart positions and unit volume can change significantly between First Alert and the Complete data available the following week.

First Alert – The First Alert system provides almost immediate information, compiled
from Nielsen VideoScan’s early reporting retailers. First Alert data is available three
days after the close of the reporting week. However, First Alert does not contain
sales data from Wal*Mart, Grocery and Drug.


I think this is basically accurate but not current
http://www.avsforum.com/avs-vb/attachment.php?attachmentid=75416

BTW , the same document also mentions that the data is a "comprehensive view" of the VHS and DVD sell-through business. It never implies the data is scaled up to provide a complete market picture.

I beleive it is what it is and is provided as is to client companies for them to do their own projections of for total market sales.

VideoScan and ACNielsen (a partnership since January 2000, and sister companies under the VNU banner since 2001) seamlessly integrate point-of-sale (POS) data collected by both companies to provide a comprehensive view of the VHS and DVD sell-through business.

no spike, actually the opposite (for both formats, it is the dip in the graph). Part of the issue is that ranking is order base while the VS numbers are POS. Since many of them ended up out of stock they count as rank but not POS

There were also people like me who ordered in tiers. For instance, I just received three titles from Amazon just this week that were part of the 50% off sale, and I have five more of the sale titles coming to me later this month (April).

~Alan

Originally Posted by MovieSwede
Im I the only one that think its more interesting to judge/compare sales on movies released on both format the same time?

Of course will a movie like CR sell alot of movies, and with numbers this smal, a single blockbuster title can really push the numbers. That dont really say som much about the potential of either format. CR would also sell very well on HD DVD if it were released today. WB sales their movie sales are about equal on the two formats. It sounded like they were talking about movies released on both formats - they were comparing individual titles
Look at the Happy Feet thread and the Amazon tracking sites.

Happy Feet is one of the few titles released the same tme on both formats.

about even

http://img252.imageshack.us/img252/5457/hfsatvr7.png

discussion here:
http://www.avsforum.com/avs-vb/showthread.php?t=825255

eproductwars.com (thedvdwars.com ranking)

rank as of this hour
Blu-ray : 69
HD DVD: 62


hdgamedb.com

rank as of this hour
Blu-ray : 58
HD DVD: 71


hour to hour volatility means little, the AMazon sites are showing different current rankings, because of when they last updated their stats. ;)


daily stats start to show trends,
http://img252.imageshack.us/img252/5457/hfsatvr7.png

Kinda neck and neck

Im I the only one that think its more interesting to judge/compare sales on movies released on both format the same time?

Of course will a movie like CR sell alot of movies, and with numbers this smal, a single blockbuster title can really push the numbers. That dont really say som much about the potential of either format. CR would also sell very well on HD DVD if it were released today.

Of course the potential of HD-DVD would be higher if they had Fox/Disney/Sony movies. But they don't. Content is a very real advantage of Blu-ray, so why exclude it in comparisons?

If anything, movies released on both should be selling better on HD-DVD precisely because there is less choice for HD-DVD owners on what to spend their Hi-Def software budget.

BTW , the same document also mentions that the data is a "comprehensive view" of the VHS and DVD sell-through business. It never implies the data is scaled up to provide a complete market picture.

I beleive it is what it is and is provided as is to client companies for them to do their own projections of for total market sales.

Exactly, a studio don't care much EXACTLY how big the total volume for a format is anyway. At least not for weekly numbers. They can use Nielsen to check how well their titles are selling.

They can check the sales figures for each of their own titles and compare that to how many copies they have shipped to the SAME retailers. Of course they would get this eventually anyway but it would take much longer waiting for returns from retailers etc. They can also extrapolate based on what they've shipped or the normal percentage difference to retailers not in the Nielsen numbers and get their own total.

They can also compare their own titles to specific comparable titles from other studios. And of course get total market share in percentage compared to other studios and total market.

Nielson has been about 4 to 1 for 3 weeks now. Any thoughts on this guys? Traditionally 3:1 or 4:1 is a point where retailers start stocking one over the other.

If anything, movies released on both should be selling better on HD-DVD precisely because there is less choice for HD-DVD owners on what to spend their Hi-Def software budget.

That is only true if you accept there are equal # of people who want to buy HD DVD and BD movies. Do you ? :p

That is only true if you accept there are equal # of people who want to buy HD DVD and BD movies. Do you ? :p

No. If, say, there are twice as many people willing to buy BDs, but they have three times the movies to choose from, individual HD-DVD titles should sell better than their BD counterparts (if all buyers have the same software budget).

Nielson has been about 4 to 1 for 3 weeks now. Any thoughts on this guys? Traditionally 3:1 or 4:1 is a point where retailers start stocking one over the other. But 1st quarter numbers are traditionally low in sales volume, there is a major factor to account for in the numbers of titles released in both formats, professionals know the impact of the coupons and promotional offers and the installed base of HD DVD and Blu-ray players is increasing rapidly.

In short, its not a static situation and most people making decisions need more information and much much more time and sales volume to make any decisions.

Just the HD DVD surge back with a few releases may indicate that ratio is not fixed in stone. ;)

But 1st quarter numbers are traditionally low in sales volume, there is a major factor to account for in the numbers of titles released in both formats, professionals know the impact of the coupons and promotional offers and the installed base of HD DVD and Blu-ray players is increasing rapidly.

In short, its not a static situation and most people making decisions need more information and much much more time and sales volume to make any decisions.

Just the HD DVD surge back with a few releases may indicate that ratio is not fixed in stone. ;)

Titles released is true, but as we've seen by an apples to apples comparison of the same movies blu still sells more.

Is the HD-DVD player base expanding rapidly? As low as the add-on sales are which is the majority I don't know if I'd phrase it that way.

Both sides are using promotions pretty handily if not just giving movies away.

Surge? I surmise you are referring to Amazon (I mainly attribute this to the Matrix being available). But if you look at DVD empire for the week those numbers are still about 4 to 1 blu.

I don't think it's fixed in stone either. But volume is not necessarily a determining factor in a retailers decision as ratios are. If they see they can get 4 times as many sales using the same space it's a relatively simple decision.

No. If, say, there are twice as many people willing to buy BDs, but they have three times the movies to choose from, individual HD-DVD titles should sell better than their BD counterparts (if all buyers have the same software budget).

Only if all movies have same attractiveness ...

This is the reason why PS3 games in the US don't sell better than the 360 games.

Only if all movies have same attractiveness ...

This is the reason why PS3 games in the US don't sell better than the 360 games.

Wouldn't the 360 having sold >4x as many units in the US as the PS/3 have something to do with that?

Gary

Wouldn't the 360 having sold >4x as many units in the US as the PS/3 have something to do with that?

That was my first point.

Titles released is true, but as we've seen by an apples to apples comparison of the same movies blu still sells more.

Now, if HappyFeet sells more in HD than BD, what would you say ?

Now, if HappyFeet sells more in HD than PS3, what would you say ?

I would say, that's about right since the Blu-ray version is gimped on purpose by Warner. A lot of people including me bought the HD DVD version.

Now, if HappyFeet sells more in HD than PS3, what would you say ?I think he would say that the Happy Feet video game must have come out early on the PS3 ;). If you meant Blu-ray though you might want to wait and see which version of Happy Feet really does sell best. Personally speaking I think most of the movies released on both HD formats will get better sales on Blu-ray.

I would say, that's about right since the Blu-ray version is gimped on purpose by Warner. A lot of people including me bought the HD DVD version.
Or ignored it completely as I did, due to this shaft from Warner.

Now, if HappyFeet sells more in HD than PS3, what would you say ?

I assume you meant BD? Anyways ... with some serious playback issues for some, the question came up as to how (if at all) Neilsen VideoScan handles returns?

I would expect it to be a debit (for a return), credit (for a sale)? What happens if someones returns one format and gets it in another. Those would be different SKUs, so I assume a debit (for SKU1) and credit (for SKU2)? Are my assumptions likely correct?

Wouldn't the 360 having sold >4x as many units in the US as the PS/3 have something to do with that?

Gary


Wait, if you use Stormprophet's logic, since the 360 has a 4:1 ratio over the PS3, retailers should just about be ready to drop the PS3 inventory from their store shelves... :D

Wait, if you use Stormprophet's logic, since the 360 has a 4:1 ratio over the PS3, retailers should just about be ready to drop the PS3 inventory from their store shelves... :D

Well, being that 360 launched a year before PS3, I would say that your analogy between video game consoles and HD disc sales is incorrect.

Well, being that 360 launched a year before PS3, I would say that your analogy between video game consoles and HD disc sales is incorrect.

What about in Japan ? Wii is heavily outselling PS3.

What about in Japan ? Wii is heavily outselling PS3.

As Nintendo and Sony have made clear on many occasions, The Wii and PS3 are not competing platforms. They are complementary platforms. Plus, I don't see anyone saying how Japan is 95% Blu-ray, so American retailers should be dropping HD DVD now.

Wait, if you use Stormprophet's logic, since the 360 has a 4:1 ratio over the PS3, retailers should just about be ready to drop the PS3 inventory from their store shelves... :D

That is a much different argument. And retailers do take notice, by expanding and reducing the size of the department.

PS2s department was huge in comparison to GC and Xbox.

Currently in most retail stores BD and HD-DVD inhabit the same size floor space. Maybe they wouldn't pull space entirely, but they should think about pulling some of it.

Wii is gonna be getting more space at the moment I think. 360 is not outselling PS3 4 to 1. If you look at the overall figures maybe. In monthly sales it's about 2 to 1. Wii is outselling 3 to 1 right now.

The other thing is in the long run no matter how much space you give one gaming system (PS2 for instance), the other systems will sell as much per square ft if the size is reduced enough.

Now, if HappyFeet sells more in HD than BD, what would you say ?

I guess we will have to wait for some official figures. So far one of the only ones we have to compare is "The Departed"

Mostly what people compare is the Amazon sales which is no where near the complete picture.

As Nintendo and Sony have made clear on many occasions, The Wii and PS3 are not competing platforms.

Just spin ... that doesn't mean they are not competing to earn gamer's money.

I guess we will have to wait for some official figures. So far one of the only ones we have to compare is "The Departed"

Mostly what people compare is the Amazon sales which is no where near the complete picture.

Looks like from Departed experience, Amazon is representative.

In terms of top 5 movies in both formats, amazon definitely does not match videoscan numbers. Eragon and Rocky Balboa never moved much in amazon, but are last week's strong sellers according to videoscan. I'm beginning to suspect there are a lot of people who don't mind paying the higher B&M prices and etailer sale for highdef may not be representative of the general public, who may be driven more by impulse buying than good planning.

Top 5 BD
1 Casino Royale 100.00
2. Rocky Balboa 62.66
3. Eragon 57.72
4. The Departed 19.73
5. The Prestige 9.48

Top 5 HD DVD
1. The Departed 100.00
2. Batman Begins 69.69
3. Babel 54.55
4. Troy 50.26
5. Goodfellas 38.45

Looks like from Departed experience, Amazon is representative.

like hd dvd departed leading blu-ray for several days on amazon ?, and we know the results..., it will be interesting if same will happen to happy feat...

Marek

Looks like from Departed experience, Amazon is representative.

I'm not so sure. Around Departed release date (Feb. 13), amazon sales rankings were neck-and-neck as well. It was after Feb. 18 that the HD DVD started to lose ground.

I'm not so sure. Around Departed release date (Feb. 13), amazon sales rankings were neck-and-neck as well. It was after Feb. 18 that the HD DVD started to lose ground.
Not to mention Rocky Balboa and Eragon total sales were a lot healthier than amazon would indicate... Troy on HDDVD is also dead on amazon but still alive in B&M.

Other than total blowouts like CR launch, I suspect correlation between amazon and B&M is low.

Just spin ... that doesn't mean they are not competing to earn gamer's money.

Another thing to consider is that Microsoft and Sony are looking at new revenue streams from online content. Which Nintendo isn't really looking at.

It strikes me that they are in fact competing for different segments within the gaming market for many reasons.

A lot of Hi-def games won't even be on the Wii, and the Wii isn't designed to take advantage of the Hi-def market. I think the Wii was designed to convert DS disciples to the console fold.

Ultimately, nex generation, both Sony and Microsoft will have entertainment systems and if Nintendo decides to stay on its path it will be the only "console" left. And then they really will be competing in different markets.

Sony and Microsoft are much more focused on winning the living room completely and Nintendo is more focused on just making games.

Looks like from Departed experience, Amazon is representative.

That's not completely true. Initially Departed Sales on HD-DVD were greater. And they tend to be, I think because of the shortage of Hi-def releases available recently.

HD-DVD users have much more incentive to get newly announced releases as they haven't had much selection recently. Blu-ray supporters have to divide their monies over a larger selection overall, and a larger selection of new releases.

For many other examples, as pointed out, Amazon is not currently representative of a lot of the actual top sellers.

Another thing to consider is that Microsoft and Sony are looking at new revenue streams from online content.

I don't think so. Online is just to attract more sales.

Also, it seems a lot of PS games will be available on wii what with several devs shifting their dev focus from ps3 ro wii.

Nielson has been about 4 to 1 for 3 weeks now. Any thoughts on this guys? Traditionally 3:1 or 4:1 is a point where retailers start stocking one over the other.

Really ? When you say traditionally - it looks like it happens all the time. Can you give us examples of when this has happened ?

I can think of numerous examples where it is not true - like macs being stocked with just 5% or less of sales. Or the multitude of MP3 players which sell a fraction of what iPod does.

I don't think so. Online is just to attract more sales.

Also, it seems a lot of PS games will be available on wii what with several devs shifting their dev focus from ps3 ro wii.
PS3 ~= Xbox 360 for processing and graphical power. PS3 has lost a few exclusives to the X360 due to how small the market is for both, but that's not unexpected.

Both systems vastly outperform the Wii.

Thus far, all of the ported games to the Wii have been no better than mediocre, and quite a few are simply awful. I don't see any shift to the Wii for existing game franchises.

Stromprophet is correct. The PS3 and Xbox 360 are going after a similar segment of the market, expanding into more of a multimedia server role rather than simply a gaming appliance. The Wii is a more traditional look at console gaming, with a twist on interface.

The PS3 and Xbox 360 are going after a similar segment of the market, expanding into more of a multimedia server role rather than simply a gaming appliance. The Wii is a more traditional look at console gaming, with a twist on interface.

I understand that thats what 360 and PS3 want / target. But the reality finally trumps careful market segmentation. For all this talk of Wii being not a compitetor I'm sure 360/Ps3 teams wish the Wii owners had paid more and bought their consoles.

Let me put it this way. DVD-A and SACD always thought they were compitetors. But finally MP3 won. Just as we keep talking about HD DVD/BD and finally streaming/download may win.

I understand that thats what 360 and PS3 want / target. But the reality finally trumps careful market segmentation. For all this talk of Wii being not a compitetor I'm sure 360/Ps3 teams wish the Wii owners had paid more and bought their consoles.
More so for the PS3 than the Xbox 360, though I'm sure both are affected by the Wii in some respects (the 360 is less affected only because it had a 1 year head start on sales). Of course, I'm sure Nintendo is wishing Sony had abandoned the PS2 in favor of producing the PS3, since the Wii's real competition is from the PS2, which costs half as much and has many times as many available software titles mostly for under half of the cost.

It will be interesting to see how long Nintendo can keep the gravy train running on the Wii without good software ( I should note that I own a Wii, and the dearth of quality titles makes the March drought for HD DVD pale in comparison).


Let me put it this way. DVD-A and SACD always thought they were compitetors. But finally MP3 won. Just as we keep talking about HD DVD/BD and finally streaming/download may win.
This might very well be true, but I think a lot of people are overestimating the impact of online downloads. There just isn't enough bandwidth out there right now to support widespread HD downloads at anything resembling a reasonable amount of time with sufficient accuracy, and there won't be for a number of years yet.

I doubt any of us are under any illusion that BD or HD DVD are here to stay forever; something new will always come along. But for now - the next 5-10 years - they're probably the best thing we're going to get in that time period for the masses.

I understand that thats what 360 and PS3 want / target. But the reality finally trumps careful market segmentation. For all this talk of Wii being not a compitetor I'm sure 360/Ps3 teams wish the Wii owners had paid more and bought their consoles.

Let me put it this way. DVD-A and SACD always thought they were compitetors. But finally MP3 won. Just as we keep talking about HD DVD/BD and finally streaming/download may win.

Not downloads... HD VMD!

Too lazy to walk the mile to Blockbusters, but if the numbers here are right, shouldn't I expect to see 2-4 times more BD than HD disks there? Are there?

In terms of top 5 movies in both formats, amazon definitely does not match videoscan numbers. Eragon and Rocky Balboa never moved much in amazon, but are last week's strong sellers according to videoscan. I'm beginning to suspect there are a lot of people who don't mind paying the higher B&M prices and etailer sale for highdef may not be representative of the general public, who may be driven more by impulse buying than good planning.

Top 5 BD
1 Casino Royale 100.00
2. Rocky Balboa 62.66
3. Eragon 57.72
4. The Departed 19.73
5. The Prestige 9.48

Top 5 HD DVD
1. The Departed 100.00
2. Batman Begins 69.69
3. Babel 54.55
4. Troy 50.26
5. Goodfellas 38.45

Something is a little smelly here, Troy & Gooodfellas, which have barely had a pulse for the past 3 mo. on amazon, are ranked 4th & 5th for HD DVD on this week's N/V #s. This cannot be true.

Something is a little smelly here, Troy & Gooodfellas, which have barely had a pulse for the past 3 mo. on amazon, are ranked 4th & 5th for HD DVD on this week's N/V #s. This cannot be true.

not at all. The issue is you are assuming top 5 means something. The relevant data is the number at the end. Troy sold 50.24 of departed and GF 38.45 of departed. Lets assume HD DVD sold 9k (a bit more then what Icemage has in his chart but easier for the math) Now the top5 add up (roughly) to 3x (departed) if we assume all the other movies give an other 3x Departed(to keep the math easy) then we get that Departed sold ~1.5K. 50% of that is ~750 for Troy and 40% is ~600 copies.

and the numbers are most likely a bit less then that for those titles. So a few new players, a few people that finally decided to get it because they got the rest..... and any title can be top 5 any week

Something is a little smelly here, Troy & Gooodfellas, which have barely had a pulse for the past 3 mo. on amazon, are ranked 4th & 5th for HD DVD on this week's N/V #s. This cannot be true.Looking back at the rankings for each week that are listed in the second post at the beginning of this thread, it looks like "Troy" has been in the top 5 for HD DVD every week we have the top 5 for in 2007, except for one week. It doesn't seem to be popular on Amazon, but that doesn't mean it isn't selling well overall.

--Darin

Looking back at the rankings for each week that are listed in the second post at the beginning of this thread, it looks like "Troy" has been in the top 5 for HD DVD every week we have the top 5 for in 2007, except for one week. It doesn't seem to be popular on Amazon, but that doesn't mean it isn't selling well overall.

So, looks like it is an impulse buy or perhaps popular with older buyers (more unlikely to use online ) ?

PS3 ~= Xbox 360 for processing and graphical power. PS3 has lost a few exclusives to the X360 due to how small the market is for both, but that's not unexpected.

Both systems vastly outperform the Wii.

Thus far, all of the ported games to the Wii have been no better than mediocre, and quite a few are simply awful. I don't see any shift to the Wii for existing game franchises.

Stromprophet is correct. The PS3 and Xbox 360 are going after a similar segment of the market, expanding into more of a multimedia server role rather than simply a gaming appliance. The Wii is a more traditional look at console gaming, with a twist on interface.A bit off topic, but for what it's worth I don't entirely agree.

Anyone who has used the Wii just once, immediately understands why it has broken the mold completely and is a refreshing and completely new approach to video games. Very different from the "same old" arcade games with the "same old" twiddle thumb controllers.

The waving around of the wireless controllers is so easy and so intuitive that even my 4 year old is having a blast. I don't have any ported games, so I can't comment on that, but I CAN say that the 8 games that I DO have are fab (I am counting Wii Sports as 1 game).

Re: the 360 and PS3, the PS3 has a slightly more powerful CPU, but the 360 has a rather more powerful Graphics engine (it's really an early design of the soon-to-market ATI 600 chip, which is an absolute MONSTER).

A bit off topic, but for what it's worth I don't entirely agree.

Anyone who has used the Wii just once, immediately understands why it has broken the mold completely and is a refreshing and completely new approach to video games. Very different from the "same old" arcade games with the "same old" twiddle thumb controllers.

The waving around of the wireless controllers is so easy and so intuitive that even my 4 year old is having a blast. I don't have any ported games, so I can't comment on that, but I CAN say that the 8 games that I DO have are fab (I am counting Wii Sports as 1 game).

Re: the 360 and PS3, the PS3 has a slightly more powerful CPU, but the 360 has a rather more powerful Graphics engine (it's really an early design of the soon-to-market ATI 600 chip, which is an absolute MONSTER).
I own a Wii. It's really not that cool. The games on it are generally weak offerings. I've barely touched it in the last month.

Waving the controller is sometimes intuitive, but sometimes it's simply tedious and inaccurate.

Graphics-wise, I don't really see much difference between the 360 and the PS3, except that the PS3 can do 1080p (but few games offer this so it's not really an advantage), and developers are now a bit more familiar with the 360 graphics so they can do a bit more with the engines.

---

Back on topic, I'm currently working on a modification to my calculation spreadsheet that allows simultaneous calculation of the second set of data so a comparison can be made, rather than just have the second set be a magnification of the assumed Nielsen values. I'll release the new version when we have new Nielsen data later this week.

A bit off topic, but for what it's worth I don't entirely agree.

Anyone who has used the Wii just once, immediately understands why it has broken the mold completely and is a refreshing and completely new approach to video games. Very different from the "same old" arcade games with the "same old" twiddle thumb controllers.

The waving around of the wireless controllers is so easy and so intuitive that even my 4 year old is having a blast. I don't have any ported games, so I can't comment on that, but I CAN say that the 8 games that I DO have are fab (I am counting Wii Sports as 1 game)..Try doing combos in Prince of Persia before saying Wii control is intuitive.

Graphics-wise, I don't really see much difference between the 360 and the PS3, except that the PS3 can do 1080p
As can the 360...

Sorry if this question is OT or has already been answered, but how are Amazon pre-orders captured?

Nielsen tracks only disks that are sold and left stores, not preorders.

Thanks Neo.

Anyone who has used the Wii just once, immediately understands why it has broken the mold completely and is a refreshing and completely new approach to video games. Very different from the "same old" arcade games with the "same old" twiddle thumb controllers.Bull. Try playing "Cooking Mama" or the previously-mentioned PoP.
The waving around of the wireless controllers is so easy and so intuitive that even my 4 year old is having a blast. I don't have any ported games, so I can't comment on that, but I CAN say that the 8 games that I DO have are fab (I am counting Wii Sports as 1 game).That's because the games are made especially easy, particularly Wii Sports. For example, the bowling game is so easy as to make duckpins look like the PBA Championships.

Time.

Stop the Videogame talk or move it to the correct forum.

I don't think so. Online is just to attract more sales.

Also, it seems a lot of PS games will be available on wii what with several devs shifting their dev focus from ps3 ro wii.

How is it not a revenue stream? 360 is selling movies, music, game extras, a lot of the stuff is purchased. And with a couple million customers that provides another revenue stream they didn't have before. Sales that would have otherwise gone to retailers on some things instead Microsoft is becoming the retailer.

"Some" devs have shifted some resources. 1) Months, if not more than a year away from releasing any games, and 2) The games are dumbed down IMO. No GTAIV, RE5, FFXIII...remakes, or off name titles, like FF Chronicles

Really ? When you say traditionally - it looks like it happens all the time. Can you give us examples of when this has happened ?

I can think of numerous examples where it is not true - like macs being stocked with just 5% or less of sales. Or the multitude of MP3 players which sell a fraction of what iPod does.

You're dealing with an electronic equipment there. Much different than a DVD or CD. When did they abandon Betamax to VHS? Or when did stores stop stocking new VHS titles in favor of only DVD?

I'm quite certain it was at a tipping point where it made sense from a retail space and sales comparison to drop 1 in favor of the other.

The reason you can't drop say Xbox/GC when PS2 was pwning them is the product was different. Essentially HD-DVD and Blu-ray are the same final product. There is nothing really proprietary to either when you boil it down. (except Titles available or studios supporting).

Everyone will buy movies. And if there was only 1 format, they would certainly buy their movies still, on one format. With something like say consoles, that's not always true. If there was 1 format, you may be missing sales of people who don't want what the console maker is offering. Movie = Movie / Xbox no = GC no = PS2.

Another thing is that electronic equipment makers have more control over contracts to retailers than DVD publishers do. It's probably a lot easier for a retailer to pull Blu-ray or HD-DVD than it is to pull HP Computers, Or Macs, or Sony TVs because the legal obligations are likely greater.

Haven't retailers (Especially Bestbuy) been outspoken about only wanting one format? I don't see them clamoring for only 1 computer style, or 1 TV, because selection is important in these categories. A movie is a movie and having to stock more than 1 is just holding back sales.

You're dealing with an electronic equipment there.

The point I'm making is where did you get the figure when "traditionally retailers stop stocking". I'm questioning whether such a beast exists. If so what is the basis for the figure ? How do you know it is not 10:1 or 20:1 ?

How is it not a revenue stream?

The question is - is it just breaking even / making money / losing money. We all know they get a few bucks ... but what are the costs ? Have you done any calculations - have you seen any ?

Currently the models, I beleive, are more like the iTMS - apple doesn't make much moeny (if any at all) of the store - but sells a ton of iPod. So, iTMS becomes cost of doing business.

You're dealing with an electronic equipment there. Much different than a DVD or CD. When did they abandon Betamax to VHS? Or when did stores stop stocking new VHS titles in favor of only DVD?

I'm quite certain it was at a tipping point where it made sense from a retail space and sales comparison to drop 1 in favor of the other.

The reason you can't drop say Xbox/GC when PS2 was pwning them is the product was different. Essentially HD-DVD and Blu-ray are the same final product. There is nothing really proprietary to either when you boil it down. (except Titles available or studios supporting).

Everyone will buy movies. And if there was only 1 format, they would certainly buy their movies still, on one format. With something like say consoles, that's not always true. If there was 1 format, you may be missing sales of people who don't want what the console maker is offering. Movie = Movie / Xbox no = GC no = PS2.

Another thing is that electronic equipment makers have more control over contracts to retailers than DVD publishers do. It's probably a lot easier for a retailer to pull Blu-ray or HD-DVD than it is to pull HP Computers, Or Macs, or Sony TVs because the legal obligations are likely greater.

Haven't retailers (Especially Bestbuy) been outspoken about only wanting one format? I don't see them clamoring for only 1 computer style, or 1 TV, because selection is important in these categories. A movie is a movie and having to stock more than 1 is just holding back sales.

Dropping VHS did not happen until 2005, 8 years after DVD's introduction (shelf space started shrinking around 2002). The 2005 phase out was planned at the end of 2004, when retailers saw that big release titles (Spiderman 2, Shrek 2) were averaging a ~15:1 ratio in favor of DVD. Up until that point, VHS was still worth stocking.

I think for one format to "win", it will need to show sustained sales (average over the course of a year) of at least 10:1 (10:1 was a about the lowest ratio between DVD and VHS, with the highest at about 20:1) . At that point, retailers and studios may stop pursuing that format. I don't see any events for this year that will drive the ratio anywhere near 10:1 between BD and HD-DVD. I really don't see it happening in 2008 either.

My prediction is that by the end of the year, both formats are between a 1.5:1 - 1:1 ratio for the year. Where it goes from there is up to the consumer and studios. It depends alot if one format has sub-$200 players and the other does not, or if exclusive studios go neutral, or neutral studios exlcusive, etc.

My prediction is that by the end of the year, both formats are between a 1.5:1 - 1:1 ratio for the year. Where it goes from there is up to the consumer and studios. It depends alot if one format has sub-$200 players and the other does not, or if exclusive studios go neutral, or neutral studios exlcusive, etc.

yes it depends alot IF, IF, IF, IF..... and if nothing from this happen, we can end up with 5:1 easily for BR

Marek

Something is a little smelly here, Troy & Gooodfellas, which have barely had a pulse for the past 3 mo. on amazon, are ranked 4th & 5th for HD DVD on this week's N/V #s. This cannot be true. not at all. The issue is you are assuming top 5 means something. The relevant data is the number at the end. Troy sold 50.24 of departed and GF 38.45 of departed. Lets assume HD DVD sold 9k (a bit more then what Icemage has in his chart but easier for the math) Now the top5 add up (roughly) to 3x (departed) if we assume all the other movies give an other 3x Departed(to keep the math easy) then we get that Departed sold ~1.5K. 50% of that is ~750 for Troy and 40% is ~600 copies.

and the numbers are most likely a bit less then that for those titles. So a few new players, a few people that finally decided to get it because they got the rest..... and any title can be top 5 any week Real sales numbers could be much higher than the Nielson/Videoscan survey numbers. At the low volume of the survey capture, there can be a lot of variation as you scale those numbers up to the entire market.

But about the appearance of the older releases on the HD DVD sales charts.

The thing to remember is the relative strength in sales of the older releases in HD DVD (like Goodfella's and Batman Begins) could also be indicative of a steadily increasing installed base of new owners buying HD DVD players.

If you just bought a HD DVD player, in effect the entire already released catalog is in effect new releases..... ;)

Perhaps the best we can do is just track the Nielsen numbers. Meanwhile, imo, predicting the future is for a different forum and thread.

So far, we have tracked several weeks of Nielsen with sales ratios of 3:1 - 4.4:1, whether this becomes 3.5:1 or not, there should be some objective agreement how long a 3:1 or greater sales lead has to be maintained before some consensus is formed for who has the best chance of taking on DVD.

I think the matrix release complicates this since we now have to see what happens between now and may 22nd. But even if there's a stalemate that just means we drag it out longer, and JQP stays uninterested longer. There's nothing that says JQP has to buy into either format before they are ready and before the two formats decide to do something about some unification strategy.

At this point, I would even suggest that they start writing EVOBs under HVVIDEO_TS folders on a BD-ROM media (as long as the 28Mbps peak limit is extended to 40Mbps). I don't see any difference between m2ts and EVOB that would warrant all this mess.

Have the BD players handle EVOBs and M2TS demux with different software and be done with it.

But about the appearance of the older releases on the HD DVD sales charts.

The thing to remember is the relative strength in sales of the older releases in HD DVD (like Goodfella's and Batman Begins) could also be indicative of a steadily increasing installed base of new owners buying HD DVD players.

If you just bought a HD DVD player, in effect the entire already released catalog is in effect new releases..... ;)We already know that HD DVD players are selling, but the big question is the pace of sales. We saw one thing that said "Batman Begins" sold 4100 copies for the first 3 weeks of this year. I believe that its average ranking on Amazon has fallen since that period. It is too bad that we don't have the top 5 for the week ending February 18th, as we could compare its sales to the 13k reported for "The Departed" for its first week. For the next week "Batman Begins" sold at about 1/5th the pace of "The Departed". If the latter did 7k for that week (a reasonable droppoff from week 1 to week 2), then that would be 1500 for BB. If HD DVD players have been selling as well as the 200k Toshiba players you mentioned indicate (not even counting XBOX360 add-ons), then it seems that the attach rate for BB to those players would not be all that high. I already know you can work backwards and assume a poor attach rate for it to justify those hardware sales though. :)

--Darin

The thing to remember is the relative strength in sales of the older releases in HD DVD (like Goodfella's and Batman Begins) could also be indicative of a steadily increasing installed base of new owners buying HD DVD players.

If you just bought a HD DVD player, in effect the entire already released catalog is in effect new releases..... ;)
Correct. BD got a boost from some new PS3 users who saw the whole catalog as a new option.

But now, HD DVD player sales have continue to grow strongly, and we are seeing a balance return to a more even position.

BD's efforts to reduce player costs will be too little too late to prevent HD DVD from taking the sales lead again, IMO. The players that they are reducing are mostly the same old 1st gen-type players which are lacking in features that HD DVD players already have.

Furthermore, it seems obvious to any objective observer (say that fast :) ) that a big factor in BDA pushing back the mandatory BDJ player profiles is to allow them to continue selling the "cheap" feature-less machines for a longer time.

So even if they get the prices down to $600 this year, it'll be on less desirable machines that do not meet the 1.1 or 2.0 profile specifications.

At the end of the day, this is why HD DVD is clawing back the ground so fast...

At the end of the day, this is why HD DVD is clawing back the ground so fast...
I take it you mean amazon? The last nielsen weekly numbers are not exactly a photo finish. Don't know what this friday's number will be, or next week, or the week after that.

What do you think? When would the weekly sales ratio be back to 2:1 instead of the 4:1 that it is today?

This is a rather blindered argument.

Given that HD DVD and Bluray movies sales are so close right now, yet Bluray has 15 time the "players" in the market (counting the illustrious PS3, of course), are you SURE that you want to attempt a debate about "attach rates"?? ;)

aid Batman Begins" sold 4100 copies for the first 3 weeks of this year. I believe that its average ranking on Amazon has fallen since that period. It is too bad that we don't have the top 5 for the week ending February 18th, as we could compare its sales to the 13k reported for "The Departed" for its first week. For the next week "Batman Begins" sold at about 1/5th the pace of "The Departed". If the latter did 7k for that week (a reasonable droppoff from week 1 to week 2), then that would be 1500 for BB. If HD DVD players have been selling as well as the 200k Toshiba players you mentioned indicate (not even counting XBOX360 add-ons), then it seems that the attach rate for BB to those players would not be all that high. I already know you can work backwards and assume a poor attach rate for it to justify those hardware sales though. :)

--Darin

What do you think? When would the weekly sales ratio be back to 2:1 instead of the 4:1 that it is today?
Wow - 4 to 1 - - link please? not living in the past, eh? :)

This is a rather blindered argument.

Given that HD DVD and Bluray movies sales are so close right now, yet Bluray has 15 time the "players" in the market (counting the illustrious PS3, of course), are you SURE that you want to attempt a debate about "attach rates"?? ;)Sure. If someone wants to prove that they aren't very intelligent (even after just saying they were) by making claims about attach rates that show a lack of intelligence, then they can go ahead. I do wonder if those making claims like 15 times the players where they count players in Europe that just sold in the last 2 weeks and sales in Japan while only counting software sales in the US are doing things like that purposely in order to deceive, or if they really don't know enough about this to keep from making claims like that.

Please, back up your 15 times the players by telling us how many of each and then explain why you would use overseas players. Blinded or purposely deceptive would describe that claim, IMO.

BTW: If you didn't know a year ago that the PS3 strategy would result in lower attach rates per player but likely higher software sales overall then ...

--Darin

Wow - 4 to 1 - - link please? not living in the past, eh? :)

No, apparently, he is living in the today ... current ... whereas predictions like yours are starting to sound more and more like what you bombasted Blu-ray about in the earliest days ...

"Just wait until ..."

<fill-in-the-blanks>

"Just wait until ..." <fill-in-the-blanks>"Just wait until... "? Like "Just wait until BDJ 1.1 and 2.0 launches - THEN you'll be sorry!" :p

or "Just wait until BD has true IME on their releases - ooooh, you gonna be sorry then!"

or "Just wait until The Matrix launches on BD!"

"Just wait until BD player prices and capabilities and specs match HD DVD, you'll be sooorry!" :)

or what about.. nah, too easy :)

rdjam,

Are you going to tell us what numbers you use for, ".... yet Bluray has 15 time the "players" in the market (counting the illustrious PS3, of course),... "? You've made claims like this enough that you should be able to give us some ballpark numbers and an explanation of why you chose to use those numbers.

What number do you use for PS3s?
What number do you use for HD DVD players?

Or did you just make the 15x up because it sounded good to you?

--Darin

"Just wait until... "? Like "Just wait until BDJ 1.1 and 2.0 launches - THEN you'll be sorry!" :p

or "Just wait until BD has true IME on their releases - ooooh, you gonna be sorry then!"

or "Just wait until The Matrix launches on BD!"

"Just wait until BD player prices and capabilities and specs match HD DVD, you'll be sooorry!" :)

or what about.. nah, too easy :)

LOL. I love your spunky (never-say-die) style! Let me match you, or at least try ... LOL

1) The vast majority of peeps buy movies for the movie, not for additional content (OK, I'll admit to buying the BD "LOEG", just so I could try the BD-J game -- guess what, it worked, but it also sucked big time);

2) Just wait until those Chinese HD DVD players (with 14-day warranties arrive).

For any rational person, that was a silver bullet, but for 'lil 'ole you, I'm sure it wouldn't be ... Perhaps a wooden stake? or maybe some garlic ;)

So far, we have tracked several weeks of Nielsen with sales ratios of 3:1 - 4.4:1, whether this becomes 3.5:1 or not, there should be some objective agreement how long a 3:1 or greater sales lead has to be maintained before some consensus is formed for who has the best chance of taking on DVD.


I see no reason why sales (very small amounts at that) ratio has anything to do with who can take over DVD. One format can clearly with this hidef "battle" only to lose the war to DVD.

What really needs to happen to take on DVD is (not just my idea, but some MR I've read) :

- Players in the $200 range
- Movie prices no more than SD (may be just a couple of bucks more)

The format that achieves these first, can take on DVD and try to win.

All the PS3s in the world will not help you get there. I can clearly see how BD can win the pygmies war but be beaten by DVD & HD DVD in the long run. This is like chosing your primary candidate - do you want someone with partisan appeal or general appeal ;)

It's just that I don't see a true price delta in the player beyond the blue diodes.

It is really a question of strategy. Toshiba strategy is to produce cheap players - so they focus on that strategy and get it to work (MS + Broadcom platform for instance). Sony strategy is two fold - PS3 and buttering studios.

We will have to see what wins out. There is no reason why a BD player can't be made for the same price as a HD DVD player - but does Sony (and others) think that is important and spend effort and money towards that goal ?

It is really a question of strategy. Toshiba strategy is to produce cheap players - so they focus on that strategy and get it to work (MS + Broadcom platform for instance). Sony strategy is two fold - PS3 and buttering studios.

We will have to see what wins out. There is no reason why a BD player can't be made for the same price as a HD DVD player - but does Sony (and others) think that is important and spend effort and money towards that goal ?

The HD players are not that much cheaper from a components standpoint. Toshiba is just taking a loss or near loss on the sales in an effort to make it up on software sales of HD-DVD discs.

It's about the same as Sony selling the PS3 at a pretty big loss. But most of the BD manufacturers aren't wanting to sell at a loss, and they don't have to in the position their in.

Before you accuse Sony, note they are selling the PS3 at a big loss to the cost of the components, it's really quite a value.