Dish Network Predicts Clinton Victory using Viewers to Voters Data - AVS Forum | Home Theater Discussions And Reviews
Forum Jump: 
 23Likes
Reply
 
Thread Tools
post #1 of 37 Old 11-07-2016, 07:34 AM - Thread Starter
Mark Henninger
 
imagic's Avatar
 
Join Date: Dec 2005
Location: Philadelphia, PA
Posts: 12,475
Mentioned: 278 Post(s)
Tagged: 0 Thread(s)
Quoted: 6611 Post(s)
Liked: 11384
Dish Network (Incorrectly) Predicts Clinton Victory using Viewers to Voters Data

When you watch TV, your service provider is watching your viewing habits. Normally, this sort of data would be used to determine advertising rates and factor into whether a show makes the cut or gets canceled. The Viewers to Voters predictive model was co-developed by Dish’s Data Science and Media Sales divisions and is the first project to come out of the company's Wisdom of Crowds initiative. The program seeks to explore verifiable correlations between viewing habits and people's actions in the real world, including those of voters.

With its Viewers to Voters predictive model, pay-TV provider Dish took viewer behavior data, filtered it, processed it, and used it to issue predictions for tomorrow's US presidential election. Based on that methodology, the company is calling for a Clinton victory.

Aside from picking the next president, the model also predicts that Republicans will maintain a majority in the nation's 115th Congress, with Democrats only gaining two seats in the House of Representatives.

Per Dish, the Viewers to Voters model was applied to the 2014 House elections, and its reliability was pegged at 98%.Warren Schlichting, exec. VP of media sales, marketing and programming noted that "With so much focus around national polling, we thought it’d be interesting to see if we could find a correlation between how our customers interact with DISH and how they may vote. We recognize that our call on the distribution of seats in the House may be an outlier. Yet when we tested the model against 2014 House elections, we found that we were able to predict the outcome at a 98 percent reliability point."

The program analyzed the relationship between viewing habits and political affiliation. According to Dish, "The model determined that customers who watched more sports, religious or family-oriented television were more likely to vote Republican. Examples include NFL Football, GEB America’s Giving Hope, Sharing Life and PBS' Sesame Street."
Dish then went on to describe the viewing habits of Dems: "In contrast, customers who watched more series/specials, education, or music-oriented television were more likely to vote for Democrats. Examples include ABC’s Scandal, Discovery’s MythBusters and NBC’s The Voice."

I have a big favor to ask everybody commenting on this thread: Please don't turn this into a partisan discussion. It's against AVS Forum rules to explicitly discuss politics, so please stick with discussing how viewership data and algorithms might be used to predict the behavior of large groups of people. In the context of the election, this post is about TV viewing habits and analytical models and whether applying algorithms to that data can compete with polling. If the conversation goes downhill, I'll be forced to ask the moderators to shut the thread down. Soon, we will know how well this genuinely new approach to predicting election results works. So again, please keep it civil, discuss the topic at hand, and when tomorrow rolls around… vote.
sdurani and rramacha like this.

Mark Henninger, Senior Editor at AVS Forum

Last edited by imagic; 11-09-2016 at 03:31 AM.
imagic is offline  
Sponsored Links
Advertisement
 
post #2 of 37 Old 11-07-2016, 08:13 AM
Senior Member
 
rramacha's Avatar
 
Join Date: Dec 2009
Location: Santa Clara CA
Posts: 325
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 121 Post(s)
Liked: 34
This is interesting read from a data science perspective. I am still not sure how true this generalization would be. Assuming majority watching Nascar will vote Republican and majority watching Le Mans will vote Democrat seems over simplifying in my mind. In this case we don't have to wait too long to see how these predictions turn out. Even if its turns out correct, I am not sure if that kind of generalization holds true.

Cheers
rramacha is offline  
post #3 of 37 Old 11-07-2016, 08:32 AM
AVS Forum Special Member
 
Join Date: Jan 2015
Location: United States
Posts: 1,999
Mentioned: 20 Post(s)
Tagged: 0 Thread(s)
Quoted: 963 Post(s)
Liked: 1134
How pathetic,

I come to this site to NOT hear political agendas and what do I see? You betcha! Politics on the front page...

Most voters under 35 DON'T have cable, dish or whatever...they don't count. Sports TV is declining, news TV is declining... so that screws up the results. I guess it makes sense for the people that actually watch TV--for most people, they use the net so there ya go.

The other flaw is the "two party" system, there are four candidates so Dish just dropped that ball. The list goes on and on.

Not that AT&T and Dish don't pay off politicians--that will never happen!

My advice? Go back to what your title suggests you are and leave the political, equipment and whatever else advertising alone. Use that time to measure the SVS sub data-bass style--throw a little science and testing into it.

So don't start the problem and get offended when you get called out. I've already voted and there are plenty of people from around the world that come to this site--they as well as I, prefer to not hear the political shills and bogus polls throw in our face.

Time to get back to the regularly scheduled program.
noshank, Guygasm, Class A and 1 others like this.
18Hurts is online now  
 
post #4 of 37 Old 11-07-2016, 08:46 AM - Thread Starter
Mark Henninger
 
imagic's Avatar
 
Join Date: Dec 2005
Location: Philadelphia, PA
Posts: 12,475
Mentioned: 278 Post(s)
Tagged: 0 Thread(s)
Quoted: 6611 Post(s)
Liked: 11384
Quote:
Originally Posted by 18Hurts View Post
How pathetic,

I come to this site to NOT hear political agendas and what do I see? You betcha! Politics on the front page...

Most voters under 35 DON'T have cable, dish or whatever...they don't count. Sports TV is declining, news TV is declining... so that screws up the results. I guess it makes sense for the people that actually watch TV--for most people, they use the net so there ya go.

The other flaw is the "two party" system, there are four candidates so Dish just dropped that ball. The list goes on and on.

Not that AT&T and Dish don't pay off politicians--that will never happen!

My advice? Go back to what your title suggests you are and leave the political, equipment and whatever else advertising alone. Use that time to measure the SVS sub data-bass style--throw a little science and testing into it.

So don't start the problem and get offended when you get called out. I've already voted and there are plenty of people from around the world that come to this site--they as well as I, prefer to not hear the political shills and bogus polls throw in our face.

Time to get back to the regularly scheduled program.
I'm not offended at your being agitated. If you think you are calling something out, it's your prerogative. I'll count this as a protest vote. Also, don't expect CEA-2010 numbers from me. Someone else is gonna do that, and there's no real reason to duplicate the effort. Thanks for the suggestion though.

Mark Henninger, Senior Editor at AVS Forum

Last edited by imagic; 11-07-2016 at 08:53 AM.
imagic is offline  
post #5 of 37 Old 11-07-2016, 10:08 AM
AVS Forum Special Member
 
wxman's Avatar
 
Join Date: Apr 2003
Posts: 8,180
Mentioned: 24 Post(s)
Tagged: 0 Thread(s)
Quoted: 4536 Post(s)
Liked: 3115
I wear a tin foil hat while watching tv. They will never be able to read my mind.

LG 65E6P FW 05.30.02
LG 65EF9500
wxman is online now  
post #6 of 37 Old 11-07-2016, 10:24 AM
AVS Forum Special Member
 
Join Date: Jan 2015
Location: United States
Posts: 1,999
Mentioned: 20 Post(s)
Tagged: 0 Thread(s)
Quoted: 963 Post(s)
Liked: 1134
If you would rather review than test, I get it.

Testing is a serious pain in the butt, takes a lot of time and with heavy subs--not back friendly. I figure one of the forum members will test it, eventually. Maybe you guys could hook up the forum members to test and verify it's specs? Throw a link to the forums where the testing is done for one stop shopping.

Looking forward to your opinion on the SVS sub and post election wrap-up.
18Hurts is online now  
post #7 of 37 Old 11-07-2016, 11:02 AM - Thread Starter
Mark Henninger
 
imagic's Avatar
 
Join Date: Dec 2005
Location: Philadelphia, PA
Posts: 12,475
Mentioned: 278 Post(s)
Tagged: 0 Thread(s)
Quoted: 6611 Post(s)
Liked: 11384
Quote:
Originally Posted by 18Hurts View Post
If you would rather review than test, I get it.

Testing is a serious pain in the butt, takes a lot of time and with heavy subs--not back friendly. I figure one of the forum members will test it, eventually. Maybe you guys could hook up the forum members to test and verify it's specs? Throw a link to the forums where the testing is done for one stop shopping.

Looking forward to your opinion on the SVS sub and post election wrap-up.
I did not mean that in the abstract sense; someone else who knows what they are doing is testing a PB16-Ultra using CEA-2010 methodology. I'm not at liberty to say who, only that I know it's happening.

I am limited by the simple fact I live in Center City, Philadelphia. There is no back yard or field I can use, everything I do has to include the room in the equation. And yeah, with this specific sub it's not merely that it's not back friendly, it's the fact that it fully exceeds the threshold of what I can reasonable move around on my own, and sadly I do not have an assistant. Plus, none of my friends who will give me an occasional hand are bodybuilders.

Anyhow, my understanding is that reliable measurements are on their way, and I'm sure links will be shared when that happens.

Mark Henninger, Senior Editor at AVS Forum
imagic is offline  
post #8 of 37 Old 11-07-2016, 12:34 PM
 
NorthSky's Avatar
 
Join Date: Jan 2013
Location: Star of the Northern Hemisphere
Posts: 16,653
Mentioned: 26 Post(s)
Tagged: 0 Thread(s)
Quoted: 7012 Post(s)
Liked: 3557
Hi Mark,

Dish Network; is this limited to USA or North America?
NorthSky is offline  
post #9 of 37 Old 11-07-2016, 01:48 PM
AVS Forum Special Member
 
Join Date: Jan 2015
Location: United States
Posts: 1,999
Mentioned: 20 Post(s)
Tagged: 0 Thread(s)
Quoted: 963 Post(s)
Liked: 1134
Quote:
Originally Posted by imagic View Post
.

I am limited by the simple fact I live in Center City, Philadelphia. There is no back yard or field I can use, everything I do has to include the room in the equation. And yeah, with this specific sub it's not merely that it's not back friendly, it's the fact that it fully exceeds the threshold of what I can reasonable move around on my own, and sadly I do not have an assistant. Plus, none of my friends who will give me an occasional hand are bodybuilders.
Get Ricci to do it, that nut builds two 500 pound subwoofers and hauls them to a field fine. I would hazard to guess the DSL TH221 cinema sub won't hit the review block anytime soon. I build all my subs out of 3/4" plywood for that reason, the initial cost savings is far outweighed by back strain expected by moving it around for it's expected life cycle. Kept modifying my garage subs and eventually they exceeded one of my standards: they should not weigh more than me. 185 pound subs, 165 pound guy so I installed handles on them and use a dolly.

Friends don't let friends buy subs made of MDF. The back you save might be your own.

Northsky, Dish is owned by AT&T which is a company that was a monopoly then broken up in the 1980's. They have never been known for ethics but do have deep pockets.
18Hurts is online now  
post #10 of 37 Old 11-07-2016, 01:53 PM
Super Moderator
 
teachsac's Avatar
 
Join Date: May 2006
Location: Elk Grove, CA
Posts: 9,854
Mentioned: 33 Post(s)
Tagged: 0 Thread(s)
Quoted: 861 Post(s)
Liked: 1397
Directv is owned by AT&T not Dish. Charley still owns Dish.

Please use the report post button to alert staff to problematic posts. Never quote or respond to them yourself.

Oppo UDP-203 -> Yamaha CX-A5100 -> Sony XBR-75X940C; Mediabridge 6' and 15' HDMI cables.
teachsac is offline  
post #11 of 37 Old 11-07-2016, 03:57 PM
AVS Forum Special Member
 
Sean Spamilton's Avatar
 
Join Date: Jul 2012
Location: Toronto, ON
Posts: 1,248
Mentioned: 3 Post(s)
Tagged: 0 Thread(s)
Quoted: 376 Post(s)
Liked: 431
Predicted results aside, I can't see how using a population that includes ONLY those who can afford Dish-TV would be a representative sample of the population of all voting Americans. Honestly the amount of bias introduced here makes this result completely meaningless.
brian6751 and RickyDeg like this.
Sean Spamilton is online now  
post #12 of 37 Old 11-07-2016, 04:16 PM - Thread Starter
Mark Henninger
 
imagic's Avatar
 
Join Date: Dec 2005
Location: Philadelphia, PA
Posts: 12,475
Mentioned: 278 Post(s)
Tagged: 0 Thread(s)
Quoted: 6611 Post(s)
Liked: 11384
Quote:
Originally Posted by Sean Spamilton View Post
Predicted results aside, I can't see how using a population that includes ONLY those who can afford Dish-TV would be a representative sample of the population of all voting Americans. Honestly the amount of bias introduced here makes this result completely meaningless.
It seems far-fetched, but if the model is accurate then is it still meaningless? Won't know until tomorrow.

Mark Henninger, Senior Editor at AVS Forum
imagic is offline  
post #13 of 37 Old 11-07-2016, 04:20 PM
AVS Forum Special Member
 
Sean Spamilton's Avatar
 
Join Date: Jul 2012
Location: Toronto, ON
Posts: 1,248
Mentioned: 3 Post(s)
Tagged: 0 Thread(s)
Quoted: 376 Post(s)
Liked: 431
Quote:
Originally Posted by imagic View Post
It seems far-fetched, but if the model is accurate then is it still meaningless?
Yes.

I could have a race with my goldfish with the winning fish predicting the results of the election. Just because the election results match those from my goldfish race doesn't mean that my predictive model has any worth beyond the enjoyment gleaned by watching them race.

Sorry
brian6751, Guygasm and zoetmb like this.
Sean Spamilton is online now  
post #14 of 37 Old 11-07-2016, 04:26 PM
AVS Forum Addicted Member
 
sdurani's Avatar
 
Join Date: Oct 2001
Location: Monterey Park, CA
Posts: 25,092
Mentioned: 114 Post(s)
Tagged: 0 Thread(s)
Quoted: 5146 Post(s)
Liked: 3624
IF they're discovered a way to correlate viewing habits with election outcomes, claiming "98 percent reliability", then that's impressive. There are polling organizations that would probably want to see what formula Dish is using. Guess we'll find out in a couple days how accurate they are (especially the part about only 2 seats swapping in the House of Representatives).

Sanjay
sdurani is online now  
post #15 of 37 Old 11-07-2016, 04:52 PM
AVS Forum Special Member
 
brian6751's Avatar
 
Join Date: Mar 2006
Location: Canton, MI
Posts: 6,028
Mentioned: 33 Post(s)
Tagged: 0 Thread(s)
Quoted: 1173 Post(s)
Liked: 1158
Quote:
Originally Posted by Sean Spamilton View Post
Predicted results aside, I can't see how using a population that includes ONLY those who can afford Dish-TV would be a representative sample of the population of all voting Americans. Honestly the amount of bias introduced here makes this result completely meaningless.
Quote:
Originally Posted by Sean Spamilton View Post
Yes.

I could have a race with my goldfish with the winning fish predicting the results of the election. Just because the election results match those from my goldfish race doesn't mean that my predictive model has any worth beyond the enjoyment gleaned by watching them race.

Sorry
Right. The best they could hope for is predicting who most Dish customers will vote for. I don't think the Dish customer base is a good sample of American voters lol. This is too basic to even be on this site IMO. It's just silly.
Guygasm likes this.

------------------------------------------------------------
That's not science! - Cyborg

FOR SALE: nothing at the moment

Rivers Edge Theater
brian6751 is online now  
post #16 of 37 Old 11-07-2016, 04:53 PM
AVS Forum Special Member
 
brian6751's Avatar
 
Join Date: Mar 2006
Location: Canton, MI
Posts: 6,028
Mentioned: 33 Post(s)
Tagged: 0 Thread(s)
Quoted: 1173 Post(s)
Liked: 1158
Their criteria for deciding who watches what is insulting actually.
Guygasm likes this.

------------------------------------------------------------
That's not science! - Cyborg

FOR SALE: nothing at the moment

Rivers Edge Theater
brian6751 is online now  
post #17 of 37 Old 11-08-2016, 05:04 AM
AVS Forum Special Member
 
Class A's Avatar
 
Join Date: Jun 2010
Location: NH
Posts: 2,481
Mentioned: 5 Post(s)
Tagged: 0 Thread(s)
Quoted: 434 Post(s)
Liked: 1739
Question

Was this registered voters or couch potatoes likely to sit on their ass??
Class A is offline  
post #18 of 37 Old 11-08-2016, 05:12 AM - Thread Starter
Mark Henninger
 
imagic's Avatar
 
Join Date: Dec 2005
Location: Philadelphia, PA
Posts: 12,475
Mentioned: 278 Post(s)
Tagged: 0 Thread(s)
Quoted: 6611 Post(s)
Liked: 11384
Quote:
Originally Posted by brian6751 View Post
Their criteria for deciding who watches what is insulting actually.
I must admit, those examples left me baffled as well. Hey, Dish sent out the press release and I posted it because it legit is topical. But, I'm more than happy to have the comments examine the efficacy of the company's approach. I do hope there is a follow-up press release that discusses how well the predictive model worked.

Mark Henninger, Senior Editor at AVS Forum
imagic is offline  
post #19 of 37 Old 11-08-2016, 05:14 AM
Member
 
Join Date: Jan 2015
Location: South of Heaven, East of Eden
Posts: 24
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 11 Post(s)
Liked: 6
Sample Size

Quote:
Originally Posted by Sean Spamilton View Post
Predicted results aside, I can't see how using a population that includes ONLY those who can afford Dish-TV would be a representative sample of the population of all voting Americans. Honestly the amount of bias introduced here makes this result completely meaningless.
Last time I checked, <15% of the TV watching population were OTA. Dish has 13.6mm subs, so the ability to get a statistically significant sample size is easily there. There is a lot of precedent in cognitive neuroscience for figuring out political affiliations by other preferences. I am impressed by the effort to do this, and tie it to elections, but a little concerned about the potential for violating someone's privacy.

Angling for perfection with tangential success...
Absorbed & Diffused is offline  
post #20 of 37 Old 11-08-2016, 02:41 PM
AVS Forum Special Member
 
Sean Spamilton's Avatar
 
Join Date: Jul 2012
Location: Toronto, ON
Posts: 1,248
Mentioned: 3 Post(s)
Tagged: 0 Thread(s)
Quoted: 376 Post(s)
Liked: 431
Quote:
Originally Posted by Absorbed & Diffused View Post
Last time I checked, <15% of the TV watching population were OTA. Dish has 13.6mm subs, so the ability to get a statistically significant sample size is easily there. There is a lot of precedent in cognitive neuroscience for figuring out political affiliations by other preferences. I am impressed by the effort to do this, and tie it to elections, but a little concerned about the potential for violating someone's privacy.
I'm not sure why you've got Sample Size bolded above; though sample size is indeed an integral component of statistical models (hypothesis test, confidence interval etc) - we really don't need a very large sample for a proper estimate. A sample size of ~400 people would be sufficient for the WHOLE of the United States, this is large enough for the distribution to approximate a normal model as long as the sample has been sufficiently randomized.

I'm talking about Selection Bias, whereby the pool the sample is drawn from is NOT representative of the population you're looking at.

This sample is only drawn from those who own DISH Television. Dish Television subscribers (13.9 million as of April 2016) aren't necessarily representative of the demographic distribution or political leanings of the 320 million people who reside in the US. If one could prove within a statistically significant margin of error that they WERE, then that would be a different story - however they don't mention that here. Based on their assertion that 'those that watch sports are more inclined to vote Republican' alone I have serious doubts about the randomness of this sample.

If you were to stratify those groups that watch satellite, cable, don't have televisions, watch OTA, internet etc. - then run a hypothesis test, I'd be more inclined to give credence to the results of their testing, but based on their (questionable) assertion that 'those that watch sports are more likely to vote republican' alone, I have serious doubts about their methodology.

I'd be interested to know exactly how they came up with their model. If it's based purely on Dish subscribers, randomization has not been achieved. Regardless, I do share your concerns about privacy - but it seems voters in the US, Canada and elsewhere are willing to forgo that (and their rights) based on fear.

Last edited by Sean Spamilton; 11-08-2016 at 02:45 PM.
Sean Spamilton is online now  
post #21 of 37 Old 11-08-2016, 03:05 PM
Member
 
Join Date: Jan 2015
Location: South of Heaven, East of Eden
Posts: 24
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 11 Post(s)
Liked: 6
Quote:
Originally Posted by Sean Spamilton View Post
I'm not sure why you've got Sample Size bolded above; though sample size is indeed an integral component of statistical models (hypothesis test, confidence interval etc) - we really don't need a very large sample for a proper estimate. A sample size of ~400 people would be sufficient for the WHOLE of the United States, this is large enough for the distribution to approximate a normal model as long as the sample has been sufficiently randomized.

I'm talking about Selection Bias, whereby the pool the sample is drawn from is NOT representative of the population you're looking at.

This sample is only drawn from those who own DISH Television. Dish Television subscribers (13.9 million as of April 2016) aren't necessarily representative of the demographic distribution or political leanings of the 320 million people who reside in the US. If one could prove within a statistically significant margin of error that they WERE, then that would be a different story - however they don't mention that here. Based on their assertion that 'those that watch sports are more inclined to vote Republican' alone I have serious doubts about the randomness of this sample.

If you were to stratify those groups that watch satellite, cable, don't have televisions, watch OTA, internet etc. - then run a hypothesis test, I'd be more inclined to give credence to the results of their testing, but based on their (questionable) assertion that 'those that watch sports are more likely to vote republican' alone, I have serious doubts about their methodology.

I'd be interested to know exactly how they came up with their model. If it's based purely on Dish subscribers, randomization has not been achieved. Regardless, I do share your concerns about privacy - but it seems voters in the US, Canada and elsewhere are willing to forgo that (and their rights) based on fear.
Bold was unintentional - subject line only. I understand the nuance about selection bias, but I am genuinely curious about why you think Direct TV would be drawing a sample that is not substantially representative of the voting population of the U.S.? What would be the differentiating factor(s)?

Angling for perfection with tangential success...
Absorbed & Diffused is offline  
post #22 of 37 Old 11-08-2016, 03:47 PM
AVS Forum Special Member
 
Sean Spamilton's Avatar
 
Join Date: Jul 2012
Location: Toronto, ON
Posts: 1,248
Mentioned: 3 Post(s)
Tagged: 0 Thread(s)
Quoted: 376 Post(s)
Liked: 431
Quote:
Originally Posted by Absorbed & Diffused View Post
Bold was unintentional - subject line only. I understand the nuance about selection bias, but I am genuinely curious about why you think Direct TV would be drawing a sample that is not substantially representative of the voting population of the U.S.? What would be the differentiating factor(s)?
Without knowing the geographic, demographic and psycho-graphic breakdown of their subscribership it would be difficult for me to give you an exact answer, however some factors could be:

Income level: do the people who can afford DISH have potentially differing political agendas than those that don't?; Geographic distribution: are subscriber locations biased more towards the west coast / northern states as opposed to the Midwest?; Psychographic characteristics: do those that prefer to spend more time enjoying a 'green' lifestyle spend less time watching television than the majority of Dish subscribers - and how do those characteristics influence their political leanings?; Age: Are retired people apt to watch more television than those in the workforce, and how does that confound the study?; Ethnicity: is the ethnic distribution of their subscribership representative of that of the United States as a whole?

I'm sure there are others, but that's what I can come up with off the top of my head.
Sean Spamilton is online now  
post #23 of 37 Old 11-08-2016, 05:18 PM
Member
 
Join Date: Jan 2015
Location: South of Heaven, East of Eden
Posts: 24
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 11 Post(s)
Liked: 6
Quote:
Originally Posted by Sean Spamilton View Post
Without knowing the geographic, demographic and psycho-graphic breakdown of their subscribership it would be difficult for me to give you an exact answer, however some factors could be:

Income level: do the people who can afford DISH have potentially differing political agendas than those that don't?; Geographic distribution: are subscriber locations biased more towards the west coast / northern states as opposed to the Midwest?; Psychographic characteristics: do those that prefer to spend more time enjoying a 'green' lifestyle spend less time watching television than the majority of Dish subscribers - and how do those characteristics influence their political leanings?; Age: Are retired people apt to watch more television than those in the workforce, and how does that confound the study?; Ethnicity: is the ethnic distribution of their subscribership representative of that of the United States as a whole?

I'm sure there are others, but that's what I can come up with off the top of my head.
All very good points, with geographic being the one that I think might be the strongest (and quite possibly correlated with ethnicity, psychographic and income). I took a glance at their filings and they don't disclose subscriber geographies that I have seen. However, I would have to think that there would be controls for that in their analysis. It doesn't help your predictive model much if 70% of your sample is from Blue or Red states, correct? I believe Nielsen pretty much set the standard for demographic analysis of viewership, and DISH is taking it one step further. Without more info, we are speculating, unfortunately.

Angling for perfection with tangential success...
Absorbed & Diffused is offline  
post #24 of 37 Old 11-08-2016, 10:27 PM
Senior Member
 
rramacha's Avatar
 
Join Date: Dec 2009
Location: Santa Clara CA
Posts: 325
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 121 Post(s)
Liked: 34
Quote:
Originally Posted by rramacha View Post
This is interesting read from a data science perspective. I am still not sure how true this generalization would be. Assuming majority watching Nascar will vote Republican and majority watching Le Mans will vote Democrat seems over simplifying in my mind. In this case we don't have to wait too long to see how these predictions turn out. Even if its turns out correct, I am not sure if that kind of generalization holds true.

Cheers
Dish screwed this up badly
rramacha is offline  
post #25 of 37 Old 11-09-2016, 01:48 AM - Thread Starter
Mark Henninger
 
imagic's Avatar
 
Join Date: Dec 2005
Location: Philadelphia, PA
Posts: 12,475
Mentioned: 278 Post(s)
Tagged: 0 Thread(s)
Quoted: 6611 Post(s)
Liked: 11384
So much for predictive models

Mark Henninger, Senior Editor at AVS Forum
imagic is offline  
post #26 of 37 Old 11-09-2016, 03:17 AM
AVS Forum Special Member
 
Class A's Avatar
 
Join Date: Jun 2010
Location: NH
Posts: 2,481
Mentioned: 5 Post(s)
Tagged: 0 Thread(s)
Quoted: 434 Post(s)
Liked: 1739
Talking

Dewey DEFEATS Truman!!!!!!!!!!!!
Class A is offline  
post #27 of 37 Old 11-09-2016, 06:59 AM
Member
 
noshank's Avatar
 
Join Date: May 2005
Posts: 42
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 7 Post(s)
Liked: 12
DIRECTV user here. Bwahahaha


Sent from my iPhone using Tapatalk
noshank is online now  
post #28 of 37 Old 11-09-2016, 07:10 AM
AVS Forum Addicted Member
 
thebland's Avatar
 
Join Date: Jan 2001
Location: Detroit, Michigan USA
Posts: 25,765
Mentioned: 25 Post(s)
Tagged: 0 Thread(s)
Quoted: 1676 Post(s)
Liked: 1012
With his Tax policy promised cuts and repealing Obamacare, Im expecting a lot of extra Home Theater cash in my pocket!!!
Edllguy likes this.

Goodbye to a great audio and video genius and writer... JOHN GANNON. I enjoyed your friendship, wit and a nice long run we took around Indianapolis at CEDIA years back... and for buying my Runco 980 Ultra years back... you saved my ass! Rest in peace.
thebland is offline  
post #29 of 37 Old 11-09-2016, 08:11 AM
AVS Forum Addicted Member
 
sdurani's Avatar
 
Join Date: Oct 2001
Location: Monterey Park, CA
Posts: 25,092
Mentioned: 114 Post(s)
Tagged: 0 Thread(s)
Quoted: 5146 Post(s)
Liked: 3624
Quote:
Originally Posted by imagic View Post
...Democrats only gaining two seats in the House of Representatives.
Dish was off with that number as well, since the gain was triple what they predicted. Like you said, so much for predictive models.
imagic likes this.

Sanjay
sdurani is online now  
post #30 of 37 Old 11-09-2016, 09:11 AM
AVS Forum Special Member
 
skriefal's Avatar
 
Join Date: Jan 2000
Location: Sandy, Utah (SLC)
Posts: 1,094
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 97 Post(s)
Liked: 80
Quote:
Originally Posted by thebland View Post
With his Tax policy promised cuts and repealing Obamacare, Im expecting a lot of extra Home Theater cash in my pocket!!!
Hold off for a bit before you choose to buy that new 4K HDTV! There's not likely to be much (if any) extra cash in our pockets, regardless of which candidate had been elected.

And yes, Dish did get this one wrong. Problems in their statistical models? Failure to fully account for the electoral college versus popular vote? Probably a combination of a lot of things.
skriefal is offline  
Sponsored Links
Advertisement
 
Reply Latest Industry News

Tags
Dish Network , election 2016 , viewers to voters

Thread Tools
Show Printable Version Show Printable Version
Email this Page Email this Page


Forum Jump: 

Posting Rules  
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off