Originally Posted by isa
That summary is stunningly at odds with the scientific method.
So is this:
It doesn't mean we don't post it.
Let's first just address first semester freshman engineering material (which I exaggerate slightly: l learned this in high school): no single test proves a hypothesis.
Good thing then that we did far more than one test. Once more:
1. There were two sets of tests. One created by Mark/Scott, and another by Arny. Each created with completely different methodology, tools and personnel.
2. Mark/Scott's test consisted of three clips converted down to 16/44. Arny used one clip converted to 32 and 44. So that is a total of 5 different tests.
3. To get the probability of chance to zero, I ran 20 trials of Arny's tests.
4. To get the probably of chance to zero, I ran 53 trials of Mark/Scott tests.
5. So the number of "tests" I conducted was 73, not one.
6. Multiple forum members here and on WBF have conducted their own tests on Arny's files and likewise. I let you count those but the number of trials is well above 100.
7. The testing was done using the proposed tool: Foobar2000 ABX. This is a plug-in that randomly assigns X and Y to A and B. All the reporting was automatically generated by Foobar2000. I have offered people to come and witness me taking the test but hopefully we are not calling each other's integrity into question in the name of "science."
In summary, this is not remotely as you characterize. It was not me guessing right once. Or twice. Or ten times. It was not just one clip but four. It was not just me but a number of others.
At best one disproves a null hypothesis, and only with statistically significant results among other key criteria established over the 400 plus year development of the scientific method.
And statistical significance was reached and then some. In my case, probability of guessing was 0%. Not .5%, not 5%, but 0%. I can tell the files from each other 1000 times in a row if you like.
Yet you continue to quibble, deny, fight, ignore, cherrypick and propose stawmans to refute the very foundations of the scientific method. Noted.
Noted what? Your post is information-free. It adds no data to the conversation. It simply aims to add doubt to the results. "Oh, scientific method was not used." You mean other than running double blind tests, using computer generated samples and reports?
What you are practicing is called FUD. IBM used it decades back to dissuade people from buying cheaper Amdahl computers. "Oh, it is not an IBM and it may have problems."
It is also the way too often practice of turning lack of data into data. "Oh, you have only run one test." Let's say I had. If you have no test data of your own, then you have no argument. Period. You can't imply that lack of additional data takes away the data that does exist.
If you are genuinely interested in science and adding data to the conversation, then run the tests. Don't just randomly guess. Remember that others have passed some of those tests. So try hard to identify the differences. Report that. Experience that. Then your posts will be constructive. As it is, they are just debating tactics and noise.