Not to be trolly, but white noise would almost certainly be the worst.
http://www.ncbi.nlm.nih.gov/pubmed/17435360

In reality, testing a music hypothesis on poker performance would be almost trivially easy.

Yea, I am not convinced white noise would be best. I said someone suggested it. That's why I didn't write it in the post.

That wouldn't be a good study. No one thinks that 100 hands is even close to a good sample of poker hands. It wouldn't be representative of a normal session, nor of a normal situation in which people play.

Ok man whatever. You can focus in on the minutia and pick it apart over details or you can see where the general idea is going.

This is turning into a bit of a derail, but it would be (again) trivially easy to set up an experiment that had a reasonable sample size (hint: one would NOT need the same number of hands to determine an individual's "true" win rate) and could test the effect of music/lack of music/type if music/whatever you want your input to be vs. win rate/fMRI patterns/whatever you want your output variable to be.

Scientific studies are about minutia and details.

To truly do a scientific study it would take a lot of time and money. You would need to get 10-20 players to play 50k hands with one type of music and then have them transition into another type of music and then play another 50k hands. So in reality it would be easy, but actually doing it would be much harder.

Ever read a Nature paper?

Fine - you're right - you win - clearly that matters to you.

I'm sure as the research teams discuss what would be a viable test they immediately dismiss an entire framework suggestion based on the fact that a fine point of the framework could use tuning.

/out

lol touche.

This is wildly inaccurate, sorry. If I cared enough, I could bust out a power analysis, but you would need AT MOST 200ish hands, given that you had players with similar win rates. You would likely need 50ish players, but that's just a guess. You don't care whether they're necessarily winning players, and you're not trying to establish a "win rate." Neither would you necessarily want the same players subjected to both conditions (although you could make an argument for it).

Really though, it's not at all anywhere close to difficult.

Why would you not want to know a true win rate? it seems like you would want to know what someone is winning (or losing) in one condition and then what it is when you change that. (You have to give them enough time to transition to get used to a different background. I think if you wanted to get this passed through a peer-review in the psychology field it would be rather difficult.

No, it's pretty easy to account for individual subject variation. If it weren't, there would be no such thing as educational research. Win rate is just another covariate to account for.

what is a poast?

Hats off to those PT4 guys.

Someone identifies the need for a new filter, and they're on it, pronto:

That is amazing. Wow, utterly amazing.

I don't see how with that kind of experiment we would be able to tell an effect on winrate. What am I missing?

I have no idea what you're confused about, but essentially you're looking at changes in win rate as a function of music type. I'm not gonna set the experiment up for you, but there are multiple ways of doing it.

Here's something to get you started.

http://www.csulb.edu/~msaintg/ppa696/696exper.htm

What that study shows is repeated prolonged exposure to loud white noise is more stressful than doing simple math problems without the white noise. It does not determine whether the stress is due to the nature of the noise or the volume of the noise. Neither dos it determine whether white noise is more or less stressful than structured noise, specifically music. The white noise used in the test was at 80dB, which is as loud as a telephone dial tone. I know I'd be stressed to have 80dB pumped at me constantly, but nobody said white noise would have to be a loud as 80dB to be useful for a poker player's mood/concentration. Finally, adding white noise to an otherwise soundless environment is unlikely to add anything of value for poker, but it may be less distracting than many typical aural environments. So I don't think that your are drawing the correct conclusion from this study.

lol, I'm not about to read the entire article and critique the design. The point is that there's zero a priori reason to expect white noise to have any beneficial effect on any cognitive task (e.g. poker), and finding evidence to support negative effects of white noise is trivial.

But feel free to do a review paper for us suggesting that white noise is good.

Yes, and in order to be able to tell a small difference in winrate you are going to need a large sample. 200 hands is not a sample of any significance. We are talking about such small difference that we need a large sample and need to be able to make sure nothing else is the cause for the change in winrate.

In a "controlled" experiment, one "controls" for external variables to a reasonable degree, and statistically accounts for reasonable uncontrollable external variables. You don't need a statistically significant change in win rate for every individual. You need consistent patterns in win-rate variation. This is why we do statistical analysis.

Which takes thousands of hands to prove.

Nice OP.

All the music stuff reminds me of the music/driving studies.