Quote:
Originally Posted by smithcommajohn
Saying AMEC's results are unlucky but normal seems a bit understated. I think running worse than roughly 95% of all players is running HORRIBLY, not simply unlucky. I mean, at what point can we call his sample abnormal? The point is, we never can. For all his hands, the results are just one blip on a larger curve where we can simply say, "someone has to run that bad". So even if he was -5 SD, he's still normal, just more unlucky.
I think the point AMEC should be taking from this is, he was absolutely justified in thinking he was running like crap, because he was. His one sample (even if it had 10 billion hands) can never prove anything except for how lucky or unlucky he was.
Am I right or wrong here, spadebidder?
I don't agree that this is very noteworthy at all. Your example of -5SD would be extremely abnormal, and would mean that only 1 person in 3.3 million (with this sample size) ran worse. There aren't 3.3 million players who have over 370K hands played on Stars, and it would be a freak thing worthy of scrutiny. But at 1 in 18, there are many regular players with the same minor bad luck.
A normal curve doesn't have to be unconstrained for the tails to go out forever, they don't. There are only so many poker players.
I might be one of those bottom 5.5 percenters in all-in luck, and I'd never know because this level of offset isn't noticeable. You wouldn't know either unless you check. You are saying that over the course of 371,000 hands, over 3 years and 2000 hours of play, that you would know if you lost an extra 77 all-in hands more than expectation. That's 2 extra lost hands a month. No, that would not be perceived. By any human.
The bottom line is that this had very little effect on his actual winrate. He has been over and under all-in expectation over the course of this hand history, and it just looks like noise against the win line. His perception was not justified or accurate. I won't post his other information here without permission, but it shows this clearly.
Edit: on your note about not proving anything with even 10 billion hands. Checking someone's all-in luck is only a small part of checking the randomness of the deal. That wasn't the objective here. Incidentally I also looked at his hole card distributions and flop distributions, and they were all normal. And there are many other things you can check.
Last edited by spadebidder; 03-12-2010 at 01:24 AM.