Whether it "evens out" or not I suspect is player dependent. Maybe not as much at high stakes where competition is more equal, but I am sure (haven't looked at the data) that in my small stakes career, I am on the pair side of pair vs Ax more often.

Ax vs pair is one example that is easy to see/explain.

I don't know.

I believe there is a deviation from the ev your program calculates from the true ev of a particular player, which is player dependent.

If you're expecting something to converge to zero, but actually converges to a nonzero slope line, even ever so slight a slope; the more data you put in, the more bizarre the results look.

I certainly don't know. If I'm picking one of your choices to explain yours' (and others') seeming misfortune; it's that your program is not as "independent" as you think it is.

Very interesting thread.

I am not an expert, but this seems OK to me. This means that on average your are down about 5.3BI over 1000 games. I remember that from my last 1k games I was unlucky according to Juk something about 15BI. I assume that 5BI average is possible, though not likely, over such a large sample. It seems that you are really on the wrong end of variance.

Great post. Nothing to add.

I think hood nailed it. Alongside all-in luck, you have to consider he-got-dealt-a-calling-hand luck, I-am-coolered luck, getting all your good cards when you have a huge stack rather than with a small one, etc etc.

I feel luckiest when I have 7BB at t200, someone shoves and I have AA. I feel unluckiest when I don't play a hand all STT until the AT I shove with 4BB and get snapcalled by AJ. Of course the AA gets cracked sometimes, and the AT wins sometimes, but luck is still more than whether you can or can't do better than your hot and cold equity.

I seem to remember seeing a thread similar to this somewhere else where it was found that many regulars seemed to be having a very 'unlucky' run, and there was discussion about sites being 'rigged' against the regulars. My first thought was: the fact that you are a regular means you are more likely to play through your downswings, becuase you understand that varience happens and you just need to keep making the correct play. But the all in 'luck' over an entire site sums to zero, right? If you get unlucky for 50$ of equity someone else gets lucky for 50$ of equity. How many players play for fun, run hot (hundreds or thousands above expected equity) and then quit once they hit a 5 BI downswing?
It seems to me that if there are a significant number of players doing this, then the players who stick around would be expected to run slightly below equity because of the players who 'quit while they were ahead'. Or am I just making no sense? I can't tell. Just my 2c.

I'm glad this thread is civil, if you posted this on the internet gambling forum you'd have a bunch of trolls insulting you.

Does someone have a link to the other threads where high volume players complained about being unlucky over a huge hand sample.

Variance has no memory. If some recreational player who doesnt understand variance starts off on a huge heater, yeah chances are that he will keep playing, and if he runs bad, he might stop. But it isn't like has has to "pay back" for his heater, and that, because he didn't, someone else needs to pay for it. Someone (actually, of course it was a load of people, but just to drive the point home, lets personify this group) already did pay for that, back when he was on that heater. Who that unlucky guy was depends one which players where playing back then. But, relatively, it might as well be another fish as it might be a regular. Or at least, that's what we hope (with poker not being rigged).

Maybe it was a fish who stopped shortly afterwards partly because of this. Maybe its a fish that didn't notice because he was running good other wise. Maybe it was a fish who never actually got a good start to the game. It doesn't matter, it isn't like "fish use up the luck", because everyone has "the same chance of getting lucky", and that chance "resets" itself every time a card comes of.

Also, I found this interesting:

I hope that I'm interpreting this right (if not I'm sure i'll be corrected). But basically you seem to be saying that the odds of a freak occurrence happening do not matter because there is a large sample size, so anything that is in the realm of possibility might happen. Or; A non-zero probability says that something might happen, this has a non-zero probability, so it might happen, and now it did, so what?

That is of course true. But OP suggested some reasons, other than pure chance and his sample being some freak outlier, for this to have happened. Most importantly; "It's rigged!" (which of course most people think about when running bad, but most people (including OP) think is unlikely). So then the question becomes the following. When do we say, "this is so goddamn unlikely, it must be rigged!"?

Since the chance of "online poker being rigged" is obviously non-zero itself, there must be a such a point. And its obviously somewhere before OP plays 75000 games and loses every single all-in he plays (included retardedly extreme case in post? Check!). Where exactly is this point? I'm really wondering about this, although I can imagine no one can answer this with any kind of certainty. But still, from this point of view, the odds matter. Because with some odds we think, "someone has to be the unluckiest bastard out there" (LOL @ that btw ), while after the odds get slim enough we start thinking "it's rigged!!!" (and rightfully so).

BTW really liked Hoods post also.

Seems hard to believe, I hope there's something simple you're figuring wrong.

Read the first superuser thread about AP. The statistical anomaly of the superuser was so out of bounds (winning 3 consecutive powerball lotteries was the sound byte, I don't remember the details now) that lead some to conclude that some form of rigging was more likely that this occurring in a normal deal. In the original "investigation", there is some excellent statistics and discussion as it all was uncovered.

so after 75000 flips you've lost 'x' flips more than you should've (not sure what x is)?

If its like 300 is it so shocking that results of flipping a coin 75000 times are 37500+-300 or so?

Almost seems standard to me.

Note: the above post was made by a moran, my apologies to those who my stupidity has offended.

"I think hood nailed it. Alongside all-in luck, you have to consider he-got-dealt-a-calling-hand luck, I-am-coolered luck, getting all your good cards when you have a huge stack rather than with a small one, etc etc."

Yeah this line of discussion leads me to think he could be running even worse than he believes, and that the luck factor is bigger than these programs can/do calculate.

Don't forget that 99,999999% of the posters complaining about bad luck are Small Stakes fishes that played less than 300 games. Also, he's not saying things that are obvious BS.

He's a winning player with a good sample and a question, not yet answered.

If someone has doubts about a poker room fairness, they have to start with a sample and after that, try to show if the hypothesis can be denied or not.

I think PS should allow data mining to re-assure us about the game fairness. I understand their concern about data mining, but allowing people to test the randomness of the system is much more important then these concerns.

Yeah, but shouldn't each of these even out in a long run?

Still I doubt that these programs are 100% accurate. Say we have a hand like this:

BB: t1245 with 10 10
UTG: t1695 with K 6
MP: t1745 with A 2
CO: t1350 with A 6
BTN: t6155 with K 7
Hero (SB): t1310

Pre Flop: (t180) Hero is SB with A K
4 folds, Hero raises to t1310 all in, BB calls

Your program counts this as a coin flip right?

Again, I'm not saying that this totally explains your results, but it must have some effect over 80k hands.

Can we project how unlucky the 'unluckiest player in the world' would be? In other words, there will be one player (or a group of players) that will experience worse variance, either upwards or downwards, than any other player(s). So,... :?

Have you done an analysis on which hands were winning?

I don't have any where near as many hands as you but my stats show that overcards are favored over pairs in Stars. I think it may be because there is a bias toward high cards on the board.