I'll bet you that we don't find a statistically significant deviation of the expected outcome. We'll have to define the parameters of the bet of course, we might use a 95% confidence interval for example.

And we're saying that since there's already 10 threads on this topic - all with ZERO proof or anything - that you should shut it until you bring us something new. Asking the same question as every other goddamn noob on this forum IS NOT HELPING ANYTHING BOLD TEXT LOUD NOISES.

maybe you should read this thread again. i'm also going to post my stats shortly and i'm sure MicroBob will have some new evidence to poke holes in. he'll probably blame it on my ISP or something.

id rather u spend your time analyzing cake's mtts/sngs, but if you come up with a bet, ill consider it.

If you post your HH I'll be happy to analyze your cake data

ok, got them zipped, how do i get them to you?

edit: sent, i hope i did it right. thanks again.

Email them to me - I pmd you my address

OK lol you guyse I think Markusgc just found the problem with Cake poker which explains why that Cake graph is so screwed over so many hands.

The problem is that Cake only records the villians hand when he wins the all-in confrontation. PokerEV only considers hands where both players' cards are known and thus a major screwup with Cake poker.

I've yet to check out the hands from NoLimitLeagues but if they have the same problem then I think we found the source of the anomaly.


EDIT: False alarm. The hands from NNL does show mucked hands at showdown. Ignore my post please.

It does seem that new hand histories do have the mucked cards. I just grabbed the top couple of files on my hard drive and PM'd them to Devilboy to show him the problem he mentioned. When he showed me one of NLL's hands I checked some very new ones and they were in the same format.

But ya see - we examined some data and came to a conclusion. Crazy how that works, huh?

Hi allurit - I don't think you should speak for everyone.
Some people really are saying that Stars is set up for the regulars to lose more often. I was mostly referring to myturn's posts from a different thread though.

I would not blame it on your ISP or something.
I'm surprised some of you think my theories are so far-fetched as to how the graphs could be skewed. It seems pretty basic to me.


tautometer - Calculating my own pot-odds on a hand with 47 remaining unknown cards is obviously how I go about looking at the way to proceed in the hand. Sorry you don't understand what I'm getting at but it's clear you have some serious reading comprehension issues.
You're making up stuff about burn-cards and other things that I hope you understand I don't think is a factor.
The effect of card-removal in the long-run though could conceivably skew results. My inability to explain this clearly is probably holding you back on this so my apologies for that.

I do think that others in this thread understand the subtle differences of the effect of card removal that I am driving at. So if anyone has a better way of explaining it so that this guy can understand it a bit better and hopefully get the hell off my back I would sure appreciate it.
I'm dedicating way more time to this rigged silliness than I normally would otherwise in an effort to explain my long-standing theories of the potential bias inherent in those pokerEV graphs and how dangerous I think it is to take them too seriously and/or at face-value without considering other factors.

nolimit - I will play my hands in the best way I can although I'm not really much of an SNG player. I will not be pushing randomly with A6 or anything weird like that until we get down to enough players or get to high enough blinds where I think it's appropriate. Since I'm only so-so on SNG's that will almost certainly not be the ideal time. But I will be playing the SNG's on there to the best of my ability and I would also be willing to bet with you that the favorite will hold up more often than the underdog.

I think that because of the weirdness of the hand-histories there I will have to write down the all-in situations.

You are talking about just 200 hands right? Not 200 all-in's?
I hope so. Because I really don't feel like playing a zillion hands on Cake just to handwrite by pen 200 all-in situations just to satisfy you.
Frankly, you're kind of lucky that I'm willing to do this because I think your arguments about this are pretty silly. But I'm in a mood to show some flexibility and will do my best to write down the all-in HH's as I go along.

all i want is others to do the same thing i did. pay attn to any all in preflop between exactly two players that isnt two overs vs a pocket pair. if the favorite wins, tally it in the favorite column on a piece of paper, if the underdog wins, tally it in the underdog column. post your results. just do as many as you can without affecting your life too much. even if you can only do 50 at least its a start. im hoping i can convince others to do it too. and i will do some more in the next few days.

like i said earlier this study isnt going to be very accurate, but if the results come back like mine did its obvious something is going on and we can investigate furthur.

i might have missed it, but did you guys see slugger5x's thread (another example of someone running outrageously bad): http://forumserver.twoplustwo.com/sh...ht=2008+so+far

the thing about all in luck is that the samples required to properly analyze results are MASSIVE for certain formats. if you were to analyze hands from formats like cap nl or sngs, it wouldn't be quite so bad, because you'd actually be getting all-in at a pretty high frequency. in 100b buyin nlhe, you're only getting all your chips in every once in a while, so your true sample is actually much smaller since most of the hands have no effect whatsoever.

back in march/april, i was hurting for cash pretty bad due to getting ripped off on an apartment in moscow, having to buy a suprise ticket back to the US, and a few other things that happened all at once. i decided to mass grind nl50 playing a super nitty, super conservative style for some easy money. these were the results:



as you can see, over a 200k+ sample, i ran 32 buyins over expectation according to HEM. part of the reason that i can be that far from normal over a sample of that size is that the proportion of hands i was getting all my chips in was quite small in relation to my total ammount of hands played, due to the way i was playing, but more moreso due to the format of the game. for a game like nlhe, i really don't think it should be that much of an anomoly to still see pretty high deviation even beyond 2 million hands.

I shall go back to the statistic we have been presented with.

At peak hour we have about 150 000 players at Pokerstars. 0.38% of the players = 570 players.

Just out of no where I assume that 570 players run as bad as you are running. You are not alone.
Logic or not? Rigged or not?

I also assume that there are many there who run much worse than that. In poker the brutal bad beat swing can be extremely hard to accept without additional explanation.

I have been messing around a little on Cake for a few weeks now so thought I would give this a go. Just went to log on and for the 1st time it is asking me to download the new version. Coincidence? Now, I am about as far away as you can get from thinking a site would rig their RNG but I thought it was interesting. Yes, I know updates go on periodically but the timing is unfortunate don't you think? anyone got a tin foil hat I can borrow?

As the maker of PokerEV I was asked to comment on this thread. I've only read a bit of the thread, here's a few points to add:

1. To my knowledge there are no flaws in PokerEV's all-in EV calculations for the big sites. It's been around for a long time and I've been very careful with it. Exceptions below:

2. If you short stack, the line will skew and it will look like you're running bad on all-in EV. Consider this scenario:

Preflop: You go all-in, two remaining players call and see the flop
Flop: player 1 bets, player 2 folds, you see a showdown with player 1

The only cards available for analysis are those of player 1 - the folding has introduced a bias. You can turn off multiway situations by selecting "show option to ignore multiway all-ins" on the settings tab, and then clicking the box that becomes available on the luck graph. Several run-bad short stacks have found that this removes all of their bad luck. It has basically no effect for people who play full stack most of the time, even over large samples. It will change the line since it limits it to all-ins, but it doesn't appear to affect how good or bad people run in aggregate.

3. If you're running Cake hands through a converter and combining them with PokerStars hands, your EV line will go crazy. To my knowledge, mucked hands are not shown on Cake or not shown reliably, and so a huge bias develops. I wouldn't use Cake hands in PokerEV at all, they probably have all kinds of inaccuracies.

4. Given the number of users of BBV and 2+2 (in the thousands), the lack of majorly outlandish probability graphs (the bell curve) makes me think that rigging is either small or unlikely. There's only been one > 4SD that I know of, and there should be several at least given BBV's regular traffic, PokerEV usage and people's propensity to post bad luck.

5. I think the only way to truly test what's going on is to take a *random* sample of 20 or so trusted regulars from one of the strategy forums (not volunteers), and make them post a graph of their all-in luck for a particular date range. Perhaps the graphs can be given to a third person for anonymity. If the samples are large enough (50+K hands), we should be able to see anomalies or lack thereof.

OK, I've just read MicroBob's objections about the bunching effect as a theoretical flaw in all-in EV calculations. I actually posted about this a long time ago. His objections are theoretically valid.

Like Mason, I don't think it's significant, but the only way to answer this question definitively is to run a simulation with known folded cards and compare to the unknown card calculations. I don't have the time to do this right now but I do have the basic code for it and it won't take terribly long, so I'll do it in the next 3-4 weeks if no one else takes it on. I'll release the results and code open source so that others can look at it.

I haven't seen a lot of huge sample PokerEV graphs. There's fgator's EV graph of 1.5 million hands, which had expected and actual off by something like 5% over a large sample, which is statistically very normal. He's very weak tight and the theoretical bunching effect should be maximized. I've also seen people run hot over hundreds of thousands of hands. So the bunching effect should be less than the order of swings. That's not to say it couldn't be significant - it could - but I don't think the graphs where people run very bad are caused by this. Perhaps contributed to, depending on how large it is, but not the major contributor.

this has exciting implications, we all strive for legit sites and let's put them to the test

Cool thank you Phil!

This is no longer the case, FYI. Cake updated the software so that all hands that go to showdown are shown in the HH. I believe the update came about 1 1/2 to 2 months ago.

-red

MicroBob's effect is obviously real.

A classic equity calculation when two hands are known and no other information, assuming 48 totally unknown cards is correct. This is true as to a heads up match or when every other player is acting totally at random.

However, in a nine handed game with non-random opponents where it folds to the blinds who then get all in, we have other information in that we can put every folding player on a range of hands same as we put those that don't fold on a range. From that we have information about the fourteen cards in the muck as opposed to the cards left in the stub.

This makes the classic assumption that the dispersion of the cards in the stub matches the dispersion of all cards unseen incorrect.

So, at the least, any claim that an RNG is off because it does not match predictions based on this assumption is quite flawed.

The game security specialist from Stars sent me like a 3 page email. I've only played about 7k hands in that time so the results don't mean much. I've asked him to analyze the previous year and if they do so I'll post the results here in the thread. I can also forward on anything to him for analysis or make requests for information if anybody wants to pm me anything.

lol if that truly does skew the results, that forever needs to be referred to as The MicroBob Effect imo

The burden of proof for the fairness of the shuffle is on the sites. This is my opinion as well as that of other reasonable posters.

It is curious a majority of active posters in these threads (though not necessarily a majority of readers/poker players) appear supremely confident of the randomness of the deal...that when presented with evidence to the contrary, they're quick to dismiss it as ``you must've done wrong then''!.

The problem of course is there do exist paranoid posters on this topic. They open themselves wide open to ridicule, while cutting away at the credibility of others with worthy observations.

Anyway, I have the following proposal for sites' faithfuls: if you lay me 20 to 1 at 1k (since you're so confident), then I can have a greatly respected/famous academic authority (there should be no contest as to the weight of his opinions in this area) judge the reliability of analyses provided by you and me of the SAME datasets of randomly chosen all-in hands from various sites.

The dataset (from each site) should be large enough (in the authority's opinion) to allow for significant conclusions to be drawn for the particular site. The hands may include ones from playmoney tables (these are completey random, right?). The (all-in) hands should be only from small stakes NL heads-up tables. The sites should be restricted to the top 10 in traffic. The money should be escrowed.

As usual, I'll be traveling this month, but if interested, PM and we can iron out details. You may end up losing 20k, but it'll make for a great thread!

From what I recall is that Mason was speaking of the effect on strategy within a hand. This isn't so much a problem of math as much as that all the human factors in a hand, estimating ranges based on opponent behavior and so on, present a margin of error that will dwarf any difference between a pot equity calculation based on the assumption that the stub is identical in randomness to all unseen cards and one that takes opponent behavior into account.

Who cares if, assuming your ranges are correct, your equity is 50% rather than 49%, as your estimation of opponent behavior is generally not going to be so precise. This is a totally different kettle of fish than using the same calculations to make detailed criticisms of an RNG over a massive amount of trials. The larger source of error that makes the smaller error insignificant is gone.

This also seems to doom any totally precise calculation of expected ev when taking any all in hands in a game with more than two players. To do this one has to make assumptions about what an opponent will or will not fold, and there simply is no way to do this by a universal method. It is still an estimation with flaws as people are different.

This would be all fine and dandy if we assume that every player gets in the same situation the same amount of times. However, the whole point of playing poker is to not let this happen to you.

I imagine some sort of feature that tries to model typical behavior, where the end user can put folding ranges for typical opponents in typical situations will make these sorts of calculations more accurate, but it still is going to suffer from the same problem once the number of trials gets large.