[ ] 100k hands

By the way, WCGRider posted his 25NL prop bet hand history, has nearly 300k hands. Here's that thread:

http://forumserver.twoplustwo.com/78...op-bet-222539/

You wanted hand histories, there you go.

I have some stuff to attend today but I'll see if I can do it tonight.

The odds that you'll get a paired (or better) flop:

3/51 (The odds the second card matches the first) PLUS
3/51 * 2/50 (The odds the third card matches the first and second) PLUS
48/51 * 3/50 (The odds the third card matches only the first) PLUS
3/50 (The odds the third card matches the second)

= 0.177647058823529411764705882352

[Someone please check my math here, I'm not sure if I got those probabilities right - specifically I don't know if the second term is a subset of the first term. I think I did it right, but I'm not completely sure. It's pretty close nonetheless since that term is only around 0.25%]

I edited my tester program (again, posted for completeness):

And the results:

The actual occurrence of paired flops is 17.40% in my sample versus the expected 17.76%.

Figuring out how preflop holdings affects paired-edness of flops is tricky. On one hand, more people playing broadway type cards would indicate more changes lower cards flop and pair themselves. Additionally, people playing more pairs takes a disproportionately larger amount of one rank out of the distribution, indicating slightly higher paired boards. There may be reasons that the occurrence is lower, but I can't think of any immediately. Regardless, the conclusion holds as the observed value is very close to the observed value (3/10ths of a percent over 200k hands).

Conclusion: PokerStars has not yet been proven to rig flops. This is not to say that they aren't rigging it in other ways or can't or won't in the future, but the evidence given has not shown otherwise.

I just ran my programs on Riders hand histories:

Similar results.

Nice work Dax, hopefully we can direct link to that analysis *when* the tinfoil hat crew decides to make another "Rigged" thread. I doubt they will reenter itt, those types generally dont like to be in the presence of such logical reasoning. They honestly just want to bitch just for the sake of bitching.

please.analyze.samples.of.>5mil.before.coming.to.a ny.conclusions.

also.no.spacebar.atm.=(

hold alt + press 3 then 2 on num pad -> release alt to make spaces.

or

or go to Start -> Programs -> Accessories -> Accessibility -> On-Screen Keyboard

s p a c e s!

I've been wrong before, and it's always possible I'm wrong now

BUT, ask yourself this, why do suited holdings bias the color of the board? One way of framing the answer is because the suits of the remaining deck cards are no longer uniformly distributed. By contrast, paired holdings are not expected to affect the distribution of the remaining deck suits.

Consider the following extreme but helpful scenarios. 12 players, half of whom are dealt suited hearts, the other half are dealt suited spades. Then, the board is much more likely to come clubs and diamonds only. Consequently, a suited flop is more probable.

On the other hand, 12 players, 2 of whom are dealt pairs of Aces, the next two are dealt pairs of deuces, the next two pairs of treis, and so on and so forth all the way up to pairs of sixes. Since the colors of the remaining cards in the deck are still uniformly distributed among hearts, spades, clubs, and diamonds, then the flop is no more/less likely to be suited. (however, it's more likely to be paired!)

The idea is this: biases in the colors/ranks/...etc (pick your favorite attribute X) of the flop texture will be observed if caused by an event that both skews the distribution of attribute X in the remaining deck cards and induces players to see flops. Do you get the idea? (no shame in saying no, can explain further)

idiotic

Can we please stop using this ridiculous logic that a market leader would never cheat nor have any incentive too? This has been proven false repeatedly. Look at our current financial situation...huge profitable companies doing super shady things to increase revenue. Martha Stewart took a huge risk for literally pennies. According to you guys she couldn't be guilty of insider trading because why would she risk her reputation on such a small amount of money? Furthermore, stars and FT both know that the window is closing on them. When Obama wins and online poker is regulated in the US, everyone will flock to a regulated potentially more legit site, ie. MGM, Harrah's. You can be sure they are trying to maximize their revenue right now because in a couple years it will be gone...

what?

while i agree with some of what you're saying (the part about businesses doing shady things even though the rvr is bad), no one has ever really presented any sort of logical model for how action rigging increases revenues, and in my opinion, it would actually do the opposite.

just to use the simplest model possible, lets say that you opened your own site and you could rig it any way you wanted to without any fear of anyone noticing. now lets say me and microbob are the only ones that join the site, both depositing for $1000. we both sit down at an nlhe table where the max buyin is $1000 and flip for rolls on the first hand. while you may have gotten the maximum per hand rake for that particular hand, the overall result would be a relative disaster for you. even if the loser redeposited, you would make much more money if we hadn't taken the maximum action route and instead taken a route that lead to you raking a much higher volume of smaller pots.

the same kind of logical gaps are present in any sort of model that involves fish survival rigging. not only is there no evidence, what's more important in my opinion, is that the model doesn't make any god damn sense in the first place. if someone ever constructs a model that makes sense, then maybe i'd take it more seriously. until then, its just the same worn out joke its always been.

If a site wanted to earn the most from its players, it would award pots in a biased fashion to those with the lowest bankroll. This would keep the money in play longer, allowing it to be raked again and again and again...

Of course, this would mean that shortstacking with your whole roll would be amazingly profitable and people like Imsa and LOLCATS would consistently destroy real poker players.

As you say above, a player going broke is a disaster. Fish go broke, regulars usually don't. I could list a million ways to increase rake without talking about maximum action on any given hand.

some random idiot that plays 3 or 4 hours a week going broke isn't a disaster. their main value is that they attract other people to the site that put in more volume. random guy is also more likely to withdraw if he makes a big score, whereas a high volume player is more likely to move up in limits generating more rake.

also, its very common practice for high volume players to use some sort of stop loss during their sessions. if a player hits that number and cuts short their session 3000 hands earlier than they would otherwise, more rake revenue is lost for the site than whatever the recipient of the good fortune will generate.

What cards people choose to play, or not play, has no direct effect whatsoever on the flop.

Rather, the cards that people choose to play, or not, only effect the likelihood that a flop is seen.

Only in this indirect way is the flop data effected by what cards people like to play. It's impossible to accurately measure this effect- we can only make assumptions, but I think it's safe to assume that any effect is negligible anyway(?)

That model wouldnt work because ur only including 2 players. I can think of a model that would but i dont care to post it.

I'm too lazy to go through 23 pages of posts. Did OP ever put up his hand histories?

The colors are uniformly distributed but there are simply less cards to chose from. Example: The chance of getting a monochrome board is higher if we play with 13 of each suit than if we play with 3 of each suit and much, much higher than if we play with 2 of each suit (=0 ldo). So, in the example you mention there are less of each suit remaining(!) which decreases the chance of a monochrome board.

Daxonovitch,

thanks a lot for running it. There seems to be a small mistake in your math tho. I'm not 100% sure, it's not like I'm always correct in this stuff despite having a profound statistical education, this stuff is simply confusing. I'm gonna redo it. Like you (I guess), I'll refer to a paired board as one with at least two of a kind.
P(paired board)=P(excactly one pair on board)+P(trips on board)
=3/51*48/50 (third card can't match the first two)*3(the cases that cards 1&2,1&3,2&3 match are equally likely) + 3/51*2/50 (trips)
=17.176%.

So, your observed frequency is higher than the expected one, which is what I expected, weeee!

1. The first part is just another way to look at things. Without prior knowledge of peoples cards, the fact that cards are dealt doesn't make one flop more likely than another. That is correct. Many flops remain 'hidden' (unobservable) and the fact that people like to play good hands (which are more often paired or suited) biases the flops that are observable.

2. This effect should definitely not be ignored. Over a very big sample it'll be observable cause it's simply there. In a very nitty sense that also means that real hands are simply not suited (lol) to test the RNGs randomness. If you and me each played 10 trillion poker hands, it should be very likely that we observe paired boards with a significantly different frequency simply because our playing styles differ. Better data is available from prearranged flips since folding pre is simply forbidden there.

I don't believe you understand the situation. What activity specifically was shady as it related to the current financial crises?

Actually she took almost no risk. She went to jail because she lied and thus obstructed an investigation not because she choose to take advantage of inside information. Further, she didn't seek out this information it sought her out and she made a decision. That is very different than conspiring to commit a crime. A better example would be professional athletes who then get involved in drug trafficking, usury, and dog fighting.

Regardless the main difference is that with online poker there is no way to systematically cheat and not be detected by anyone with a year of CompSci and one or two years of stats. It is impossible. So any analysis of what a site does has to be premised on the fact that if they cheat they will get caught which is why it is never in their best interest to cheat until the business is close to failing.


I don't see this happening. No matter who gets elected online gambling is not an issue that anyone will address. Further, assuming your theory is correct the most likely scenario is for a market leader to be bought out by a B&M casino rather than them attempting to enter a market where critical mass is a requirement for profitability.

As was mentioned before, if monotone flops are going to occur TWICE as often as expected, that number would converge very quickly. If you want to say that they occur 2% more often than expected, yes, you'd need something like 5mil.

we are not talking about grammer here, we are talking about online poker. so if you are wanting an english grammer lesson go some were else. i think the donkey here in this case is the one that is talking about something that is not even on the same topic.