statistical tests to show that online poker is not rigged
 

I have been having this argument with some people on PStars who say that it is rigged. Now, I was just wondering if anyone knew of any high profile tests that were run to show that the distribution of the cards is statistically sound.

Any help would be amazing. Thanks!

Re: statistical tests to show that online poker is not rigged
 

Independent audit reports.

Re: statistical tests to show that online poker is not rigged
 

Could you send me a link to results that have been published online? There must be something like this available to the public.

Re: statistical tests to show that online poker is not rigged
 

these independant tests dont test if its "rigged." they only test to see if the outcomes generated by the random number generator are distributed appropriately. so in theory, pokerstars' random number generator could pass test, while they could still rig the cards on particular hands. for the record, i dont think this is happening.

Re: statistical tests to show that online poker is not rigged
 

[ QUOTE ]
these independant tests dont test if its "rigged." they only test to see if the outcomes generated by the random number generator are distributed appropriately. so in theory, pokerstars' random number generator could pass test, while they could still rig the cards on particular hands. for the record, i dont think this is happening.

[/ QUOTE ]

I actually have thoughts on this ... but I probalby wouldn't be able to sound very intelligent....and I definitely wouldn't be able to explain/defend any of my points...so I'll just keep quiet....

Though - NO - I don't think any site is 'rigged.'

A question of frequency of occurence of rare anomolies MIGHT be in question. (I said might - huge might - don't flame me)

Re: statistical tests to show that online poker is not rigged
 

[ QUOTE ]
these independant tests dont test if its "rigged." they only test to see if the outcomes generated by the random number generator are distributed appropriately. so in theory, pokerstars' random number generator could pass test, while they could still rig the cards on particular hands. for the record, i dont think this is happening.

[/ QUOTE ]

I do think that it is happenning. The number generator is random, but as you just said they could rig cards on particular hands. For instance, giving someone 4 of a kind and someone else a straight flush. They know that both parties will bet heavy and GUESS WHAT? that means that the bigger the bet, the more rake they take. It makes perfect sense. However, it is random in the sense that you have an equal chance of coming out the winner or the loser in such a situation, but it is still technically "rigged."

Re: statistical tests to show that online poker is not rigged
 

[ QUOTE ]
[ QUOTE ]
these independant tests dont test if its "rigged." they only test to see if the outcomes generated by the random number generator are distributed appropriately. so in theory, pokerstars' random number generator could pass test, while they could still rig the cards on particular hands. for the record, i dont think this is happening.

[/ QUOTE ]

I do think that it is happenning. The number generator is random, but as you just said they could rig cards on particular hands. For instance, giving someone 4 of a kind and someone else a straight flush. They know that both parties will bet heavy and GUESS WHAT? that means that the bigger the bet, the more rake they take. It makes perfect sense. However, it is random in the sense that you have an equal chance of coming out the winner or the loser in such a situation, but it is still technically "rigged."

[/ QUOTE ]

There is a max rake on every hand. For example, if you look at a 30/60 limit game on Stars, the rake is the same whether the pot is $1000 or $135. With all the money these sites make every day, I find it hard to believe they will go out of their way every once in a while to rig a hand in order to get 50 cents more rake.

Re: statistical tests to show that online poker is not rigged
 

The rake is the same at the upper level games I'll agree, but the majority of people playing at pokerstars aren't playing at those limits. They are playing at the .05/.10, or the .50/1.00, the average Joe I'm talking about. And at these low levels there is a pretty substantial difference in the rake depending on how much is in the pot. I'm not saying it is definately happenning, but if you look at some of the horrific online beats that we've all seen, and then consider this, I'm saying it logically makes sense.

Re: statistical tests to show that online poker is not rigged
 

Oh my goodness, I have created the very beast that I sought out to destroy!

Look, I don't want to argue about whether or not online poker is rigged. For the record, it's pretty clear to me that it isn't.

I just wanted to know if there are any results published online that statistically prove that the card distributions are fair.

Please start another thread if you want to debate about whether or not online poker is rigged.

Re: statistical tests to show that online poker is not rigged
 

[ QUOTE ]
but if you look at some of the horrific online beats that we've all seen, and then consider this, I'm saying it logically makes sense.



[/ QUOTE ]

You're right. I've played hundreds of thousands of hands, and I've seen literally dozens of those "horrific" beats that you speak of, 989:1 (the worst beat possible in holdem) shots hitting. That should NEVER happen. What are the odds of a 989:1 shot hitting!?! It's like a billion to one!

Re: statistical tests to show that online poker is not rigged
 

You could try the "Integrity" page -- they don't have literal published tests, but it certainly looks promising when you use hardware RNGs (not the pseudorandom, "seeded" software RNG that most home microcomputers use), which are officially verified by two security firms -- including one which hacked an online poker site in the past!

(Incidentally, I wish they did something kind of like PokerRoom's EV chart at http://www.pokerroom.com/games/evsta...php?order=card )

Seriously, the "OMG STARZ IZ RIGGED!!!1" people need to quit whining. No one seems to realize you hit an 80% shot, oh, say, 80% of the time. (which, honestly, is why I usually play limit poker)

Re: statistical tests to show that online poker is not rigged
 

What are the odds of a 989:1 shot hitting!?! It's like a billion to one!

The odds are 989:1.

Re: statistical tests to show that online poker is not rigged
 

This is a great 1st post.

Re: statistical tests to show that online poker is not rigged
 

[ QUOTE ]
The rake is the same at the upper level games I'll agree, but the majority of people playing at pokerstars aren't playing at those limits. They are playing at the .05/.10, or the .50/1.00, the average Joe I'm talking about. And at these low levels there is a pretty substantial difference in the rake depending on how much is in the pot. I'm not saying it is definately happenning, but if you look at some of the horrific online beats that we've all seen, and then consider this, I'm saying it logically makes sense.

[/ QUOTE ]

Actually, logically it doesn't make sense.
Poker sites want as many hands to be dealt as possible, more hands = more rake. Also, they don't want their players to go broke. What happens in huge pots with 4 of a kind losing to straight flushes? The hands take forever to play out, and a person loses a huge chunk of their money. Sure, the site may get max rake, but at a high cost.

Logically, if a site were to rig their site for max rake potential, they would ensure that all their pots involve the winner winning with top pair higher kicker. This way the hands takes a short amount of time, nobody loses all of their money, and there's still enough action to get alot of rake.

Re: statistical tests to show that online poker is not rigged
 

There are three distinct issues.

First is whether the hands you get in on-line poker are distributed precisely as a theoretically perfect shuffle. The more transparent sites publish enough details of their methodology to reassure people that it should be almost perfect, and some publish sample outcomes which always conform to statistical expectation.

However, if I were hired as a statistical auditor, I would concentrate on the outputs (the cards dealt) not the inputs (the theoretical perfection of the random number generator). I wouldn't just look at the unconditional distribution of pocket and board cards, I would look at what matters to poker players, the distribution of hand strength; particularly the correlations between different hands at the same table and the same player in different hands. There are a lot of steps between random number generation and hand output that could have bugs. To my knowledge, no one has ever done anything like this. I'd also look hard to see if any player enjoyed superior luck, more than could be explained by random chance. This could be an insider or someone who figured out how to hack the site.

So, on this point I would say there is some published information, but less than a skeptical person would demand.

The next point is whether the sites deliberately skew the distributions, whether to push up the rake or attract players. I don't agree with the "it's only a few pennies in rake" argument. If you push up revenue by 1% while keeping costs constant, you can push up profit by a much larger percentage. Moreover, a manager under pressure to hit some target might well compromise the integrity of the business to make a few dollars, we've certainly seen enough of that over the last few years.

No site publishes anything that would refute this. It wouldn't matter if they did, because if you don't trust them to deal the cards fairly, why would you trust their reported statistics? Even if an independent audit firm did the analysis, it's hard to know what they would look for. You could easily rig things but keep the overall distribution of cards as expected by chance.

So on this issue, I'd say there is no information and not likely to be information. Trusting people will trust the sites, suspicious people won't.

While either of those types of non-randomness would bother me as a player, neither one is unfair. Assuming they're too subtle to notice, they wouldn't affect play much. So they're only theoretical problems, even if they exist.

The third kind of rigging is the really worrying kind. The site could have hired players, or bots, to win people's money by cheating. But here there is no evidence that could be produced. The cards and hands could be perfectly random, but the bot could know what they were. An auditor might discover that some players seemed much better than others, but unless she contacted individuals, she couldn't prove it was anything more than skill differences.

The bottom line is that people who believe the sites are rigged will never see evidence to disprove that. On the other hand, there's never been a shred of evidence to prove that any site is rigged in any of the three ways.

Re: statistical tests to show that online poker is not rigged
 

[ QUOTE ]
[ QUOTE ]
The rake is the same at the upper level games I'll agree, but the majority of people playing at pokerstars aren't playing at those limits. They are playing at the .05/.10, or the .50/1.00, the average Joe I'm talking about. And at these low levels there is a pretty substantial difference in the rake depending on how much is in the pot. I'm not saying it is definately happenning, but if you look at some of the horrific online beats that we've all seen, and then consider this, I'm saying it logically makes sense.

[/ QUOTE ]

Actually, logically it doesn't make sense.
Poker sites want as many hands to be dealt as possible, more hands = more rake. Also, they don't want their players to go broke. What happens in huge pots with 4 of a kind losing to straight flushes? The hands take forever to play out, and a person loses a huge chunk of their money. Sure, the site may get max rake, but at a high cost.

Logically, if a site were to rig their site for max rake potential, they would ensure that all their pots involve the winner winning with top pair higher kicker. This way the hands takes a short amount of time, nobody loses all of their money, and there's still enough action to get alot of rake.

[/ QUOTE ]

why do you think players go broke on lower levels just because they lose one buyin? The player who lost will bring more and the player who won may create even bigger pots and more rake in the future.

Re: statistical tests to show that online poker is not rigged
 

Ok it took me a while to find it, but here is an RGP post from 2002, scroll down a bit to hand analysis, 2nd post from top.

http://www.playwinningpoker.com/rgp/02/

Not even 40K hands and everything seems to be pretty normal.

Re: statistical tests to show that online poker is not rigged
 

Aaron,

Excuse me for tossing a bouquet your way, but I just did. - You've given this thread a lot of insight and things to think about, as usual.

Now I've got a serious (burning?) question for you: Do you play online? I don't, having tried it a few years ago and not caring for it. But I'm thinking of giving it another go. I doubt your answer will affect my decision, I'm just curious.

Nut

Re: statistical tests to show that online poker is not rigged
 

Actually, it does not show that everything is normal. There is a serious deficit of Aces and Kings, more than can be plausibly explained by chance. If this is really a random sample of hands, then the site has serious problems.

The trouble with this post is that it gives no consideration to standard deviation. It shows the actual and expected numbers of various types of hands, and concludes that they look pretty close. But over 37,867 hands, things should be a lot closer than this.

For example, there were 2,264 paired starting cards. 2/13 or 348 of them should have been AA or KK. In fact, only 297 were. That may not seem to be a huge difference, but if everything is random, getting 297 or fewer AA or KK out of 2,264 paired hands would happen only one time in 780.

That's not conclusive proof that the site has random number generation problems, but I wouldn't tout it as evidence everything is working perfectly. Combined with the strong deficit of AKs, it looks funny.

Of course, there are a lot of card combinations cited, and some of them are going to differ from their expected frequency. But AA, KK and AKs are among the most important starting hands. Missing 67 of them, even over 37,867 hands, could make a significant difference. If you are a 2 BB/100 hand average player, you expect to make 757 BB over this many hands. Your AA, KK and AKs might average 5 BB wins eacy, the deficit of 67 means you lose about half your expected profit.

Re: statistical tests to show that online poker is not rigged
 

[ QUOTE ]
What are the odds of a 989:1 shot hitting!?! It's like a billion to one!

The odds are 989:1.

[/ QUOTE ]

Lee Jones article

Re: statistical tests to show that online poker is not rigged
 

[ QUOTE ]
I have been having this argument with some people on PStars who say that it is rigged.

[/ QUOTE ]

And why exactly are these idiots playing a game they believe is rigged?

As others have said here, it is possible that cheating is going on while the overall results appear normal. The logical problem with this whole topic is that it is generally impossible to prove a negative proposition. No amount of evidence will prove conclusively that online poker is not rigged. One example will prove that there is cheating.

Re: statistical tests to show that online poker is not rigged
 

[ QUOTE ]
I'm not saying it is definately happenning, but if you look at some of the horrific online beats that we've all seen, and then consider this, I'm saying it logically makes sense.

[/ QUOTE ]

This doesn't make a bit of sense to anyone who has the least bit of understanding about math and probablilty. See the Lee Jones article I cited a few posts back in this thread. Given enough events, anything that can happen will happen. We notice the rare 1000-1 shot when it happens. We don't notice the 999 times it doesn't happen.

Re: statistical tests to show that online poker is not rigged
 

for the record, like some others, i don't think it's rigged at the top 15 sites (took 15 sites to be the major ones).

but it isn't testable even with a big flow of cards.

they could just a juice a few in-house players cards fairly regularly. you'd definitely need every player's card flow and a huge number of hands for each.

you could probably prove though that the general card flow is pretty much correct i.e. worst case they are just juicing it slightly.

Re: statistical tests to show that online poker is not rigged
 

[ QUOTE ]
[ QUOTE ]
I have been having this argument with some people on PStars who say that it is rigged.

[/ QUOTE ]

And why exactly are these idiots playing a game they believe is rigged?

As others have said here, it is possible that cheating is going on while the overall results appear normal. The logical problem with this whole topic is that it is generally impossible to prove a negative proposition. No amount of evidence will prove conclusively that online poker is not rigged. One example will prove that there is cheating.

[/ QUOTE ]

you are correct and that is a good way to look at it...

but one day we may open the paper and find out it's slightly rigged. and would anyone be shocked?

enron and adelphia turned out to be scams (although they was probably some "evidence" beforehand) and i remember that cryptologic's dealing mechanism didn't work properly a few years ago (that's basically "rigged" - turned out they were the victim, but i think it qualifies for what many suspect).

Re: statistical tests to show that online poker is not rigged
 

[ QUOTE ]
[ QUOTE ]
What are the odds of a 989:1 shot hitting!?! It's like a billion to one!

The odds are 989:1.

[/ QUOTE ]

Lee Jones article

[/ QUOTE ]

How are people really this dumb. For god's sakes, my freaking location is "People Suck at Sarcasm."

Re: statistical tests to show that online poker is not rigged
 

[ QUOTE ]
How are people really this dumb. For god's sakes, my freaking location is "People Suck at Sarcasm."

[/ QUOTE ]

You're a poker player. You should be delighted that people are this dumb. They are where most of your profit comes from.

Re: statistical tests to show that online poker is not rigged
 

[ QUOTE ]
[ QUOTE ]
How are people really this dumb. For god's sakes, my freaking location is "People Suck at Sarcasm."

[/ QUOTE ]

You're a poker player. You should be delighted that people are this dumb. They are where most of your profit comes from.

[/ QUOTE ]

Good Point.

Re: statistical tests to show that online poker is not rigged
 

The following are simple tests that can be run on a large sample size and would more than satisfy anyone sane (some people will never be satisfied, they lose and don't want to accept it).

1. Preflop card distribution

Every hand you get cards and in say 100K hands you have a really good sample and you know how much you are "expected" to have of each type of hand (AA, KQs, J4o, etc.)

I only have 50K or so hands in my PT database but the distribution is near statistically pefect. Obviously there are deviations but nothing whacked out.

2. How often particular draws come in

When you flop a 4 flush and stay in for turn/river how often do you make your flush? When you flop an OESD and stay for turn/river how often do you make your straight? etc.

This can be done with fancy SQL on a PT database but I'm too lazy.

Obviously these are simle checks but would effectively prove a random shuffle.

And since everyone thinks every site is rigged, you'd probably have to repeat this test on every site.

Non random flops
 

I was in Vegas a few months ago and happened to find myself chatting with an older gentleman while waiting in line for the buffet. It turns out he was a [poker] dealer years ago. The conversation turned to tipping and I commented on how I was unable to imagine how the dealers at the smaller rooms were able to make a living. A Mirage/Bellagio dealer can afford to deal 30 minutes of 3-6 and walk away with $4 for his trouble; it all evens out if he gets to push a tourist a few winners at 15-30 at 5+ dollars per. A room where 3-6 is the biggest game will not offer this opportunity.

He broke out in a big smile and said, "you'd be suprised just how often that tourist did well when I dealt".

He proceeded to explain that he was neither capable nor willing of actually choosing who won (well, not willing - I suspect he was indeed quite capable) but that he was always on the look for a good tipper who needed a little assistance.

He used to call them "tourist flops" - a nice euphamism for those 9-6-4/rainbows that often flop big hands or draws for the visitor while missing the local rock's A-K.

He claimed he could produce them at will and a quick demonstration left me to believe that he was, if anything, understating his skills.

Cheating ?

Absolutely !


Non-random ?

Hell yes !




But how would an online site benefit from such a ruse ?

Supplementing the rake seems to be a claim the falls short of plausibility when subjected to scrutiny; not only does this theory fail to account for the fact that once the rake has maxed out there is nothing to be gained from building the pot, it also ignores the negligible (and possibly reversed) effect such behavior would have on the HOURLY drop.

$60 3-6 pots take time to deal out. I don't know whether it takes longer to play one than to play three $20 pots (probably not) but it does take some additional time.

I am not suggesting that the "online poker is rigged" bunch be silenced - perish the thought.

These imbeciles are doing the rest of us a huge favor; the louder they scream the less likely that some rogue site will decide to take the "add to the rake" vehicle out for a test drive - and for this we owe them a debt. (Extra meds and free aluminum foil for all the paranoids . . . put it on my tab).

That said, we need to see that for every "Chicken Little" there are at least three voices of reason. People are signing up to give away their cash at a blistering pace. It would be a shame if the claim that online poker was rigged became accepted as fact by as little as 10% of the population.

*

We now return you to your regularly scheduled programing.

Re: Non random flops
 

I have wondered myself. After reading a couple posts regarding sample size and number of times receiving AA or KK, I checked my PT stats. I am at work, so I dont have exact number.

I had AA 217 times and KK 216 times and QQ 226 times out of 35,000 hands give or take 2000...cant remember exactly. I was impressed that they were as close as they were.

I am gonna be in the same boat as some and always wonder, and just hope that online sites arent rigged. When someone hits that wonder card to fill their flush or 1 outer, I just have to think that its cause its online that they called the bet to begin with. I think that since its online, its not the same as actually putting money in while in a B&M. Therefore producing some of the beats people see.

On a side note, how often should you see Quads come up? I had quads 3 times one day, and saw a few others. I dont ever recall seeing that many while sitting down playing in a B&M or a home game. Just curious.

Re: statistical tests to show that online poker is not rigged
 

[ QUOTE ]


You're right. I've played hundreds of thousands of hands, and I've seen literally dozens of those "horrific" beats that you speak of, 989:1 (the worst beat possible in holdem) shots hitting. That should NEVER happen. What are the odds of a 989:1 shot hitting!?! It's like a billion to one!

[/ QUOTE ]

That is probaly one of the funniest replies if read in a long while

Re: statistical tests to show that online poker is not rigged
 

[ QUOTE ]
[ QUOTE ]
What are the odds of a 989:1 shot hitting!?! It's like a billion to one!

The odds are 989:1.

[/ QUOTE ]

Nice article...thanks for the link.

Lee Jones article

[/ QUOTE ]

Re: Non random flops
 

[ QUOTE ]

On a side note, how often should you see Quads come up? I had quads 3 times one day, and saw a few others. I dont ever recall seeing that many while sitting down playing in a B&M or a home game. Just curious.

[/ QUOTE ]

The concept of seeing something X times in a day does not have significance. You can expect on average to see quads once every XXX hands. You see a lot more hands per hour online, so you can expect to see quads a lot more often online. In a B&M, you might see 30 hands per hour. If you are just playing 2 tables online, you might see 100 hands an hour. If you play for 6 hours, you see 180 hands live, 600 hands online. So you should see quads more than 3 times as often online. That goes up if you play more tables.

Also, you remember seeing quads 3 times in a day because it is so unusual. You don't remember all the days that you never saw quads. Without a database of significant size, reports of incidents like this mean nothing.

Re: Non random flops
 

Ever get the feeling you might be able to relate to Chris Columbus when he was trying to explain to a flatlander the error in his (flatlander) thinking?

[img]/images/graemlins/smile.gif[/img]

Re: statistical tests to show that online poker is not rigged
 

Thanks for the kind words. I was slow to go to online poker. I don't like online casinos, and it took a little while to figure out the online poker sites were different. Then I was worried about the usual stuff: the site not returning my money, players colluding or hacking the system, software errors. If there had been on-line poker in the days when I played mostly for the money, I certainly would have jumped in; but these days I play for fun; and fun for me is a good high-stakes, in-person game.

I have played some on-line to see what it was like, but I don't think I'll ever get serious about it.

Re: statistical tests to show that online poker is not rigged
 

Thank you for the reply. I'd have guessed you were a B&M guy, but curiosity, and your post, caused me to wonder. again, thx