Please discredit this academic study on the legitimacy of Pokerstars

Status
Not open for further replies.
Danjwarburton

Danjwarburton

Rock Star
Silver Level
Joined
Aug 17, 2013
Total posts
140
Chips
0

This is NOT a moan and groan poker is rigged thread, but a logical request for people with more brains than me to evaluate an academic paper assessing the data from pokerstars. The data from the study supports interference from Pokerstars in the overall game.
This academic study seems to show some form of interfering with the maths of poker!


https://www.academia.edu/34059393/Online_Poker_-_Rigged_or_Not_A_statistical_case_study_Pokerstars

The guy performed a full statistical assessment of a fair sample.

For a TLDR option:

Pokersites make more money when there are very big pots. In order to do this:

1.) It has been found via analysis that once players call preflop, mathematical evidence supports the notion of a second shuffle providing more action flops than usual. The flops are intentially created to give everyone playing a "part" of it. This keeps them playing more often beyond the preflop street.

2.) Coolers happen statistically more than anticipated. The statistical evidence shows that the underdogs win more than they should! If you are the favourite, you should cash out. If you are the underdog, you should stay in the hand and not use the cashout feature.

3.) Leveling the field is a genuine mechanism that works to keep bad players on the site and cause good players to tilt and leave. The site has various ways it nudges the stats in the favour of the underdog statistically more than usual.

Overall, 3 major outcomes:

1.) Flops are purposefully designed to cause more action and increase pot sizes causing more rake for the sites.
2.) Leveling the field exists to ensure losses aren't as harsh for bad players at the expense of goodones.
3.) The website gets away with this by having the occasional extreme in the other direction fudging the figures for the weak regulators that exist out there who look at overall outcomes and see the figures overall match what they should be. The author describes this as follows:

If you need to have some water at 50 degrees C, you have half of it at 0 degrees and half at 100 but only show the assesser the overall figures which would be 50 degrees!

I'm open and hoping for this not to be the case, but wanted help from the community in helping to discredit this. Those who are good with research, can you find any criticisms of this paper?
 
Last edited:
mtl mile end

mtl mile end

Legend
Bronze Level
Joined
Oct 19, 2017
Total posts
11,293
Awards
23
CA
Chips
138
There is not a chance in Hell that I am going to read, let alone interpret that paper.

But I can notice a few things right off the bat. The Introduction was not copy edited. This leads me to believe that it is not a "real" paper. It will not be reviewed or defended.

The author is not a statistician. His Academic credentials are all Social Science oriented, if they are, in fact, valid.

Thankfully, he lists in the methodology the number of hands examined to create a sample size: 55,320 - TOTAL!! I am not a statistician, nor am I particularly proficient in mathematics, but I believe that any real math oriented academic would laugh this off as a joke.
 
PHX

PHX

Legend
Loyaler
Joined
Dec 7, 2014
Total posts
7,127
Awards
16
TT
Chips
68
I did not read the whole thing nor am I going to I just browsed the paper and the conclusions. He seemed to have put in a lot of time and work into this paper if he actually did it.

He clearly stated that there is no complete way to be 100% sure without having access to the pokerstars card distribution. (Top of page 172)

His sample size is 55,320 hands played by him alone. To draw such a conclusion you need a lot more hands and need hands from multiple sources not just one player.

If I had to make an educated guess the average online grinder plays about 600K hands per year.

If the study was 3 separate players each with one million hands analysed it would be more credible for me.
 
ventrolloquist

ventrolloquist

Visionary
Silver Level
Joined
Mar 20, 2019
Total posts
647
Chips
0
It's funny but I just remembered this study and was about to post here asking the same and found this post..

Well, I cannot discredit the whole study but after a quick look I can say that the author cannot make inferences that monster/nut hands are more common than statistically expected because the samples of boards where such hands may occurs is too small. There may have been several studies like this, chances are that one of them will have outlier results and will end up being published. But honestly I think 55k hands is a very small sample if you intend to look at rare situations among all those hands.

For example, the table that says "number of pocket pairs that flop a set". While his overall sample (55k) is big, the percent of that sample that is pocket pairs is too small to reach the conclusion he reached with a high certainty. But I could be wrong. You should search the paper for some probabilities that the results are/aren't due to random chance (not just that they're statistically significant), any respectable statistician would include that kind of info. I'll do so when I have some time.

In some of his tables, the number of showdowns for that particular situation (sample size) is too small and it could easily be random noise he's seeing.

But I may be wrong. Still need to read this thing in more depth. But someone with more solid statistics knowledge hopefully can chime in.

As far as rigging is concerned, I'm sure there are ways to favor certain players, some of which probably would still look random after taking a large hand sample. Some hypothetical possibilities are: (not saying any of these are the case).

1) Having prop players that let new players win
2) Dealing winning and losing hands out in clusters so the overall sample is random, can create cycles of tilt and satisfaction which can keep people more hooked. I joined a new site relatively recently and immediately started off with an 80bb/100 winrate for 2k hands which stayed very constant for that sample :/

But then I cannot figure out a logical reason for rigging beyond retaining new players as long as possible. There is just no incentive to rig unless there are house bots. The rake flows one way regardless. Maybe it's possible that if a site is rigged to favor underdogs vs. solid regs they will be more likely to retain both players but then at that point if you were to take a large sample of regulars and newbies it would be very easy to prove something is amiss, so that's unlikely.

The line in there about mixing cold and hot water and getting correct temperature water is one hypothetical way I've thought about that rigging that looks random might be achievable. Sometimes it does feel like the proportion of action flops that get there vs. over pairs is strangely high but then again those are the scenarios that you will remember because they hurt the most and you'll tend to forget all the times this doesn't happen.
 
Last edited:
ventrolloquist

ventrolloquist

Visionary
Silver Level
Joined
Mar 20, 2019
Total posts
647
Chips
0
If what the paper is claiming is true, perhaps it only happens to each player for the first few thousand hands in order to favor newbies who play few hands, while regs who put in massive volume are slowly less and less affected by this preflop equity leveling. Just a thought. I still think his sample is too small.
 
L

LedZeffelin

Rock Star
Silver Level
Joined
Jan 23, 2020
Total posts
123
Chips
0
All True

Although I agree with all all the author's conclusions from personal experience, more hands are necessary to reach a verdict,

My question is whether it is really possible for the software to be manipulated without anyone ever having found out. That would be way more anomalous than any off-kilter spread described in the paper---which is actually fascinating reading.
 
ventrolloquist

ventrolloquist

Visionary
Silver Level
Joined
Mar 20, 2019
Total posts
647
Chips
0
If what the paper is claiming is true, perhaps it only happens to each player for the first few thousand hands in order to favor newbies who play few hands, while regs who put in massive volume are slowly less and less affected by this preflop equity leveling. Just a thought. I still think his sample is too small.
Ive just thought about one reason why equities may seem more level than "expected". It's possible the equities author looked at are pure heads up and don't factor in card removal effects of a full table. If 4 other people fold junk the density of good cards in the deck grows, combined with fish who play any 1 Broadway hand you now have a deck that has more broadways. At least I think that's what would happen.
 
ventrolloquist

ventrolloquist

Visionary
Silver Level
Joined
Mar 20, 2019
Total posts
647
Chips
0
Although I agree with all all the author's conclusions from personal experience, more hands are necessary to reach a verdict,

My question is whether it is really possible for the software to be manipulated without anyone ever having found out. That would be way more anomalous than any off-kilter spread described in the paper---which is actually fascinating reading.
Ive come across other posts from regs whove been playing for years who feel like underdog's edges are amplified on stars compared to other sites. Agree bigger sample is needed though.
 
mtl mile end

mtl mile end

Legend
Bronze Level
Joined
Oct 19, 2017
Total posts
11,293
Awards
23
CA
Chips
138
WTF. Sample size too small. End. of. story.

Yesterday on PokerStars, I flopped set of aces versus a set of jacks. You know what happened? JACK ON THE RIVER!!!! IT'S RIGGED!!!!

PS quads suckout   slowplay AI on flop
 
Joe

Joe

99.98% Kiln dried
Bronze Level
Joined
May 28, 2016
Total posts
8,333
Awards
10
GB
Chips
116
WTF. Sample size too small. End. of. story.

Yesterday on PokerStars, I flopped set of aces versus a set of jacks. You know what happened? JACK ON THE RIVER!!!! IT'S RIGGED!!!!

View attachment 266542


The guy with ATo was feeling a little sheepish when the cards went over methinks!! :laugh:

Yeah the one-outers on the river are hard to forget, there's a particular example I've been trying to forget for a couple of years already... :eek:

This is the first time I've mentioned it without relaying the hand/runout, progress! :biggrin:

EDIT~ Sorry OP!! I forgot you! Others have mentioned the sample size. If Stars is RIGGED it'd be great to find out, but it's my opinion that will not happen, either way! Most of the western world is capitalist society. Global corporates can get away with almost anything they want and there's little anyone can do about it, sorry! Furthermore, while you are in a downswing (or are just unlucky [unprepared]) it can seem like the RNG is against you, but the truth of the matter is- when you're in a pocket of negative variance it doesn't matter which platform you play on, the same sucky things happen. In my own experience I have left online games which seemed fixed against me, gone to play live instead and the exact same stuff continued happening. Mathematics, statistics and probability (especially) are laws unto themselves.
 
Last edited:
MrrrRock

MrrrRock

Rock Star
Bronze Level
Joined
Nov 23, 2019
Total posts
179
Chips
2
the subject here is very interesting without a doubt
 
Claudiunm

Claudiunm

Legend
Platinum Level
Joined
Feb 3, 2020
Total posts
1,245
Awards
5
BR
Chips
314
I'm lazy about these 'studies' that seek to prove fraud in online poker.
I think it would be much more beneficial to use this time to study online poker and the randomness of the Rational Services Ltd (RNG) Random Number Generator.
Yes, there is a logic in the randomness of the software. But it's fair, because it doesn't choose who it will benefit. begin your thoughts with this affirmation.
Stop trying to treat poker like it's a tamed game like chess.
Poker is wild as it does not submit to anyone. We are the ones who need to learn to dance with him. Decipher it. Or he will devour you!
 
najisami

najisami

Legend
Loyaler
Joined
Aug 11, 2014
Total posts
3,261
Awards
6
MA
Chips
565
Hello,
First of all, I'm still reading the paper (halfway), just in case there's something critical I haven't seen yet. So I apologize in advance if I happen to say anything out of line.
My 1st remark concerns the sample used by the researcher (as also mentioned by few members above). It's like ignoring the hundreds of thousands, maybe millions of hands dealt regularly on pokerstars, compared to which, 55k cannot really represent much.
Then the "rigged" part, the cause of the whole research. Though I can't provide any evidence for it, I just can't think it's rigged. We know that those sites are out there to make money. But we also know that they are already making tons of it (Riduculous rake at PS for example). Thus my question : Why would PS or any other site risk their reputation, their gold mine in order to favor a player over another or stimulate the action that they already have plenty of ?
I know that this point of view might appear too simplistic or naive, but it makes some sense to me while we are waiting for some evidence....
 
Syltan

Syltan

Legend
Bronze Level
Joined
Nov 25, 2014
Total posts
2,810
Awards
8
Chips
35
Curious and intriguing, but doubtful.
 
  • Like
Reactions: Joe
F

fundiver199

Legend
Loyaler
Joined
Jun 3, 2019
Total posts
13,261
Awards
1
Chips
252
This thread was started more than 2 years ago, and back then I did actually look a bit at this "study". And as others have said, its basically a joke. Sample (55k hands) is way to small to mean anything, and he even manage to "proff" the opposite of, what he is trying, since most of his parameters are well within normal expected standard deviation. Like something was supposed to happen 20 times, and it actually happened 22 times.... The author is basically just another sour loser, and he would have been better off spending his time to study poker strategy rather than produce this "paper" :)
 
Sonxaku

Sonxaku

Rising Star
Bronze Level
Joined
Feb 14, 2022
Total posts
16
Chips
0
I unlucky cant show u the guy who showed me that EVERY HAND in online poker is exactly the right % of possibility.

He was analyzing 2 million online hands and the results of EACH HAND are the same results wikipedia will show u if u google how much chances you have after this and this hand.

I dont really think its rigged, but i also think its not ABSOLUTLY possible to write a code which ONLY is based on "the-luck-factor".


I wanna share my sickest hand in online poker i've ever experienced.
Please Note: Preflop All-In!

https://ibb.co/3cTfxbs
 
Last edited:
C

canabero

Legend
Bronze Level
Joined
Aug 30, 2014
Total posts
1,515
Awards
10
Chips
10
I did not read the whole thing nor am I going to I just browsed the paper and the conclusions. He seemed to have put in a lot of time and work into this paper if he actually did it.

He clearly stated that there is no complete way to be 100% sure without having access to the Pokerstars card distribution. (Top of page 172)

His sample size is 55,320 hands played by him alone. To draw such a conclusion you need a lot more hands and need hands from multiple sources not just one player.

If I had to make an educated guess the average online grinder plays about 600K hands per year.

If the study was 3 separate players each with one million hands analysed it would be more credible for me.

55,330 hands are more than enough to have conclusions. take in mind that normally guys in this forum take conclusions for less than 10 hands played.
and, yes i have not read the article.
 
A

AllOut

Rock Star
Bronze Level
Joined
Feb 20, 2021
Total posts
276
Awards
1
Chips
53
55k hand size is okay and the guy speaks truths. No need defending online poker, you don't even know who you're playing against in the tables. Even this factor alone makes me think how dishonest this environment can be. Would you even realize if they put their own accounts into mtt ITM places to snatch piece of prize pools? You can tell me you have no proof, but do you have any assurance? Like most casinos did these acts throughout history, the probability of pokerstars setting such robbery traps is high in my opinion. I imagine myself in the head of it, all I see are oppurtinites and clean space to steal the money.
 
Status
Not open for further replies.
PokerStars Guides: Italiano - Dansk - Nederlands - Deutsch - Français - Español - Polski - Norsk - Português - Svenska - PokerStars Mobile - Deutsch Mobile - PS Casino Mac Poker
Top