You are using words that you don't understand and in a context that makes absolutely zero sense.

has to be a level, if so gg sir

Hello,

I posted earlier in the thread about the lack of actual cheating evidence in the OP, and suggested that if collusion occurred, it was probably by one colluder folding hands preflop when the other has a good hand.

Feruell contacted me and sent me an export of his database with the suspects' hands. I also sent an email to Dooshcom after he posted and offered to look at his hands in case he wanted to clear his name. He has not replied.

The Suspects

Dooshcom, benkaremail, liverfc123, benhamcheese are the suspects per OP. There is circumstantial evidence suggesting that the latter three are the same person, including benhamcheese showing up at 500-1000 5 mins after liverfc123 busted. I defer to Feruell on this evidence.

The analysis groups all four suspects and will be hereafter referred to as "Team". However, I would like to note that Dooshcom/benkaremail's hands are about 80% of the sample, so the statistical power of the following analysis applies mostly to those two players. I will follow up with another post looking at benhamcheese and liverfc123 individually. For now, I'll say they show the same pattern but on a smaller sample size.

The Analysis

I used the following approach in looking for evidence of collusion:

1) Filter for hands when a player has (77+, ATo+, ATs+, KQs) in the SB or BB and look at how often another player opens before him from the Button or Cutoff. The following splits were compared:

(a) Team player in BU/CO opening PFR* vs Team player in SB/BB
(b) Team player in BU/CO opening PFR* vs Everyone Else (not a Team member) in SB/BB
(c) Everyone Else in BU/CO opening PFR* vs Anyone in SB/BB . This last one is not necessary to reach a conclusion, but it's nice to have.

2) Filter for hands when a player saw Showdown in SB or BB. These are not all premium hands, but they are stronger than a random hand. Again, look at opening PFR* from another player in BU or CO.
- Same splits as above

* Important Technical Note: Filtering for SB/BB holecards biases the sample in a way that affects the observed BU/CO opening PFR. For this reason I am labeling the observed value PFR* to distinguish it from "true PFR". See note (i) below. That being said, we can still do apples-apples comparisons between the PFR* resulting from the 3 splits above.

RESULTS

The chart below shows the results for the case where the SB/BB has a premium hand. Notice that Team PFR* when another Team member holds the premium hand in the SB/BB is less than half of that when a non-Team member holds the premium hand. The results for the Button are statistically significant at the p < 0.0001 level. The Cutoff results are not statistically significant due to small sample size. (ii)



More Detailed Results

The following table shows the data used in the above chart, with sample sizes included. It has additional splits, Doosh vs Team and Team vs Doosh. This is to show that it didn't matter whether Doosh was the one opening the BU/CO or viceversa, they both have a low opening PFR* when the other has a good hand in the blinds.

It also shows an "Inferred PFR" - This is an estimate of the true PFR from PFR*, which depends on the opening/calling frequencies of the players that act after the BU/CO. I used the regulars' frequencies for the purpose of this estimate. It gives an idea of the players' actual PFR. Note that we estimate that the Team was opening < 20% from the button when their buddy had a premium hand in the BB. Their overall button opening PFR was 46%.

Compare the "Team vs Team" and "Team vs Everyone Else" rows in the table below. The p value refers to a statistical comparison between those two rows.




Lastly, below are the results when the buddy had any hand that went to showdown in the SB/BB. It still shows the same pattern, statistically significant for all cases except CO vs BB which has too few samples. This suggests that the collusion was not limited to the most premium hands.

Note that estimating the true PFR in this case is more complicated. The inferred PFR is not always smaller than PFR*, as was the case when we filtered by premium hands. If anyone's interested in the technical details, let me know.



That's it for now, I think this is conclusive, more to come later.

CLIFFS: Dooshcom and team were opening about half the normal amount when a buddy had a premium hand in the blinds.


(i) The easiest way to see this is to imagine that the BB has a premium hand, and will always play as a result. When the BU/CO raise all-in, the BB will call and all those hands are in the sample. Likewise when the BU/CO fold but another player puts BB all-in. However, when all players fold to the BB, the hand doesn't go to showdown and is not included in the filtered sample. Thus the sample PFR* in this case will be higher than the true PFR since it doesn't include those times when BU/CO folded and the hand didn't go to showdown.

(ii) I'm not a stastician, but I googled around and used the tests on the following web page, which seemed to be the right tests to use. The proportions compared were the Team's PFR* when a buddy had a good hand in the SB/BB vs the same stat when the good hand is held by someone else.
http://faculty.vassar.edu/lowry/tab2x2.html

Wow that is some solid work jaytorr. A+

Nice job, jaytorr.

I wonder if dooshcom is going to accuse you of something.

Hi,


Furell’s Formula: IF dingdonk84 on ipoker = dooshcom, then dingdingdonk on 888 = dooshcom.

What mathematical formula was used here exactly to conclude this? If you think that dingdonk84 on ipoker is me, how does this conclude to dingdingdonk on 888 (a completely different network) is me and therefore I must be multi accounting! And how stupid exactly are the people that then wrote posts saying - PROVEN! Absolutely nothing is proven by this besides that the names sound slightly alike lol.

Just for a laugh if could be bothered I'd create the username thrcnbeonly2 on ipoker then claim that because I think that thrcnbeonly1 on 888 is you, then thrcnbeonly2 is also you on ipoker. This is NOT evidence.

Less aggressive/soft play was an example. My point is that statistics will be different to the normal. Obviously increased aggression is also an example of this statistical change, not evidence against.


Randomly:
1.If I wait a day or two before making a post, I am probably guilty for taking so long. If I reply instantly I am obviously guilty, why else would I need respond so fast?
2.If I write a long post it looks guilty because it's so long. If I'd wrote a short post it makes me look guilty because I didn't even care enough to write a long one.
3.If I write a post with some ‘emotions’ in it, it gets called "emotional gibberish" (FYI - humans have emotions), if I was to write a post without any emotions then I simply didn't seem like I care.


I hope this post is exactly the right length and has exactly the right amount of emotional gibberish in it to please my critics. This will likely be my last post here as I see no point in continuing when regardless of what I say or do, I am ridiculed.


To clarify a few things:

I decided to take some legal advice regarding this whole situation. Assuming the Russian mafia (http://forumserver.twoplustwo.com/18...04/index2.html) doesn't come after me, then this could have proven to be useful, especially the name defamation advice. When a player has 100's of thousands of $ in their account and the pokersite decide to close the account and confiscate the funds, what would the poker player do then? Go straight to court where the standard is "beyond reasonable doubt" and alongside getting his money back sue for name defamation, and of course compensation. The headlines read "Pokersite doesn't pay out". Headlines like this would damage the entire poker industry, in turn damaging ev for all of us.

The reason I point this fact out is because we live in a world of laws. Gone are the days where we tie a rock to a suspected witch’s leg and drop her in the ocean. If she drowns she was innocent, if she fly's away she was guilty and is now proven as a witch. Long gone are the idiotic days of ‘guilty till proven innocent’, thankfully we live in a world of ‘innocent till proven guilty’. This is the standard I would like to be judged by and I see no reason why it should be the other. In terms of poker - Sure, this standard obviously makes it harder to catch cheats, however it is also the only reason why each and every one of you have not already had your accounts closed down several times over and the funds confiscated. Poker has lots of variance, not just with regards to the financial swings. At some, or actually during 8 years of poker, at MANY points in time every player’s statistics will look off from the normal (what is the normal/is this figure the exact same for every player/who decides what this normal is. Furell?).

There is constant statistical variance in your play, caused by cards/players/tilt/change of game plan and many other things. In a game like push and fold this is enhanced even more. Just like there can be insane variance in the game (losing overpairs 10 times in a row or whatever) the same variance will happen in all of your stats, especially in a nolimit game, and even moreso in a preflop only type of game. If random data samples slightly different to the normal ALWAYS = cheating, then God help us all. Thousands of people complain daily about pokersites being rigged because of this variance, and their graphs in the given period of hands would look something like the above. (“I lost AA ten times in a row followed by losing 15 coinflicks in a row. The site must be rigged”). Luckily for pokersites they have RNG (random number generator) to PROVE otherwise. Unfortunately I am not a RNG.

Over the course of all of your careers you would have been reported multiple times for cheating. Perhaps by less known figures, but also by other regs in doubt of your play. In the case of regs, it is likely to have been with good reasons and some “evidence”. Maybe you played different for a small period in their eyes. Maybe you opened the button or 3bet certain players more often. Maybe you purposely did this for your own reasons that did not involve colluding, but none the less it happened. This is quite normal. You started running well and increased all these %'s even more on purpose. Another reg noticed your %'s changing and also notices that it seems to be frequently when player bob123 is in the game. Maybe you're colluding with bob123. You get reported as a suspected cheat and, YES, your statistics has changed during this small sample of hands, and your gameplay vs bob123 is different to other regs because you think bob123 is likely to respond to certain plays differently to other regs. Maybe it's not that you think bob123 is different to other regs, coincidence just happens to have made your stats different vs him to others Maybe your stats are also a little different vs bob456 in this small sample of hands. When playing hundreds of thousands of hands per year, of course stats are skewed at various points. Hence the term variance.

According to this thread, the only conceivable verdict on the above scenario is guilty. The outcome, freeze his account and run his name through the dirt. Accusations like this would have happened to ALL of you at some point, probably several times over. Therefore you are all guilty, as you would have all had dubious looking stats at some/many points in time vs various players. However your funds were not confiscated, your accounts not closed and your name not run through the dirt. Why not - Because that would be stupid. You probably did not even hear about the accusation because pokersites simply do not make public the information that someone has been ACCUSED and usually the accusers don't either. Pokersites will usually not even mention it to you if you’re a reg, unless it is really warranted. Because of this maybe an occasional player did cheat and get away with it, but the vast majority did not cheat, and should not be judged to have done so at such an earlie point and with such little evidence. The investigation should take place and IF found guilty, then they should be ridiculed. Not before!

This is currently the situation I am in, except some of you feel that guilty till proven innocent is the correct line of thought. Suit yourself, after this post I am done arguing otherwise. The evidence posted in this thread is barely evidence as some have realised, and some of it is just simply incorrect. Yet the verdict is guilty. And why? Because a public scandal is far more fun than no scandal. Forum threads are boring when no one posts stuff like this. Forums can be useful for many things, but let’s face it the reason that thrcnbeonly1 has posted this here is not to “warn” others about me colluding. How can they? Dooshcom is not playing on 888 since the accusation was reported to the CEO of 888 and let’s face it, if someone wanted to collude in this way, they’d simply make more usernames in the future on different networks in a different person’s name. Thrcnbeonly1 is not trying to warn anyone to avoid me as this is impossible, nor is he trying to “prove” with any legitimate proof of being cheated. All that this thread was ever meant to do was to heighten the profile of someone under investigation yet deemed guilty, regardless of the result. If it was result dependant in some way to ridiculing me, they’d have waited for the result of the investigation which actually proves some more substantial proof! The reason? Likely because he feels it MAY be true, and is really pissed off anyway at me for whatever reason, so does this. Regardless of the outcome, it has achieved its purpose of pointlessly ****ing me.

Finally, so far we have 6 unnamed individuals posting accusations. If proven wrong, they can change their username or just say "opps, my bad" and lost nothing really besides a bit of street cred in the old username that had no wider reach than 1k-2k or so people. As a WSOP bracelet winner my name has still been run through the dirt to a much wider community and future sponsorship deals stifled regardless of the outcome. So how about you guys all at least add a little bit more credit to your claim and put some real names on the table here besides mine. Who exactly is needaaaa/thrcnbeonly1 and the rest? Since you clearly wanted to make all this so public on a thread of worthless evidence add your own names to the thread so that when proven innocent it can have a wider impact on your future poker careers too. Have something to lose besides - “oops, my bad” at the end of this.


Darren Woods.

So many words, so little content.

Prior to jaytorr's post I was having a hard time making up my mind what to think about this situation. The evidence that had been presented made it seem very likely that something strange was going on. The hands posted didn't make much sense. However, specific evidence that the players in question were colluding was lacking.

Assuming the analysis in jaytorr's post is accurate, I'm convinced that they were cheating. I'll leave it to someone who has actually formally studied statistics to tell us what the probability of generating these stats without cheating is, but my intuition is that it's very close to 0.

1) When you read the original accusations in post 1 the analysis has nothing to do with them. They're not accused of folding when there's a buddy with a premium hand in the blinds.

In fact they're accused of doing almost exactly the opposite thing (of being more likely to be involved in a pot with their "partners": The second big thing we have discovered and have made a thorough looked at is that when any of benkaremail or dooshcom moves allin and gets a call from an opponent they will overcall with really bad hands that there is no question is a really bad play unless you are cooperating to win the pot.

2) Analysis has to be defined before it's done and then with an open mind you go find the results. You'd check each person against each person in the blinds and see if there's an unusual pattern of folding that only exists between specific people.

3) The reason that analysis has to be defined in advance is that lots of things vary of the average and just because you found one things (out of the many) that you can could have checked doesn't constitute proof.

4) For one thing you'd want to know how sensitive the results are to what's been defined as a premium hand (which is totally arbitrary and must be defined before analysis starts).

1) I agree. The original poster and the people who helped him begin looking into this did not do a very good job. They had an intuition that something weird was going on but made an unsophisticated and misguided investigation of it and the evidence they produced was uncompelling.

2-4) I think the criticism that the analysis was not defined before it was undertaken, with respect to jaytorr's findings, is wrong. He lays out a hypothesis in his early posts in this thread:

The best way to collude in shortstacked NLHE games is to play much tighter when your teammate is still to act and holding a strong hand (aka best-handing). If a pattern of this sort can be observed in the play of doosh and benkaremail it is very likely they are colluding.

He then goes on to confirm this hypothesis.

Still, I would agree that a more thorough investigation is called for before any conclusive judgement is passed and certainly before any money is seized or redistributed. You identify two things that I agree are very much worth looking into:

Can any patterns similar to the one observed between doosh and benkaremail be observed between any other pair of players?

How sensitive are the conclusions to the range of hands identified as "premium?" Obviously if the effect disappears if you make the range tighter or wider it is much more likely to be an artifact of sample size and variance than it is to be evidence of cheating.

It is also worth looking into how likely it is that the players in question could appear to play this way over the sample in question without actually deviating from normal play. In other words, is it possible that they did not make any unusual folds and nevertheless appeared to play this much tighter against each other in these situations?

FWIW, when OP originally contacted me about this, the collusion that he described is exactly what's alleged in jaytor's post. I also was sorta weirded out by the fact that the hands that the hand or two that he showed me (before I told him that I'd rather look on my own) didn't really match what he was alleging, but the actual qualitative description that he gave me of what was happening was that they were best handing.

The analysis jaytor did is roughly what I probably would've done. (I did some initial very simplified analysis when OP first contacted me about this and got some mixed results that suggested I needed to look into it in ways that are a PITA to do in HM.)

Jaytor,
Get my skype from OP or e-mail me at noah@subjectpoker.com letting me know your preferred method of contact.

The standard in civil lawsuits is "balance of probabilities." I believe that case has been made against you for what it's worth. I do not believe that unless all your hands are know that a "beyond a reasonable doubt" case/criminal fraud case can be made.

Has that $50,000 bet proposal suggested earlier been dismissed completely by you? If so, why so as a finding in your favor would not only earn you $50,000 without the need for legal fees or delay but would certainly clear your name to the satisfaction of everyone in this thread and beyond while painting your accusers with a very dirty brush.

So basically your defense is?

A. You're not using all of these accounts
B. Statistics will be different than normal in games with fish, therefore the abnormal statistics cited don't carry weight

I got nothing else out of your entire response besides "Woe is me"

I would love to see a criminal expert in lying look at Darren's response and give their opinion.

He does a pretty woeful job of defending himself, but with the analysis and discoveries already posted in this thread without aid of seeing all his hands I don't believe he can do much else but bang the table. ITS PROBLEM.

Certainly if it was me, I would 1) address the specific hands in detail; 2) address hands that the original accusers played with him (since they could confirm it) that show him acting against the interests of hiim and his alleged co-conspirators should they exist; 3) admit to multi-accounting and using HEM and say it was stupid to say otherwise; and 4) take that $50,000 bet offered -- it would provide vindication and a nice payout at the expense of his accusers who would then look as bad as he looks currently. Simple

Team VPIP in SB and Button after Buddy Opened

Another comparison we can make is the Team's VPIP in the SB and BU after another Team member opened, vs the same stat against everyone else. I wish I had l looked at this earlier, as this stat is simple to understand and the sample sizes are large.

Their VPIP from SB when Team opens BU or CO is less than half of their VPIP vs everyone else. It's 10% SB vs BU (25% vs others) and 7% SB vs CO (16% vs others). This is over 484 and 221 hands, p values are < 0.0001 and <0.01 respectively.

BU vs CO Open shows the same pattern, not quite statistically significant but close (p<0.10).



Note that they were probably not folding hands in the 10-15 percentile, it's just that when they had a good hand in the SB, their buddy opened way less than normal.

Additional Supporting Data
In the rest of this post I'm going to include additional data and discussion, which is not critical for the analysis but might help us understand their pattern of collusion.

BB VPIP vs Open in BU or CO

Again, their VPIP in the BB facing a single raise from BU or CO is lower when the opener is a Buddy.



There are three possible explanations for this data. I think probably all three are true to some degree.
1) Buddy opens very tight when BB has a good hand. We already saw the evidence for this in my previous post.
2) Buddy opens very loose when BB has a bad hand. There is evidence for this as well, some in the OP. Below is a chart showing their Button opens that were shown down.



Taken together these two points would explain the very loose opens observed - the above range looks to be about 67%- while their overall opening PFR remained reasonable: Their overall Button opening PFR was 46%.

3) BB folds more marginal hands when Buddy raises (to save on rake?)

This is very possible, for some reason they were folding more often blind vs blind when playing 3 handed or more, but not when they were HU. As OP argued, the HU play was for bait/cover and they might have played differently then.



Maybe we can get someone from the probability forum to look at these tables and chime in. The reason that stats tests exist is to quantify the probability that you would see a set of results due to random chance. When the test says random chance is very unlikely, you need to find another explanation for the data.

I think the data speaks for itself and the case is very strong. I will also contact NoahSD to get his input.

Also Ike- Thank you for your suggestions. I will consider doing a sensitivity analysis with a wider range.

I suppose you could call it a level in the sense that I don't really believe it. I was trying to give an extreme example of an excuse he could give and I guess I went too far, but it's not much more ridiculous than the reasoning he had for his play in that satellite thread.

As a former fraud investigator (for discover card), I have had little doubt (for non poker reasons) since his first post that he was guilty of something, but when the standard of beyond a reasonable doubt was proposed for a wager I thought I'd play devil's advocate. There will always be doubt in a case like this, it's up to poker players to decide what is reasonable.

Doosh, its disgusting to read your posts. Time to confess and suffer. To late for bla-bla-bla.

Dingdingdonk and dingdonk1984 played me during the same period of time, at the same limits, with the same style. And they are both from UK (you can see the counrty of the player on 888 and we know dingdonk1984 on ipoker is Doosh).

What are the chances they are 2 different persons? Dont insult the intelligence of the readers.






Why do you bother to lie about multiaccounting when you are about to be crucified for much more serious crimes?

Great work Jaytorr, agree on getting someone from the probability forum to look this over. Also big props to Ike and Noah.

It's great when the cheaters get exposed.

Doosh, people are still waiting to hear whether you accept the 50K bet that has been proposed, and we'd love to hear what you have to say about jaytorr's findings.

(that is if we ever hear from you again...)

I really hate playing devil's advocate, but he never actually said they were different persons, he said the similarity of their names wasn't sufficient proof that they were not. He has a habit of implying falsehood without actually lying.

Jaytorr,

Your two efforts have certainly proved the case on a balance of probabilities or by a preponderance of the evidence (whatever standard you like and, really, to anyone with a sophisticated understanding of the subject matter beyond any doubts that could be described as "reasonable" from the evidence and arguments we have seen). Certainly, any jury of intelligent poker players would say the case has been made even without feruell damaging Doosh's credibility by showing a falsehood/non-answer answer in his reply (but that helps).

I await 888's efforts at protecting its credibility.

Me posting anymore is pointless, no one in this forum believes in variance and everyone thinks that data samples of less than 600 hands are that significant enough to draw charts like the one above and somehow to be able to prove collusion.

All I can do is deny all these accusations and allow you all to continue your online slanging sessions. Maybe after this I should hire someone to datamine for a while and find a 600 hand sample that doesn't match the norm then send email accusations of colussion with "proof". Might turn out to be a fun hobby.


lol yes Jaytorr. This random statistic which is also different in the same way as all the others, was to save on rake. What a great example that you guys will shout "colluding" and use any ******ed statistic that you can to prove it. Saving on rake during a 500/1k session makes about as much sense as this entire thread and everything else that your 42-574 hand charts prove.


Please continue in my absence, this is daft and obviously the witch hunt will not finish any faster with me responding. GL with your witch hunt nameless accusers. I am done with responses.

What a pathetic end.

Based on your first post I was hoping we may witness a second coming of Scott Fitzgerald or Hemingway.

First off it wasn't a specifically found 600 hand sample it was hands they played with you, so there's no bias there. As to claiming it must be variance:

The p values basically mean that "given that you play the same against the team the same as everyone else, there is a .001% chance of seeing this data in a sample this large. So number of hands is pretty irrelevant the p value is what matters. I'm a bit rusty on my stat though but it seems intuitively the one thing missing is comparing doosh to the other individual player players. The p-test is reliant on assuming he is playing the same against alleged team members as the rest of the field when clearly there will be small adjustments and all the p-test can show is the odds of these results coming up when he plays the same.

I think the next step is to run the same tests on various regulars which whom you have enough hands to analyze and then run a p-test on the sample with team members against the sample with the lowest vpip against a person you think he is not colluding with (and have sufficient confidence in the value). I'm super rusty with my stat and I'm sure someone can clarify better but I haven't seen much discussion on this and while at the moment it seems doosh is super guilty I'm still not entirely sure the data actually proves anything is going on other than he plays differently vs. these players than the average player. You have to show that they are the only players he plays this different from average to show there is collusion, as I think Noah did in the Stox collusion analysis. If this can be shown I think it can pretty much be agreed there is close to 100% chance that there was collusion.