Active 888 account, I found this one by checking HEM DB. Everything is a match, 3bet=sqz, BB Agg=Agg etc. The RPTR stats also look similar for such small sample.

Good Job Schwein they should def consider hiring you after the work you did

was just about to post ^^^

Incredible how absolutely sht stars is at security lately, these cheaters would have kept playing for ****ing years and feeding greedy PokerStars rake.

Just disgusting, im glad they got busted now though

Good job Schwein. Now we need pokerstars update from their integrity team on what happens to all of us taken for millions by these bots. No official statement yet but I'm sure this is happening on ever poker network. I still think hole card sharing has a part of this as to build a winning PLO bot without this is hard going knowing that if you paired with Bot No.2 then you know 4 dead cards only? Hope to find out soon. Stats definitely state collusion.

What bother me now is the accounts like seregaxx and other bots are wining aswell SNG/MMTs with a huge volume; I think the boting could also happen with SNG/MTTs

glad some people got fired over this for sure! (rumours ofc though). Quite a few regs at nlhe 100/200z who also fit into a russian botring /sw are under investigation aswell! At least theyre actually doing **** now finally

It's a rumour which I think Is just that. I don't think stars at this point with what's happening would sack their "integrity team" or they would not be able to investigate Schweins good work. We all had a common incling as to why we were beaten by totally trashy hands pre-flop and blah blah. I kept wondering how on earth seargeaxx was sitting on 3-5 buying per table when he re-appeared for a brief time? He was hitting both runs on a couple of occasions 50/50 runs of 2 boards? I'm doing my own stats by end of the week and let two statisticians from Hertfordshire in the UK have a go at figuring out how they can win so much playing like donkeys. They can grasp any game these statisticians and I'll send report first to stars unless they are happy they are ok with Schweins great work. They will work on Schweins RPTR and anything else they think is useful. The games seem a lot better as of now and thinking I can beat $2.50/$5 upwards but not $1/2. (Am break even thanks to $1/2)!
As for the UKGC I think until October 1st when new chief begins is when things might happen as they have a tiny budget and not much success and Richas is your user to speak about legislation.

Hot do you calculate the distance between two accounts? What stats do you need? I wanna see how much i'm different from those guys.

Hey, guys.
My topic was closed due to the fact that there is already tread about bots. I was told to write in this topic, so I'll copy all the info that I found into this tread. Its under spoiler, so who is sceptical about it can skip it.

Ok, I got you. I'll try not to make quick decisions and maybe do it in privat.

This is the point I intend on. It's not the fact that the group that was discovered is the only one and it's very possible that there are other groups of bots which may have different key stats.

Anyway, I keep my investigation opened and as soon as I find something solid I will inform you. I have some guesses but I need more time to make deeper research.

My take on legitpoker posts.

1. The accounts I have picked were not only very similar to each other but also very similar to banned botters/TOC offenders. I think this alone makes a huge difference. Out of big winners on 100-400 only about a dozen accounts vanished and most of them were identical. My accusations in the OP were never based on just RPTR, but also on a connection to banned accounts, various HEM stats, playing schedules.

2. Some of the accounts on my list had very weird playing schedules, same number of hands per day, 5 days a week, never moving despite huge winrates, never taking shots. The 4 accounts you listed play a wide variety of stakes, some play hu etc. Their hours and numbers of hands per day are not consistent.

3. The 4 account mentioned by legit do some stuff from HEM really different. They have different distribution of aggressive actions on flop and turn. They also have different Agg in general. 3bet from SB/3bet from BB ratio is different for all of them. Probably more, not bothered to find out.

4. As for the RPTR stats. I wasnt aware that anyone else is so close, looks like its possible after all, but when it comes to different stats they have a more or less normal distribution. Its more likely to find similar accounts for what I would call average statistics. For example if we pick only accounts with both c/r flop>13 and c/r turn> 13 it will be a really small group. If we pick accounts with c/r F,T around 8-10 the group will be way, way bigger. Finding 27 very similar accounts in a small population is more alarming then finding 3-4 similar in a large population

5. My concerns were not just based on the RPTR 21 stats but also on sharing some other similaraties. I have posted some and other just shared with PS Integrity Team. The 27 accounts had 3bet=sqz, very flat Agg by position, BB Agg=Agg, very similar distribution of actions on flop and turn and so on.

6. The red lines, some people have them going up, some going down. The accounts on my list have them going up first and then switch halfway around mid 2014.

7. The 4 accounts(6 pairs) picked by legit are more different then the accounts in my thread. 3/6 account pairs have a difference of 250+, in my thread all pairs were under 250, most under 100 and some as close as 35. So the ring theory over there is a flawed one. There is one similar pair of accounts which means its worth checking. I checked that pair by looking at various HEM stats and those were really different. Case closed for me.

Honestly, I am not concerned at all here, but if you are you can always email Stars.

I have stopped looking for any new names on Stars a few days ago. In case I missed someone, Stars will not. They are really working on it. But mostly I have stopped looking, because I didnt want to become overzealous and be like, hmm how about this one also kinda looks similar.

The thing about botting is that its never 2-3 accounts, when someone creates ground breaking software its always plenty of accounts.

I dont have time to reply to every possible list someone is going to post.

I did it here to explain my methods further. Its generally a 2 step process. Pick accounts based on 21 RPTR stats being similar. Step 2 is to check everything else there is in HEM. I wanted to avoid selection bias from the beginning and that is why I was picking the accounts based on some stats and then comparing once again by looking at other stats. At first the step 2 stats were 3bet/sqz ratio and Agg distribution by position, now its a whole list. The additional step 2 stats should be uncorrelated to the stats that were used for step 1 that is picking the accounts.

For those unfamiliar with statistics and selection bias. If we select a group of accounts out of general population of accounts based on several factors and want to find out if its a random population of accounts having those factors in common simply because a large enough population will have similar pairs or groups due to its size we need to do the following.

We have to find a stat which is uncorrelated to the stats which were used for selection of the sub population. The stat has to be shared by only a fraction of the general population, the less of it has it the better. I mean if we find something which is true for just 10% of the general population it should be also true for 10% for the selected group if its a random one.

That is why I chose high 3bet/sqz ratio (>0.8). Its not related to being a winner or having high AFQ or low fold to cbet etc. The high 3bet/sqz is also shared by a minority of accounts on Stars.

So if the 27 accounts truely had the 21 RPTR stats similar by coincidence they should have the 3bet/sqz ration in line with the general population and that is not the case, they all have it super high.

I have emailed Stars Game Integrity Team with a list of uncorrelated stats which are uncommon for general population but shared by all 27 accounts. I didnt make some of those public for obvious reasons. I was looking for those stats even in my OP, found some more since then.

I hope this explains sufficiently why I am suspicious of those guys in particular and not worried by others.

For me having the 21 RPTR stats really similar means something might be going on and its worth looking into it, but its not damning enough by itself. Like I said its worth looking so I checked the 4 accounts listed by legitpoker and they have the extra uncorrelated stats all over the place which means its just a bunch of random guys that somehow happened to be similar due to having common stats in a large population.

Can't wait to receive my standard Stars mail: "You got cheated, we are not gonna tell you how it happened or who did it, we are not gonna tell you by how much, but here are $ 32.96, take it or leave it. Sincerely, your PS team"

Can someone PM me a code for Russian PTR I wanna check out some of these accounts, thx in advance

They aren't playing like donkeys. They play "better" than most regs. That's why they win. The problem is that they are apparently using some sort of automation software that allows them to consistently beat the regs while never going on tilt.

Collusion may have occured though. They sure didn't collude like donkeys.

I'm not an Omaha player so I'm somewhat confused about the stats.

Is it possible that these bots are winning by playing such an unorthodox style, and without collusion? If so, why aren't humans playing this way if it's so profitable?

Could be possible. There are a couple of legit regs playing a bit the same.
They are winning a lot because most players fold way too much against them. Game theory wise they make huge mistakes. But that doesn't matter sometimes when your opponents are not that good --> folding too much.



By the way: have not seen them playing for at least 3 days now I believe.

What is sure is they will be back soon with different stats

Email from stars:


The Bot situation that you mention is currently being investigated as a matter of urgency. I can let you know that of the accounts mentioned in the opening post of the 2+2 thread, 2/3rds of these accounts had already been closed by PokerStars. We are conducting a thorough re-review of our original investigation and also taking steps to identify ways of monitoring our tables if these players attempt to return to the site.

well 2/3rds was closed already at the start of this thread. that said didn't saw sergej, fedora and sam for 3(?) days!?

@Schwein, do you have any intention to check PLO100/200 on iPoker as well?

+1
Ipoker is shady all the way down to PLO20...

I have the intention to do it, dont have the means yet.

Uploaded a recent FTP DB 3 days ago and emailed PS Integrity Team a new name which isnt included in my OP. Turns out they have beat me to it and already found that guy. Checked Ongame, couldnt find anyone there.