Your point that a high ratio of win rate to SD is an indicator of cheating seems sensible and adds something to win rate alone. Of course in extreme cases that detail isn't necessary.

The significance of the ratio of win rate to SD obv depends on sample size (Would some t-test type approach be helpful here?). Your chapter addresses the case of marked cards (or online, a player knowing opponents' hole cards). The indicators of cheating will vary by cheating type - e.g. whether the cheater can see hole cards, future board cards, equities (as in this case) or is colluding. I would expect a good poker site to have a range of indicators, regularly review lists of outliers and examine actual play in detail where the stats highlight a suspicion.

Why is it necessary to break the sample down into discrete groups to calculate variance of win rate?

No understanding of hacking / intercepting packets etc. here...

But since many other sites provide all-in equity cash outs (stars & iPoker for instance). Would they be susceptible for this kind of cheating as well? Or is this a security breach unique to GG poker?

If it's true that this was some sort of white hat hacker group (and probably if not also) they'd probably have tried the same thing on other sites as well.

Doubt anyone knows the answer to your question, but the only place I would trust playing on completely is Stars. I'd guess that it's very, very unlikely this same thing happens anywhere else out of the bigger sites. GG has seemed like a total joke, and dodge in many ways, to lots of players since years ago, I don't think it's a coincidence that this happened there.

There's always a chance though that it's a competitor behind this whole thing and that they targeted GG specifically and don't have any interest in doing this elsewhere.

Any even mildly decent programmer wouldn't allow that information to be sent out before it's supposed to be shown, so it's unique to absolutely terrible programmers and security practices. If I were to program anything like that, I'd go through the code a gazillion times and make a ton of unit tests to make sure it never happens.

MoneyTaker69 was reported on the 25th, but I wrote about this problem on the forum.gipsyteam.ru back in November, referring to a video from May 2023 where this error is visible

I wasn't clear in my initial post. I agree with you. My point was that in this case they should have the ability to go past the statistical analysis to analyze who among the "warning sign cases" used the specific technical vulnerability they've identified Moneytaker as using. If they have this ability, then in this case the question of whether outliers that look a lot like Moneytaker (perhaps a tamed-down, less-obvious-acting version, but still clear anomalies) are within the bounds of what's possible for a non-cheater becomes less relevant. Of course they should be doing stats on their player data - that seems basic. But it's not clear that they do, since it seems we need players with less data (datamined) to signal them. That last bit is particularly concerning. Of course, a lot about this is concerning.

They’re likely churning accounts to game the PVI system.

pretty sure zoom type games like rush&cash are untracked
a smart cheater would vacata there for that reason, this is just the tip of the iceberg

Apparently the thumbs up/down feature is using third party software with known vulnerabilities, like Adode Air. At this point the theory seems to be that the modified client can exploit this in someway to get server side information such as equity state data.

I don't know much about adobe air but I seen this posted.


Others sites like Stars I am sure are not stupid enough to use something with know vulnerabilities that isn't fit for commercial software.

Unless I'm missing something, that isn't relevant here. Not entirely sure what the video is about, but if 100% was shown on the flop, it's really got nothing to do with the MoneyTaker69 exploit. Looks more like a visual bug. Not even sure the river is known server side at the time it shows 100%.

You're probably right the vulnerability has existed for a long time though; the thumbs up/down wasn't introduced recently.

the thread is cluttered with so much noise that the real questions are getting overlooked .

those results look suspicious too, albeit its micros, another instance of 500nl was shown with some suspicious wr's. how many others have been using this exploit, i think we need to do the work for gg, who wants to ransack their security department? eddie get your coat i will get the balaclavas, or just hack the servers that way its probably so easy still?

MoneyTaker69 has played 7377 games with a standard deviation of 87. An average player with stats 53/17/4 has a loss rate of around -50bb/100:



Let's assume that he was an above-average recreational player postflop with those stats and had a realistic loss rate of around -25bb/100:



The most optimistic scenario, which in my opinion is impossible, is that he was a breakeven player:



In reality, he won 70 buy-ins, so all statisticians draw your own conclusions.

Nassim Talab love to s.hit on people(always correctly) when they can't see clearly the problem but instead try to use "logic". Look at the https://en.wikipedia.org/wiki/Ludic_fallacy . Be Fat Tony and not Dr. John

I am currently analyzing the hands of suspicious players whom I have already highlighted in another thread:

https://forumserver.twoplustwo.com/s...postcount=5239

I will post the results.

Maybe the software was written in India or somewhere like that and wasn't properly tested / audited. Or maybe the software developers or someone higher up put the bug in on purpose.

In the HHs shown ITT, MoneyTaker69 was ahead on every street with a pair of 8s versus a pair of 7s or pocket 2s versus like jack high. It would be interesting to look at more HHs and see if he knew how to play at all. Also, see if he folded big hands when behind.

The graphs go more or less straight up. If it were due to random fluctuation, that would not be the case, and the chances of straight up graphs like that should be even more astronomically small than 90 BBs/ 100.

So he won 90 BBs/100 with a 53% VPIP? Sort of like Postle, playing winning with junk.

id like a statistician to tell me how many hands need to be played in the next billion years in order for a 50vpip that loses at -50bb/100 to achieve a 90bb winrate over 15,000 hands assuming a stdev of 100

pretty please

I've already posted many juicy hand histories in another thread (read from post #5237 onward):

https://forumserver.twoplustwo.com/2.../index210.html

The MoneyTaker69 case has already been clarified. I doubt he was the only one, so we should focus on other potential cheaters. GG has lost a lot of credibility and player trust, so I believe it is crucial for them to join this debate if they want to regain any trust from all those who play on their platform.

I really wonder if there are any withdrawal difficulties the next few days/weeks. I think as players the only solution is to withdraw funds to signal the site something like this is not ok. I am only a low stakes player but i did yesterday and i wonder how many other did too... i hope this wasnt the first domino

where is MM when you need him

ahhahahahah

I did the same this morning. Mainly a stars player but put some money on a month ago, took it all out again now and won't play there.

+1

Definitely something going on there. I have sent an report to GG 2 weeks ago about it but wasn't expecting that something will come out of it. And now this... I have big sample at plo 200 ( and before that at 500 too ) and I have pictures of extremely weird hands against guys with extremely weird stats - like 40% 3bet for 500+ hands and those are not as uncommon as one might think. Basically every hand in this pool is 3bet and 2/3 of them multyway.

I suspect a collusion as I know there is a software for that. And that was the reason for me to move from normal tables to Rush as I was sure I was getting colluded there. But feels the same think at Rush now. It feels impossible to win a hand at showdown in a pool that is full of fish allegedly.

I am interested to see if you share the same concerns or something else ?

I was here

How about some 10NL players?

edited: nvmind I didnt look at all of them

Besides couple with more hands, which probably look fine, all seem to rather be blatant chip dumpings, which should have been discovered after first session/during the first session by security.