nobody gonna buy your book

ur also ignoring the fact that many 50 vpip whales even if they get premium JJ+, AK all day long and hit nuts all day long still cant win bc they play like well, whales

limp pre
flop set, min bet min bet 50x jam riv
xmin raise, check turn, half pot riv
3b min, half pot, min bet, check riv

etc etc etc

you still ignoring his name I see

also I see no math

think its safe to wrap it up with this you clearly don't want the statisticians on 2+2 to judge you publicly for dodgy calcs

I don’t care about his name. I’m talking about a statistical approach that can help with identifying a possible cheater which can be applied in other situations as well. Why is this upsetting to you?

MN

youre the one suddenly signing your name wrong , maybe its you who is upset

you cant even toss out a range of deviation for a 50vpip who loses at -25bb/100 to -50bb/100

no ballpark? literally nothing? 160? 200? 120-160? certainly youve done these calcs before, no? you should have an educated ballpark guess by this point.

what I see in my friends 100million hand database is that after we filtered for 50 vpips and 15000 hand samples the highest winrate out of 100 players is -4bb/100

so how do you get 90bb? lol

we all know by now you dont actually think he isnt blatantly cheating and you cant do the math bc you know it will prove impossible

ur just shilling for ur book

Mason you can infer a player that wins 95% of their 200+ BB pots it’s going to have a relatively tiny stnd dev given their winrate. If you never lose a hand, it’s pretty hard to have a hi one

I’m sorry, but you don’t seem to understand much. In my long career, one thing that I learned is how counterintuitive things based on probability theory can be for some people.

To be specific, I don’t have an estimate for what this person’s standard deviation might be. But the poker site could certainly supply them. I’ve done these calculations for my own results and those of a few friends, but none of us had results like this.

MM

i dunno another guy who seemed pretty confident mentioned you can do the calcs on a napkin

I use a spread sheet program (which I programmed) to compute the maximum likelihood estimate for the standard deviation. However, once you have the standard deviation and know the sample win rate, the calculations are then simple.

MM

I wonder if there will end up to be a pattern in the screen names like in the past. I would be slightly more suspicious of names that have numbers at the end. People sub consciously often make names with similiar syllables too. Just a suggestion to look for. That is of course assuming only 1 person was cheating which I highly doubt

oh missed this comment that got 16 upvotes

but I eventually figured it out for myself. Now I know for the future lol

i truly cannot decide what is a worse look:

"we'll allow players to soup up their own rogue DIY client allowing them to block critical safety patches we push out"

or

"i know how to tell if they were cheating but i can't reveal that information because that could aid the cheaters but buy my book if you want to know how to tell if they were cheating"

First, I did post the chapter that I was reluctant to post. Second, as I said, when this book is available, I plan to make free kindle copies available for everyone.

MM

Mason has always been a moron but this is just embarrassing at this point. I have a friend who works for the census bureau like Mason likes to brag about. He had brain cancer and had a large part of his brain removed to try and get the cancer.

I lol'd at this

The people arguing nothing is going on are old school live players who probably know little about online poker, Mason and limon. The reason they are getting insulted is their arguments are so ridiculous. It isn't just win rate. There are HHs like calling down for 120xBB with pocket 2s unimproved and similar. The win rates are extremely consistent, which have to be at least through skill and not chance, etc.

In a way this is better than UB. So far no evidence the people doing it were insiders, like that guy who was worth his weight in silver. Unbelievable if it was software bugs and they didn't do anything about massive cheating allowed by them.

I don't worry too much about online poker, as it is difficult to play on these major sites from the US, you need to know GTO really well and use stats on players for online cash games, and the sites are flooded with bots and RTA cheaters. So no point bothering playing online now. Maybe others are still doing well with it.

Do you suspect that a large part of Mason's brain was also removed and/or no longer functioning correctly?

good god mason, you are washed. you guys had like a 40 year career out of one or two books in the 70s, time to hang it up.

go play on prime dope and see what kind of SD and winrates you need to achieve a 90/100 run over 9000 hands. there's more proof of god.

Omaha games seem the same. 50/15-ish people that do very well, also same pattern of donking 1bb in a spot that makes absolutely no sense to do it even for rec's.
I've seen a guy that is putting insane volume on R&C 100 and 200 (he playes 8 tables rush simultaneously and almost always wins the LB) min3bet with kjt7 single suit(very reg-like behaviour) then a "fish" 4bets we get it all in with aces and he is good , because the "fish" also has aces. All in all very very suspicious and dodgy pool tendencies in games that look extremely soft. I saw EddieKing mention 22 cheating accounts, I am willing to bet he had only holdem in mind, there are just as much in the Omaha Pool.

No-one making the case for sporadic access to villain's equity being, essentially, no big deal?

lol. I don't play online cash, particularly now, but I think there are regs making 3/100 at lowish stakes. 10+/100 is considered crushing the games. You couldn't achieve anything like 90/100 even in loose live games, but online, with multitabling, you are playing maybe 600 hands/hour.

2+2 put out a number of good books, particularly pre-boom, plus these forums. Some of the comments go too far. However, Mason and limon are obviously totally clueless about online poker.

I agree that GG should be able to work this out/ automate the process. You can look at the SD BB/100 for the population and calculate the probability that a player got incredibly lucky and also happened to have one of the highest SD bb/100 in the population. It is astronomically unlikely. If you assume a typical SD bb/100 for this player, and assume they are the biggest winner on the entire site, this result is 7 SDs (1/1 trillion chance) above the mean. If you assume this player is breakeven (still very generous), and assume they have 2x the SD bb/100 of a typical player, they are about 6 SDs (1/1 billion chance) above the mean. Both of these are well above the level of "reasonable outlier" when you consider the number of players who are putting in 9k hands in a given month at these stakes on this site.

Your point about selecting outliers is well-taken. The example that comes to mind is the woman in the UK who had two children die due to sudden infant death syndrome (SIDS). While it is important to be careful in settings such as these, mindful calculation would lead you to the same conclusion: this guy is cheating. WRT your other comment, I hold a PhD in statistics.

The statistics lessons are interesting (really), but that debate seems to miss the larger point of this thread. GG confirmed the hack happened and gave details. The next question should be whether this was a solo actor, which seems highly unlikely, but in an ideal world the site would be sophisticated enough on the tech side of its investigation to determine that and then do something about it. People here don't have much faith in GG's ability (or even desire?) to do this and that is the major concern.

And the primary method for discovering if he was cheating and/or was a solo actor shouldn't be hypothetical standard deviations and the rest. The existence of an extreme outlier should be the trigger for an investigation, not the primary evidence. Now that we have one confirmed cheater, the mound of player data the site has should provide them with the indicators for who else to investigate, but I would hope they have the capability to then identify technical evidence that X/Y/Z cheated and someone else didn't.

I know nothing about the tech involved, but it seems to me the job of a well-run site is to have the tech competence to take the suspicious cases and identify if the now-known hack was being used by others. Honest question for computer-savvy people, is it reasonable to expect GG to have that evidence (i.e. through records of how the site was being accessed, can we expect them to be able to pinpoint someone having used this specific vulnerability in the past?)

I didn't post the nl500 results because someone else already did earlier in the thread.

Above is an image of the GG poker microstakes cash game leaderboard over the past 60 days, some of these winrates are mathematically impossible, +37 buyins in 125 hands is laughably impossible

Suspicious results:
#4
#19
#36

Not to mention some of the other absurd winrates over the larger samples like 35-40bb/100 seem pretty insane as well.

Hi tmanto:

My comments are below.

I would think that if someone was playing honestly and they had during some period of time an incredibly high win rate, you would also see an incredibly high standard deviation. This is what I would expect from a bad playing wild player who got very lucky.

Okay. But I suspect that an extremely wild player would have a standard deviation more than 2x the SD bb/100 of a typical player. This is my opinion but it is based on a lot of years at the table (and tracking these type of statistics. However, I don't have to be right.

Let me be clear, I never said this person wasn't cheating. All I said was that there's a statistical approach that can lead us towards a definitive answer.

I have a Masters in math. But when a graduate student, I took a lot of extra statistics courses and have almost enough credits for a Masters in statistics as well. And in my working 11 year career, I always worked in the field of mathematical statistics.

Mason

While what you say is certainly correct, my initial comments came. I believe, before GG made their statement. At least my comments were before I knew about their statement.

I don't quite buy this argument. Poker sites. like live poker rooms, need to be squeaky clean. And to assure this, they should (behind the scenes) be looking at all their big winners through the type of statistical methods I have described, and when the ratio of their win rate divided by their standard deviation gets too high, they can then look at other things.

Mason

You said his results may not be a statistical anomaly before making 15 comments about how people should buy your book.

What are you even trying to argue now in your last 10 posts?

Claydol