I think the statistical evidence is actually some of the most damning evidence if it wouldn't sell well to a jury. If we are using Bayesian inference, the chance of him achieving his results by luck alone is much, much smaller than the chance of someone cheating at a poker game.

https://twitter.com/mitchellichtman/...93684610863104

We've covered this upthread, but the way to win this case is through all the evidence that doesn't have to do with winrates.

The basic storyline is this: guy starts playing on the stream and plays a relatively standard style of poker, including paying off with some second best hands in coolers, folding the best hand, and folding hands with a chance to win. Then he suddenly buries his phone in his crotch and, only on days when his buddy is in the control booth, he starts playing like he knows everyone's hands and amasses $300,000 in winnings doing it. Just as suddenly, his buddy leaves for Las Vegas and he can't do it anymore and changes his play style back to the way he did it before. His buddy comes back and he starts wearing a bulging hat which he grabs onto like an NFL quarterback and he plays like he knows everyone's cards again.

None of this is outside of the understanding of the jury, and all of it is supported by hard, irrefutable evidence.

Just remember, there were potripper deniers also

Invoking Bayes theroem is somewhat irrelevant. He's taking many unknowable statistics and pointlessly adding a bonus one (prevalence of cheating) to make a very impressive sounding number without explaining how he arrives at the inputs (this is the hard part). You can easily construct a confidence interval for a sample of normal random variables if we know the mean and the variance of those random variables. The problem is we don't know either, and there aren't really data anywhere to even begin to construct reasonable estimates of what they should be or how they could be adjusted to Postle specifically.

There probably is some kind of game theory/mathematical argument here, but it would be extremely complex and certainly not something you would squeeze in a couple of tweets. It's intuitively obvious to poker professionals that Postle's decisionmaking is overwhelming guided by his opponents' precise holdings, but constructing that argument is not easy.

There were also people that idolized Durrr, Good2cu, OmgClayAiken....and others for turning $50 in to literally millions in less than two years.

Or remember how strange the moves Isildur made were?

I 100% agree the moves look absurd. But so have been the moves of so many other players.

He is literally being filmed every hand and still can’t be caught. I particularly like the analysis of when he turns on god mode or plays without it.

All that said, I wouldn’t be interested in playing with him. Either he’s that good, or is cheating. Would just like some tangible proof other than an 8500 hand stretch of results.

Has anyone posted the actual dates that JFK was away? I haven't seen them.

He’s cheating

Zizek using Bonomo as an example of a historically clean player on a heater was an interesting example.

If that amateur made 125-250k+ in those 300 hours of play in low stakes games I'd say they were fine taking a shot for a year or two.

No. Statistically, just no. His VPIP and raise/fold vs. call frequency on river are not remotely in the realm of anything we've seen from a winning player, live or online, in any era, ever, since Potripper. Those are also very coincidentally the exact 2 stats most correlated with what someone who could see the hole cards would do.

Yeah, I dunno man... this is a tough call...

Please, guys, get your heads out of your asses. He is cheating. The evidence is irrefutable, it's only a matter of if you're capable of understanding it. All you're doing is displaying your ignorance.

Bayesian inference is irrelevant? It's a reasonable way to estimate the probability of him cheating. There is plenty of data from live poker players who keep track of their results, a lot of them friends of mine, that have showed me their winrate. There is plenty of data to estimate if Postle's winrate is plausible even for the best poker player in the world. It's been estimated by different statisticians and I've seen numbers anywhere from 6 to 12 standard deviations. Even a 7 SD is a an astronomically small probability of happening by chance. You'd have to seriously pulling the wool over your eyes if you thought it was more likely he was getting lucky than he was cheating.

There's a difference between being on a heater and playing perfect poker.

Something tells me that more than 90% of the people in this thread think he's cheating. He is so screwed in that lawsuit. His only hope is that 0% of the jury knows anything about poker and are too stupid to understand.

It's not that his moves are strange. It's that they are never wrong.

Isildur and others lost lots of big pots. I'm sure there are plenty of hands that can be shown where they tried bluffing vs. a very strong hand. That's true of any poker player who regularly bluffs -- except Postle. He only bluffs against air or weak hands. Once in a while a weak hand looks him up like when ace-high called him and was good, but that doesn't mean Mike was wrong to bluff -- he just ran into to someone that decided to hero call with a hand that would usually fold to a big bet.

How is it that he always knows not to bluff when he's up against a big hand? His opponents can't all be that transparent. That is the difference between someone who is cheating and someone who plays a new and different style than has never been seen before.

I mean, there are apparently people who have been members of 2+2 since 2008 who still aren't convinced he's cheating, so that worries me about the limits of human stupidity and ignorance.

If it did end up before a jury, given that there are already a low percentage of people who know much about poker, it should be easy for the defense to reject enough jurors to hit that 0%.

if anyone has a doubt about if mike is cheating:
i don't know if anyone posted it before but here is the top ten SHADY HANDS of mike postle:
https://www.youtube.com/watch?v=DdNbgiHZEfg&t=71s

Scary but true. A random person off the street would have no problem attributing the streak to luck.

Don't confuse Internet contrarianism with actual human reasoning.

It's much harder to be a contrarian for contrarian's sake on a jury. First of all, you have to get on it, and contrarianism tends to show up in voir dire, resulting in the prospective juror being excused. Second, you are required to deliberate. (If you refuse to deliberate you can be kicked off the jury.) You have to deal with 11 other people who basically are going to think you are an idiot, and treat you like one, if you are proposing crazy claims.

There's a reason why "Internet troll" is a thing whereas "jury troll" is nonexistent.

Again, no competent prosecutor is going to prosecute the case that way.

A random person on the street COULD definitely attribute the streak to luck, but not all the stuff with the crotch and the buddy and the way the play changed when he started winning, etc.

And there's a difference between playing perfect poker and playing with perfect information.

Let me be clear, I think there's zero chance Postle is innocent.

Bu I have some issues with the discussion of statistical evidence of cheating. A lot of people in this thread, on twitter, and in podcasts have referenced the probabilities of the win rates being possible. A lot of people have referenced that the likelihood is 1/(atoms in universe), etc.

But I'm not at all confident in the probabilities people are throwing around. Here's why.

These probability estimates assume a normal distribution of win rates (this is reasonable IMO) and require knowing the average win rate and the standard deviation of the win rates.

An average win rate of 0 is reasonable (it's actually less than 0 with rake, but let's ignore that.)

But we have no idea the correct standard deviation to use for these calculations in live poker. If we don't know the SD, all these probability estimates based upon the normal distribution are just complete guesses.

If someone can explain to me how they arrived at a reasonable standard deviation estimate of a win rate, I'd appreciate that. FWIW, I can't even figure out the formula people use. The PT4 forums reference work by Mason, but the links are dead.

I'm saying that despite you, and others, repeatedly claiming this, that comparable data doesn't exist anywhere. You have- at best- self-reported results from of a small sample of people who have a tax incentive to underreport, playing against completely different people with a completely different style. Bayesian inference only exacerbates the error of these assumptions by relying on yet another completely unknowable statistic (prevalence of cheating).

I assume that's directed at me. Say Veronica has a chihuahua named Mips, and she says Mips only barks at people who cheat at poker, and when Postle met Mips, Mips started barking. Then when someone with common sense points out how ridiculous that entire premise is, you come along and say "well if you don't trust Mips I guess you think he's not cheating! How stupid can you be?" It's embarrassing, really.

[delete]

There's no such thing as perfect poker unless you have perfect information.