I checked Mypokerintel:


So either it's not that rare to run at insane BB/100 over 10k hands and more or this site can not be trusted for accurate tracking

el oh el

the data is a lot more than not perfect. i'm fairly certain your 10 stddev claim is way off, though i don't have the math expertise to make a better one. for one thing, the stddev for normal players is going be dramatically lower than for people who play like nionio. for another, outcomes are not normally distributed. in relatively normal cases this can be fairly safely ignored, but when trying to compute the likelihood of an extreme outlier outcome, it becomes much more relevant. sorry to rip on your analysis without providing an alternative, but i'm just not qualified. brilliant statistician, please come save us!

What stakes do Toesanders and Legiao play at?

wtf, someone ran at 700BB/100 for 50k hands? that can't be right.

50% W$WSF with a 60%VPIP is winning a whole lot more hands uncontested than 53% W$WSF with a 27%VPIP. I don't know if its convincing evidence, but its pretty compellingly out of the ordinary. Working on the assumption that this guy is pretty smart and trying not to get caught, tracking his own stats and keeping his W$WSF < 50 seems reasonable.

I'm far from completely sold on this guy being a cheat, but I'm coming around from very skeptical to pretty suspicious.

This leader board can be somewhat deceiving sometimes. At UB there are 3 small NL tables (.05-.1cNL and .01-.02cNL) where there is no cap on the buy in. If two players were to buy in deep(say $100 at .01-.02) and and one lost to the other their BB/100 would be way off.
I don't believe this is the case with NioNio because he always played high limits winning at that pace. But it is all more reason to get all the hands that he played. It would be great if mypokerintel would give these up.

I am nearly 100% sure that I have run at 60bb/100 (30PTBB/100) for a stretch of 3 - 5 thousand hands when I was running white-hot. In fact, checking my PT confirms that. (Granted that's just cherry-picking the stretch of hands when i was running my absolute hottest.) But:
- I don't play 60 VPIP
- I wasn't playing vs. the best HSNL players
- I didn't 'disappear' right around when the AP scandal broke

Those factors combine to make this situation suspicious, obv, but high winrate alone isn't proof-positive.

***Just had an important thought: doesn't PokerEV allow you to evaluate street-by-street equity? e.g. just "when" the money goes in? Wouldn't a careful analysis of those stats also provide a very important part of the puzzle? Or at least supplement the "and he wasn't even running hot..." train of thought? Meh, I guess that's what the river-aggression stat does, right, which people have already looked at...? I guess what I'm thinking is that while we might not have the INFINITE river aggression that sealed DoubleDrag's fate, wouldn't it be possible to simply look at the percentage of money that went in on the river, and compare it to the percentage of such money for a "legitimate" player who's simply running white-hot?
>> Take my 3K stretch for example where I ran at around 55BB/100...I wasn't outplaying people per se -- I was just on the fortunate end of a ton of cooler hands where I'd flop bottom set to his top two, or AA > KK -- e.g. most of the money would go in on preflop or on the flop...
Cliff's notes: People have run *that* hot before over 3K stretches. But couldn't we learn something important by comparing NioNio's hot run to that of a player we know isn't cheating? Wouldn't the pattern of street-by-street aggression / equity look dramatically different?

Check his stats. He won at 60PTBB/100.

even if this guy is cheating, wtf does it matter? nothing will change, look at AP. the only thing that can happen is warning people to not play there.

It's obvious what's going on. Nothing will ever happen though.

Notably absent from this thread is Nat...
Surely in your trip down to Costa Rica (or your investigation beforehand) you must have met some contacts who you could pump for info on this, no? If you could figure out either NioNio's real name or IP address, that would be a pretty good start...

There is a reason for this.. MyPokerIntel creates a bb/100 rating over all your hands, regardless of stakes. The reason some people have ridiculous bb/100 winrates over a lot of hands is because of the "no max" tables, where you can buy in for well over 100bb.

MyPokerIntel sees you buy in, and sees you rebuy, and sees what you leave the table with. This is how it computes your stats. However, when you buy into a 10nl table with 25k, it cannot see a buy in past 100bb. So it thinks you bought in for 10 dollars, and once you leave with your 25k, it sees you as winning 24990 at a 10nl table. Thus the ridiculous winrate.

Toesanders told me on AIM that he was fooling around with a friend at a no max table and that's why his winrate is so huge. However he is one of the better players on UB anyway.

Well I've seen fish with high vpips have W$SF's of 50 ish over good samples. GimmeNutz had one of 47-48 or something IIRC and this was a significant enoguh sample. I'd say there is a pretty damn good chance this guy is cheating, but considering how hard it was to get any kind of acknowledgement of foul play from AP when it was pretty obvious makes me not so sure anything is going to be done about this.

I also think if he was smart enough to keep his W$SF a bit lower he would try harder to not make very suspicious river checks like the two pair vs rivered top set hand. I remember that DOUBLEDRAG had a W$SF of slightly less than 70 but the sample was only several hundred hands.

Like I said, I believe there is a very good chance this guy is cheating, but I don't think this evidence alone is going to sell to either UB (who would deny probably anything that happened no matter how ridiculous) or the average person.

Stevire=NioNio?

Lol. I am happy to arbitrarily exclude the top two results on that mypokerintel chart simply because "it is not possible to run at 700bb/100 over 100k hands."

To put that in perspective, that's around 40% more than the unabashed Absolute Poker cheater who was doing everything he could to win as fast as possible.

The subsequent explanation by Scansion makes sense to me.


If you exclude those two self-evidently absurd figures, our cheater in this instance is winning five times faster than the next best guy.

You may be right. In which case, if, instead of being 10 standard deviations above the mean, the suspicious player is only 8 standard deviations, it is only a one in many quadrillions occurence. Whether it is 1 in 10^15 or 1 in 10^21 is, to be honest, pretty marginal.

Other people saying that "lots of people have run hot" and so on (and this isn't directed at you, Ike) are, I think, failing to recognise just how hot this player ran.

Sure, swongs are one thing, but this is truly extraordinary. Either this player is one of the best players in the world and has run hotter than anyone else on the planet, or their doing something dodgy.

Either way, UB should investigate it.

Sure. If you'd won 2½ times more money over the same stretch, you would be comparable.

Out of the millions of hands that you, a truly good poker player has players has played, at lower stakes and against lesser opposition, your win rate (60bb/100) is not even close to the suspicious player's win rate (150bb/100).

Yep - which is why UB needs to properly investigate it.

No, I don't think they have.

I'm obviously open to changing my mind - but I don't think this is right.

Yes, I think this is good.

He does know of the thread, I understand.

wow, mypokerintel seems like a terrible site. 50/month for misinformation. buncha scam artists imo.

I find it scary that it's only the stupid ones that are being caught (and even then people still remain unconvinced)...

It seems to me that this is further reason to play at a large & reputable site (eg, Stars, Party or FTP) if you're playing for significant stakes.

maybe not to the insane degree of AP, but those stats show a ridiculously obvious hold card reader. Again, someone who didn't even try to hide his cheating... running through the stats:

Any VPIP over 40-45% at 6max will be unprofittable (30% is avg). He's at 60%. AF is typically similar on most streets for most players. His is <2 on avg for flop&turn, but on the river it's >6 (when he knows 100% whether he's won or not). A bb/100 of 15 would make you the best pro at those stakes, his is 80. As for the sample hands, gross value bets are the only evidence you'll be able to see as there is no "master hand history" like with potripper... there must be showdowns. The stats show 100% cheating... if there was a 20000+ hand sample size. With a small sample size it may have just been a gross run by a poor unknown player, but 3000 hands is by no means a small sample size either.

And beyond the stats, the anecdotal evidence is not to be ignored. Never getting action on your hands, but always seeming to get raised or floated when your bluff. That doesn't happen very often.

fyp

1) you're talking about pokertracker big bets/100.

2) i assume that no one has ever run at 15ptbb/100 at 50/100 over their (significant) lifetime sample size.

Ya sorry, was referring to PT's big blinds/100, not big bets/100

A PTBB/100 is a big bet/100. That is equivalent to 2x big blinds/100.


I don't mean to be a nit about this, it's just that these numbers are so large that multiplying them by two makes an issue.

In PT you can set bb = big blind for NL... anywho 80bb/100 is absurdly high by either definition.