Quote:
Originally Posted by MyrnaFTW
lol gg poker.
good thing for the forums as its still a place to put this out into the poker world when someone doesn't have long reach on twitter
and lol at MM saying he has a method to scrape for this as a way to hawk more books.
Identifying a Cheater
I have now come to a topic which I’m a little reluctant to write about. The reason for this is that writing about it can help a potential cheater defeat this method of detection. This assumes he’s aware of this chapter and understands what is being said. And in the past, whenever I told anyone about this approach, my request was always not to repeat it.
On the other hand, today there always seems to be a new and current cheating accusation. Most of these, in my opinion, are probably just “sour grapes” because someone got lucky. But we have seen in the past a few of these possible cheating events where it was clear that real cheating did take place, and it was quite costly to the honest players. Also, my best guess is that this pattern of constant cheating accusations will continue in the future.
So, the question arises, is there a good way to identify a likely cheater? And the answer, in my opinion, is yes. Let’s see what it is.
To start, this book has been constantly talking about expectation, which in this case we’ll refer to as the win rate, and the short-term luck factor, which most players today refer to as variance. But for this discussion, we’ll talk about the short-term luck factor in terms of the standard deviation (which, again, is the square root of the variance).
But for this discussion, we’ll be looking at the relationship between the win rate and standard deviation. That’s where the key information lies.
This leads to the following question. What will happen to these two parameters, win rate and standard deviation, if the cards were marked and someone could accurately read these marks? The first is that their win rate would be very high. I think the reason for this is obvious.
But what about the standard deviation? The answer here is that the standard deviation would be very low, especially when compared to what we normally see relative to the expectation (in the short run). However, it won’t go to zero since there will still be some luck in the game since an opponent can always draw out. But when this happens, the cheater can always save the bet on the end, and in no-limit hold ’em, this can be a large amount.
Now what would happen to an honest player, particularly one who plays too many hands and goes too far with their hands, if they have a streak where they do exceptionally well. The answer is that their win rate, as with the cheater above, will also be very high, but the difference is that their standard deviation will be high as well. So, the ratio of their win rate to their standard deviation won’t look (too much) out of balance.
Today, when I see accusations of possible cheating, if the standard deviation is mentioned, something like, “His results are several standard deviations above where they should be, hence, this must be a cheater” is often stated. But this is wrong.
And the reason it’s wrong is that with lots of people playing poker, there must be someone, in some period of time, who was the luckiest. And you can expect this player to not only be a big winner, but to have results, from a standard deviation perspective, that is way above the mean for a typical player. But this does not mean he was cheating.
In fact, in my opinion, if their expectation was very high and their standard deviation was low, this, in itself, does not mean they were cheating. That’s because an excellent player, who’s also an outstanding hand reader, should have this profile, at least to some degree. But it does make them, again in my opinion, a possible candidate for cheating if the results are extreme.
Now to clearly show that they were cheating, I think you need to look at specific results of their play. This is something an Internet poker site can do or perhaps something that could be done if the player in question was a longtime participant in a live poker show where video of all their plays is available.
Specifically, look to see if they were lucky. That is, in the run of great results, did they flop sets at a higher rate than expected and/or did they make more than a fair share of flush/straight draws. And on an Internet site, they could look at things like did their flop top pair hold up at a higher rate than what it does for a typical player and did their big pair (under aces) win at a higher rate when an ace flops than it does for a typical player.
If the answer to the above (and similar) questions is no, that is they don’t appear to be lucky, and the ratio of their win rate to their standard deviation is way off the charts, there’s a good chance you’re looking at someone who probably should not be allowed to play.