Absolute Soulreading/Rigged thread #3

[Drag]

@ flight2q

The analysis that you provide is not of a statistical nature. There is no such thing as 'running over the table' when we are talking about statistics.

A STATISTICAL way of analysis should look something like:

80% VPIP player normally (from analysis of statistical evidence for all the known players) is going to have a negative winrate of less than -20PTBB/100. Now, take a very high standard deviation of 100PTBB/100 (I know only of people with SD = 50-60 PTBB/100). The probability of winning 200PTBB/100 over 300 hands with -20PTBB/100 mean and SD=100PTBB/100 is less than 0.1%. (I guess we can do the exact math in this case.)
Add a similar stats for limit poker, i.e. too high winrate for that style over a large sample. Add evidence for tournaments: a probability to win a tournament with 100 entrants is about 1/100, a probability for a bad player playing 80VPIP style to win a tournament is much less than 1/1000.

Now you have to multiply probabilities: taking like 10^-3 for NL, 10^-3 for Limit; 10^-3 for tournaments, and you get 10^-9 as a VERY CONSERVATIVE estimation that a player in question got lucky. When we increase the number of hands the probability drops even further.

Then you can add the evidence from individual hands, circumstantial evidence about chip dumping (which is not of a statistical nature BTW) to come to a very solid conclusion that cheating took place.