Unethical: Communicating about hands while in progress, softplaying, and whipsawing.


Ethical: Combining bankrolls with people you trust to enjoy

lower variance, and a higher earn by being able to play higher.


I personally know of an example of the second sort on Paradise.(I'm not in it, a lone wolf by nature. I was on a blackjack team once. When it works its very very cool, when it doesn't it can get very very stessfull.) My point here is obviously that the mere existance of teams does not automatically mean cheating. And Ironically, this would be the sort of team that would declare itself openly without a sense of shame.


So, lets say you are a reasonably intelligent, fairly suspicious

poker player, sitting at a table and starting to suspect a couple players. How long would it take to find a few hands that

are suspicious? Is it that impossible for the focused player to

find some credible evidence? I've been thinking about it since

I started this thread, and I know its a significant handicap not to have the mucked hands.


But...


The most obvious, in which two players whipsaw and one folds quietly at the end, is pretty simple. I think I'd be almost satisfied with two occurences between the same two players in

a session, and three would clinch it. But NOT one though. Theres

just to much testorone pumped wackiness to get suspicious after

a single breakout of compulsive mouse clicking.


Softplaying might stand out too, if two normally aggressive players slow down after its head's up. This would need some statistical evidence, but still not too hard. The following metric might suffice: For every hand that players x or y saw the showdown, cull the hands that were heads up after the flop or turn and were multiway on the flop. Determine the percentages of

aggresive(bet, raise) and passive (check call) actions. Compare that to the percentages in the hands they are head up. If these

numbers are very far apart (say 80% aggressive normally, 5% aggessive with each other) this eveidence should be come statistically credible fairly quickly. A measurement error SD

would be ~.7/(num_actions)^.5 , so maybe a 40 hands with 4 or 5

x and y showdowns could give a 95% confidence level.


The hardest seem to be just using info adjust drawing odds info.

Folding AQ utg, beacuse of the AK three doors down. Not bothering to draw to split the pot with identical 1-card open straight draws. Knowing your flush draw is dead, or 2 of your outs are gone.


Bad news: I think its nearly impossible to catch this without revealed hands. Paradise claims to have a filter to flag the most obvious cases, like the AQ fold to the AK, which I believe,

but I'm sure there's a limit to its sensitivity. Because such information is relevant for relatively close decisions, it would take a long time for convergence of statistical evidence; certainly more time than we can reasonably expect from our diligent slueth.


Good news: Such information is relevant for relatively close decisions, so it doesnt affect your winrate in the 1st decimal place, in my opinion. Players that can accurately use this information are already not going to be dropping much for you to pick up anyway. And players that have the mental endurance to

accurately process and communicate this info, while still staying intently focused on the game for extended periods of time, may rare, perhaps rare to the point of fiction. I find it fully credible that people are out there, im'ing furiously back and forth, folding smugly while rubbing their hands together with cackling glee, but until I see some hands showing they kept their own house in order during all their scheming, I say I'll take the risk.


zooey


P.S. I got off track. The reason I worte is I'd like to discourage public accusations without _some_ evidence. I mean, if you suspect somebody then write to Paradise. They have kicked people off and will continue to do so. If you're going to accuse people in a public forum, credibility should demand that some small burden be bourn.