Hand by Hand Analysis

$1000 + $50 Buy-In No Limit Holdem Tournament

POTRIPPER's first 99 hands played, with hole cards shown for all players.

Analysis

I don't believe that any single hand below can be taken as proof of cheating. Any single action has plausible explanations; good read, good laydown, tight play, hyper-aggressive play, a donkey play, dumb luck, even a misclick or misread. However, when viewed together, the proof is indisputable.

If I had to pick just a couple hands out-of-context, I'd pick #86, where he folds KQ in late position on an unraised pot, big-blind happens to have pocket Aces, and hand #89, where he makes an all-in bluff on the river with Jack high.

If that doesn't convince you (and it shouldn't), then let's put his play in context. He raised or re-raised preflop 55 of 97 hands, including any hand dealt (34o, 35o, 56o, 62o, 72o) in any position. Put the KQ fold in that context.

He willingly folded only 9 other hands preflop. Hand #51 was only the second hand folded preflop in 49 consecutive hands. QQ had made a small raise. The first hand folded preflop was a check/fold against KK in hand #17. He also gave up hand #52 for two consecutive preflop folds, the player behind him had KK. He continues to demonstrate an uncanny ability to fold preflop only against strong hands: #57 QQ, #60 JJ, #63 AA, #74 AKs, #78 JJ, #86 AA. Note, none of these folds were against sizable raises, many pots were unraised or not even opened.

Despite playing nearly every other hand, he never failed to quickly get out of trouble when his opponent made a hand. He folds after the flop to top pair or better with literally 100% accuracy, 14 of 14 times (#13 (two pair), #16 (set), #19 (two pair), #23 (top pair), #26 (top pair), #28 (top pair), #35 (top pair), #40 (top pair), #43 (top pair), #61 (top pair), #70 (top pair), #75 (top pair), #76 (two pair), #82 (set)).

Finally, if his opponent did not hit a strong hand, he rarely fails to bluff, including some ridiculously large overbets when he's behind but opponent has weak hand. #9 (ace high against middle pair), #27 (ace high against bottom pair), #29 (ace high against underpair), #34 (queen high against king high), #36 (jack high against middle pair), #37 (middle pair against king high), #45 (queen high against ace high), #58 (ten high against 2nd pair), #79 (nine kicker against bottom pair), #81 (two undercards against underpair), #84 (4th pair against 2nd pair), #89 (jack high against king high), #91 (bottom pair against ace high), Only once did his opponent not fold (#58).

Despite his prolific bluffing ability, he never once fell for the bluff of another player. In hand #91 he calls a large bluff on the turn with only bottom pair against a paired board. On the very next hand (#92) he calls an all-in bluff on the river against a paired and suited board with only an Ace paired and no kicker.

I think the above better demonstrates the fact he had access to hole cards, and how it was used to his advantage; than just the improbability of someone happening to get this lucky in poker. However, Michael Shackeford (The Wizard of Odds) has calculated the probability as one in 1.88 × 10^44. "It would be easier to buy a 6/49 lottery ticket in six different states, and hit the jackpot all six times."

Conclusion

I am convinced, beyond all reasonable doubt, he had access to the other players' hole cards.

I'm also 100% certain this is not an RNG flaw (like the PlanetPoker debacle in 1999). Why? Two reason, first such a flaw would require input of the player's hole cards and flop cards before the deck could be singled-out from the possibles; just knowledge of two cards is insufficient. In the above case the player definitely seems to know all the hole cards before the flop. Secondly, this flaw would usually provide knowledge of the turn and river cards. Twice our player was demolished by river cards (hand #54 and hand #73).

What that leaves is "exposure" (info is leaked to the poker client, perhaps encrypted), "monitoring" (real-time decryption/analysis of all traffic in front of AP's server -- least likely possibility by far), or "intrusion" (insider or backdoor access to AP database). And for reasons I will outline, this requires either a conspiracy well beyond the scope of a rogue programmer, or total incompetence and negligence on behalf of their system security and design.

"We have researched these hands as well as numerous other hands involving these players and have found that the pattern of play of the players did not lend credibility to the claim that the players knew the cards of other players at the table" — Official Statement from Absolute Poker (09/21/2007)