Theory: 9's are more likely than any other rank to fall on the river in Stud/8, because hands containing 9's don't tend to continue all the way to 7th street ( Read more... )

timprov noted that for the purposes of playing Stud, not all 52! deck orderings are distinct. One reason is because the order of your downcards doesn't matter. There are two ways the downcards can be arranged in the deck, per player, so the number of distinct Stud games is at most 52!/16 52!/(2^8 ( Read more... )

A while back, Scott Aaronson posted his slides from a "crazy idea session" asking why theoretical physicists get 8 billion dollar machines while theoretical computer scientists tend to be funded on a less generous basis. What might theoretical computer scientists find to do with $8B in hardware? His suggestion is to look for some lower bounds on ( Read more... )

Absolute's Razz hand evaluator is screwed up and awarded the pot to a pair of fours instead of a pair of aces, spawning this thread ( Read more... )

junk_game_junky gives a couple graphs comparing starting hand strength in Hold'em, second-best Hold'em, and true inversion Hold'em: heads-up and 5-handedI particularly like the illustration of the way true inversion Hold'em completely distorts high hand values. It's also interesting that 2nd-best flattens out hand values (making post-flop play particularly ( Read more... )

Chinese Poker played high only, or played with deuce to seven in the middle?

dmorr won a prop bet a BARGE by successfully predicting which his chosen expert "JP" would pick: 2-7. This led to some interesting discussion ( Read more... )

Checkers has been solved by Jonathan Schaeffer of the University of Alberta. Article to appear in Science in 2007. (I have on my shelf an earlier paper on this work, from 1996.)

He has provided an online version of the proof--- you can query the endgame database and step through the games that the proof examined ( Read more... )

Is Badugi unique among poker games in that your opponent can only be holding one of your possible outs to beat his hand? (At least when you are only drawing one.) In other games we can come up with scenarios where the number of possible outs held by an opponent are equal to the number of cards he holds ( Read more... )

So, I realized (duh) that if there are fractional outs you can't just plug them into the same formula you use for full outs... Fortunately in Badugi your opponent can have at most one of your outs, so it is easy to break the cases out separately and calculate using either N or N-1 outs, weighted appropriately ( Read more... )

My TD2-7 third-round solver has provided its first answer. This case is 7432 vs 7432 with 3.5 big bets in the pot:

***************** Iteration 273 ( Read more... )