Devil CardsDue to a string of horrible beats lately I’ve decided to take some time off and run some math on various poker hands/outcomes instead of playing. I’ll be using Maple 8 and I’m sure most of the methods will be plagarized from various poker sources and put into one document, which I will call:

Jake’s Review of Poker Math

Last night I was doing some thinking and ran some simulations with Wilson’s holdem software and came up with some numbers that basically matched what I thought. This study has little scientific basis - randomized 1,000,000 hands to showdown with no betting. Later on I include some theory which no doubt creates a poorly devised experiment-theory hybrid.

So here is the first in the series:

All In Preflop : 2-2 vs AK Showdown:

AK Details A K Wins 2 2 Wins Split
Unsuited 0 Suit Matchup 46.87% 52.84% 0.29%
Unsuited 1 Suit Matchup 47.27% 52.43% 0.30%
Unsuited 2 Suit Matchup 47.49% 52.17% 0.33%
Suited 0 Suit Matchup 49.90% 49.79% 0.32%
Suited 1 Suit Matchup 49.51% 50.16% 0.33%

Data Set Size = 1,000,000 hands

But really - to get a total average standoff percentage we need to realize that AK suited is much less likely than AK unsuited, or matching at least 1 suit between the two is more likely than not. So given that all cards have an equal chance of being dealt, we can write a few lists to discover the probabilities.

And Finally:

Averaging To Get a Final Winner:

AK Details Prob. Occur Y 2-2 Y A-K Y Split
U 0M 0.1 5.28% 4.69% 0.03%
U 1M 0.4 20.97% 18.91% 0.12%
U 2M 0.1 5.22% 4.75% 0.03%
S 0M 0.2 9.96% 9.98% 0.06%
S 1M 0.2 10.03% 9.90% 0.07%
Totals 1.0 51.46% 48.22% 0.31%

Y - product of the probability of occurrence and the win probability

So AK is a dog by 2.8% over 1,000,000 hands. Nothing really new there but interesting to see it evolve.

Just a few notes about split pots - these can only occur when a 5 cards appear on the board that beat both hands. So Straight, Some Full Houses, Straight Flush, Flush (Four of a kind is possible to split only when the overcard connectors do not contain an Ace). Of these, only the flushes depend on suit, and as the data indicates - as matching suits increase, so do the split pots (Albeit slightly, but due to more ‘other’ suited cards existing in the deck to give the non-matching flush).

Also consider that AK are the worst connecting overcards to have (by a slight amount) when going up against a low pocket, due mainly to straight possibilities. I’ll prove this to you in future posts, but if you think that you’re against low pockets - you’re better off playing J-10 and the like. However, if we showdown AK and J10 it would be quite apparent the good money was on big slick.

On to the next problem!