Tuesday, April 10, 2007

1000 hands from the button

The easter holidays have just ended. I was going to go skiing in the countryside, but because of the weather (and me being too lazy) there's been more poker than skiing. I played a lot of poker hands from the button's point of view in a hypothetical game with three decent players (Yeah. I was the dealer and controlled all three players). I was playing pretty close to what Wiesenberg suggests in his articles in Card Player. The first 500 hands, I raised with pairs of 10's and better. The next 500 hands, I included 9's and 8's. I also raised with any flush draw and open ended straight draws containing a queen.

I was finished with the 1000 hands after a little more than 12 hours of playing. I confess that I probably used less than 10 seconds to shuffle between each deal, but I don't think that affected the randomness too much. I also didn't bother to continue with the hands where the button folded.

I think I had straight flush draws three or four times, but didn't hit on any of those. The best hand I saw was quad 10's in one of the blinds. He had quads before the draw.

The biggest confrontation was when the big blind was dealt tens full of aces. The button raised with eights-up and called a reraise. He got eights full after the draw, and raised the bet from the big blind. The big blind reraised. At that point the button figured that he was probably behind, so he just called the reraise.

I tried to analyze the most thought-provoking hand in my last post (three-way action). The most interesting coincidence happened in a hand where the button raised with a flush draw. Both blinds had the same garbage hand, and the cards were arranged identically (4537A). Did I say garbage? That's a double-gutter...

Anyway, I just calculated that this should occur about once in 757500 hands (I gave the small blind a hand with 5 cards of different ranks and calculated how often the big blind would have the same ranks in the same order). That means that the probability of this happening is slightly smaller than the probability of being dealt a royal flush. I hope my calculations are correct. (47*46*45*44*43/243 = 757 500,740740...)


Online draw strategy by Michael Wiesenberg can be found here:

http://www.cardplayer.com/author/article/all/14/6406

If the above address doesn't work, you can look for the article manually at www.cardplayer.com

Edited: The sample size here is too small to draw any conclusion about what hands to raise from the button three-handed. The button made 202 times the big blind with tens as the minimum raising hand, and 234 big blinds with eights as the minimum raising hand. My own thoughts before I started the experiment was that it's a little too tight with tens as minimum raising hand and a little too loose with fives. I think I'd raise with a pair of sevens with a couple of big sidecards as the minimum hand. If I knew that the opponents were too tight, I might raise with something as weak as fives. If they were too loose, I think I might fold some pairs of sevens and eights (especially against extremely aggressive opponents).

No comments: