Wednesday, March 02, 2011

A calculation mystery

So I played VeryJosie a quick series of four heads-up matches tonight, of which she won three. Nice job, congrats, etc. Not surprisingly, she quickly writes up a blog post about it, in which she claims that in one key hand she got her money in "goot," and that she was "very slightly ahead."


It's true that she was ahead in terms of having the best hand on the flop (when it all went in), but I was actually a favorite to win, with 15 outs twice. She had As-Kh versus my Qd-9d on a flop of 6s-4d-5d. In order to provide her proof of the math, I went to cardplayer.com and entered the cards into their odds calculator. I was puzzled by the results:



It says I'm only 39% to win there, which I was sure wasn't right.

For confirmation, I went to PokerStove, which provided what I was confident was the correct answer:



That's 55% me, 45% Josie.

But I'm left with this mystery: What's wrong with the CardPlayer odds calculator? I really can't figure out how it came up so far off.

7 comments:

Josie said...

When you factor in my good karma, I was waaaaay ahead. :)

THETA Poker said...

Simple answer: buggy software!

The odds the cardplayer.com calculator gave are the preflop odds. Adding the flop didn't do anything, but you can get it to show you the right answer by adding turn and river cards then deleting just the turn card.

Kahomono said...

The 55% assumes she can't improve. But if she hits an A or K on the turn your Q and 9 outs are gone.

Not sure that's enough to reduce your overall % that much, though....

THETA Poker said...

P.S. - The bug may well be browser-specific. The calculator didn't work at all for me in Safari, and I got the behavior I described in Firefox (both on Mac OS X).

Rakewell said...

Buggy software is right. I did it again and it came out right.

taximike44 said...

I've found this site to be less flash, but more substance as far as poker calculators.

http://twodimes.net/poker/

James Antill said...

The one I see used most often, on 2+2 is:

http://www.propokertools.com/simulations

...mainly due to the ability to input ranges easily.