Quote:
Originally Posted by GaryTheGoat
scylla,
Great software!
I was working through the situation which you posted hereinabove:
http://forumserver.twoplustwo.com/sh...05#post3215705
I had previously "solved" the problem via beta5 in agreement with the hereinabove post. However, when I ran the same problem through beta6, I was presented with differing results from flop forward of BB. I even ran my b5 savefile through b6 and the results varied.
Which results are correct? Can the ev calculations could be exported to text file function like Lateski suggested here?
gTg
Hi Gary,
1.0.6 is correct.
1.0.5 had a horrendous bug in it that made >=tp accept ANY pocket pair.
Not only have I removed this bug, I have gone over this entire part of the code with a fine comb, removed a couple of other tiny bugs and am now confident that there are no more.
I have checked this by looking if for instance >=mp was always ahead of pp>3rd , and every other possible scenario. It was not a fun day
One of the other bugs was that a 2 pair hand like QJ on a 22QJA board was accepted as >=tp, which of course is not the case. Another small bug was that AA on a 22QJ board was considered worse than 2 pair. There were some others, but you get the picture.
Incidentally, these bugs were in the condition-check functions and NOT in the hand valuation functions that work out equity and best hand and such. The latter are 100% certain bugfree and match all documented statistics that I have been able to get my hands on.
Anyhow, it's irrelevant now. I don't see how the condition-checks could contain any more bugs, it borders on the impossible.
As for the file output, I'm more than willing to implement that, I just don't have a good idea what such output would look like. Could you give me an example of what it would look like for the examplefile in question?
Last edited by scylla; 03-24-2008 at 06:09 PM.