Poker-AI.org

Hi guys. I just joined here. I'm currently creating my first pokerbot for my Masters dissertation project.
I'm currently trying out the Klaatu hand evaluator since I figure there is no point reinventing that particular wheel given the good work that has been put in on it. However, I'm wondering why the top hand, a royal flush, gets a different ranking depending on the number of cards evaluated.
5 cards - 7461
6 cards - 6074
7 cards - 4823
The worst possible hand always evaluates to 0. (23457, 234578, 2345789 respectively without flush).
Can anybody explain why that would be?

Also, I'm thinking that rather than just getting a current hand ranking, it would be useful to know the best possible hand ranking on any given flop, turn or river. In that way I could assess hand strength relative to the best hand possible, such that the nuts would come out at 100% hand strength. Does anybody know if any work has been done on anything like that?

Thanks
CaptnCanary

Welcome.

The best 5 out of 2345789 no flush beats the best 5 out of 234578 no flush. You say Klaatu ranks these the same at 0, so if there are no gaps in the ranking, the top ranks must be different.

Many people build tables of results using evaluators: search for LUT. I have tables of strength for every hand v every hand for flop and turn.