Author: Dan Honeycutt
Date: 08:28:51 06/16/04
Go up one level in this thread
On June 15, 2004 at 23:05:26, Andrew Wagner wrote: >On June 15, 2004 at 23:04:43, Andrew Wagner wrote: > >>On June 15, 2004 at 22:55:39, Dan Honeycutt wrote: >> >>>On June 15, 2004 at 19:28:36, Andrew Wagner wrote: >>> >>>>I've heard it said before that it's not good to compare node counts between >>>>engines, and that node counts aren't a good indication of strength. So, I've >>>>been staying away from that a lot. >>>> >>>>The other day, I was chatting with my good friend Jaime Benito de Valle Ruiz. We >>>>see each other a lot on ICC and compare notes on our engines. We decided to play >>>>a fixed-depth game between our engines, to test eval strength. In the process of >>>>the game, he noticed that my node count was ridiculously higher than his. For >>>>example, in one position, where I was getting 277k nodes, he was getting like >>> >>>What kind of nodes, alpha-beta or q-search. See Bruce Moreland's discussion on >>>q-search explosion. >>> >>>Dan H. >>> >>> >>[snip] >> >>Alphabeta nodes. Currently about half ny total nodes are qnodes, which isn't >>bad. > >I should also note that I'm failing high on the first move about 80-90% of the >time, so my move ordering isn't too awful. Those don't sound too bad. If you get 277k where other engines get 11k to 50k that's like 10 times their nodes. The other suggestions in the thread are all good, but I doubt if any or all will get you a 10 fold reduction. Hate to say it Andrew, but I suspect a bug. My suggestion would be to single step about a 3 ply search, watching where the nodes come from. It's tedious but that has helped me root out search bugs. Dan H.
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.