Search found 89 matches

by FlavusSnow
Sun Oct 17, 2010 5:51 pm
Forum: Computer Chess Club: Tournaments and Matches
Topic: Fixed Depth Tournament
Replies: 10
Views: 679

Re: Fixed Depth Tournament

I guess the same could be said about a fixed node count... depends on what the author of the engine considers a node.
by FlavusSnow
Sun Oct 17, 2010 4:02 pm
Forum: Computer Chess Club: Tournaments and Matches
Topic: Fixed Depth Tournament
Replies: 10
Views: 679

Fixed Depth Tournament

Would there be any reason to do a tournament with a fixed search depth and unlimited time? My thought is that a fixed depth rating list would indicate more accurate evaluation functions. The problem is that sometimes speed compensates for accuracy and those two things work in concert. Anyone else ha...
by FlavusSnow
Mon May 10, 2010 6:56 pm
Forum: Computer Chess Club: General Topics
Topic: Komodo 1.0
Replies: 4
Views: 1099

Re: Komodo 1.0

Ah. I didn't realize Komodo was only 1 thread.
by FlavusSnow
Mon May 10, 2010 6:30 pm
Forum: Computer Chess Club: General Topics
Topic: Komodo 1.0
Replies: 4
Views: 1099

Komodo 1.0

Has anyone else noticed that Komodo appears to be better at long time controls than short ones? see IPON list vs. WBEC Ridderkerk. IPON is based on short games and Komodo is 100 ELO weaker than Stockfish, but on WBEC Ridderkerk (standard time controls) Komodo is performing better than Stockfish in t...
by FlavusSnow
Thu May 06, 2010 10:13 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Stockfish - material balance/imbalance evaluation
Replies: 88
Views: 22320

Re: Stockfish - material balance/imbalance evaluation

I'm curious if anyone has tried evaluating material balance indirectly. An example would be by counting the number of moves each piece has and multiplying it by a coefficient specific to the type of piece. In the end, the side with the most pieces will likely have a higher score, but in the case tha...
by FlavusSnow
Sat Apr 10, 2010 3:42 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Tiered Search and MTD(f)
Replies: 6
Views: 878

Re: Tiered Search and MTD(f)

Yes, only at the tip nodes. I figure with iterative deepening it might be doing the same work as what you described, though, depending on the pruning techniques. And yeah, if you've filled your hash table it'd be slow. I would probably keep the MTD(f) hash isolated from whatever hash scheme was used...
by FlavusSnow
Sat Apr 10, 2010 1:50 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Tiered Search and MTD(f)
Replies: 6
Views: 878

Re: Tiered Search and MTD(f)

yeah, and I think what you've described is similar in concept. The only reason I had thought of MTD(f) was that it visits fewer nodes, so your penalty for opening the window on these nodes should be less. In the end, though, I tend to agree that maybe there is not much benefit in doing this. I'd lik...
by FlavusSnow
Sat Apr 10, 2010 12:39 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Tiered Search and MTD(f)
Replies: 6
Views: 878

Re: Tiered Search and MTD(f)

yes, but to a relatively shallow depth. I agree its expensive, but the question is if its worth the expense. I couldn't imagine a 2 core machine doing very well with it, but maybe a 32 core machine would do better this way than by extending a conventional search by one ply
by FlavusSnow
Sat Apr 10, 2010 2:30 am
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Tiered Search and MTD(f)
Replies: 6
Views: 878

Tiered Search and MTD(f)

Has anyone tried a tiered search? It seems to me that a search using MTD(f) as its first tier, and PVS as the second tier might be beneficial. I understand that this could add inefficiencies into the search, because you're applying the static evaluation to nodes that you might otherwise not; however...