Search found 2248 matches

by Milos
Fri Jan 01, 2010 10:08 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Using Heinz in 2010 is not optimal
Replies: 6
Views: 1752

Using Heinz in 2010 is not optimal

Many are still following Heinz results. R=2 at root, R=3 at leaves. It's simply wrong! Few simple facts. The depths when this was established and tested were 8, 10 and 12. Typical depths today are 15-20. Even with 8, 10, 12, adaptive model R=2/3 worked the best. So, not only 2 or 3 was the best. Con...
by Milos
Wed Dec 30, 2009 11:33 pm
Forum: Computer Chess Club: General Topics
Topic: Has Stockfish 1.6s DC Finally Dethroned Rybka?!!
Replies: 29
Views: 6880

Re: Why only Rybka?

alpha123 wrote:
2: RobboLito 0.85g4 1 CPU 6.0/12 2-1-3 ····· 1-2-3 36.50
Where did you get g4?! :shock:
Misha forgot to mention, g4 was on 32bits, while others were 64. :D
by Milos
Wed Dec 30, 2009 11:15 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Smooth scaling -- an explanation
Replies: 39
Views: 9372

Re: Smooth scaling -- an explanation

I guess that each and every program will have its own optimal R factor, and also that the factor will change if they add LMR and it will change if they add other reductions. This is exactly the point. Neither, 2 nor 3 nor 4 is carved in stone. It has to be determined for each program what is the op...
by Milos
Wed Dec 30, 2009 11:06 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Smooth scaling -- an explanation
Replies: 39
Views: 9372

Re: Smooth scaling -- an explanation

Regarding your explanation I was just saying that when you write in your example depth = 5, actually it is depth = 2,5*OnePly so that when you say that with depth = 5 you have the same reduction by 4 of the original, this is not correct because the original with a depth = 5 (aka 2,5*OnePly) gets a ...
by Milos
Wed Dec 30, 2009 9:06 pm
Forum: Computer Chess Club: General Topics
Topic: Smooth scaling stockfish
Replies: 67
Views: 16583

Re: Little test with both stockfish 1.6s versions!

Delta is a huge value in Tord's program because it is in centipawns. I do not know that log base is is the right squeezer. I suggested cube root in my original post as an alternative. This factor also is a rich area for possible experimentation. I guess that when a final formula is found, it will b...
by Milos
Wed Dec 30, 2009 8:17 pm
Forum: Computer Chess Club: General Topics
Topic: Smooth scaling stockfish
Replies: 67
Views: 16583

Re: Little test with both stockfish 1.6s versions!

I suspect that it is too much. It also may be a good idea to adjust the constants. I actually just eyeballed the equation constants from a curve I drew by hand an calculated by inverting a small matrix. I guess that great improvement can be had by adjusting the formula in a rational and scientific ...
by Milos
Wed Dec 30, 2009 2:39 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Smooth scaling stockfish
Replies: 98
Views: 23202

Re: Smooth scaling stockfish

I want to congratulate to Dann. It's not 150elo of course, but the idea is brilliant. Just implemented it in Robbo, and it looks like 5% additionally.
Great job!
by Milos
Thu Dec 24, 2009 8:08 pm
Forum: Computer Chess Club: General Topics
Topic: Stockfish 1.6
Replies: 73
Views: 40638

Re: Stockfish 1.6

Why would anyone need to go to all that trouble? How hard would it be to take Rybka 2.2, for example, and have it identify itself as Rybka 3? You think ppl are that stupid that cannot see the difference between 7.5MB and 2.5MB file??? Or you think they cannot see the the difference in output? Oh, I...
by Milos
Thu Dec 24, 2009 8:01 pm
Forum: Computer Chess Club: General Topics
Topic: Stockfish 1.6
Replies: 73
Views: 40638

Re: Stockfish 1.6

Rybka's +45 Elo perf is in line with one extra core, so so far Robbo is performing exactly as Rybka 3 would with one core. And your error bar is, hm, only 60 Elo. Seams your conclusion is very reliable. On the other hand just to inform you, because you kind of pretend not to know. Going from 1 core...
by Milos
Thu Dec 24, 2009 6:03 pm
Forum: Computer Chess Club: General Topics
Topic: Stockfish 1.6
Replies: 73
Views: 40638

Re: Stockfish 1.6

Anybody who downloaded a pirated copy of Rybka 3 (and I'm not suggesting that anybody in this thread did) should realise that fake copies were deliberately put on the net. This could well explain some of the poor results posted around the forums. If anybody with a fake Rybka 3 tries out the same po...