Search found 1621 matches

by ernest
Sat Aug 08, 2020 3:09 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: New SF performs differently than SF NNUE
Replies: 6
Views: 565

Re: New SF performs differently than SF NNUE

Ovyron wrote:
Sat Aug 08, 2020 10:56 am
Human developers just need to figure out how to make regular eval match what NNUE shows.
Well, good luck ! 8-)
by ernest
Wed Jul 01, 2020 9:35 pm
Forum: Computer Chess Club: General Topics
Topic: TCEC Question
Replies: 46
Views: 1892

Re: TCEC Question

Leo wrote:
Wed Jul 01, 2020 7:12 pm
Much to my surprise LCO is really getting beat up. AB engines are not dead. Not even close.
I do repeat my question : what about the LeelaRatio of the current TCEC configuration ?
Maybe this is a clue to the match outcome...
by ernest
Mon Jun 29, 2020 1:03 am
Forum: Computer Chess Club: General Topics
Topic: TCEC Question
Replies: 46
Views: 1892

Re: TCEC Question

the Stockfish Knodes/sec were more than 2000 times those of Lc0 The "nodes" of Lc0 have absolutely nothing to do with the "nodes" of Stockfish because the architectures of both programs are completely different. Note the definition of NPS: Nodes per second, including NNCache hits and terminal node ...
by ernest
Mon Jun 29, 2020 12:36 am
Forum: Computer Chess Club: General Topics
Topic: TCEC Question
Replies: 46
Views: 1892

Re: TCEC Question

Leo wrote:
Mon Jun 29, 2020 12:24 am
I haven't heard anyone complaining about the fairness for a long time.
Not complaining, just asking ! 8-)

(looking at the Knodes/sec)
by ernest
Mon Jun 29, 2020 12:08 am
Forum: Computer Chess Club: General Topics
Topic: TCEC Question
Replies: 46
Views: 1892

Re: TCEC Question

And if SF is stronger, it is not by an enormous margain, as evidenced by the draw count. Any idea if TCEC-hardware wise, the current Lc0-Stockfish contest is "fair" ? Looking for some LeelaRatio clues, I noticed that going out of book, the Stockfish Knodes/sec were more than 2000 times those of Lc0...
by ernest
Thu Jun 18, 2020 7:41 pm
Forum: Computer Chess Club: General Topics
Topic: What is the best way to obtain the 7-piece tablebases?
Replies: 44
Views: 6265

Re: What is the best way to obtain the 7-piece tablebases?

Wrong. RTBW files work fine without RTBZ files. The other way around, not. RTBW files give win/draw/loss information (plus something about the 50-move rule which we can ignore to keep things simple). As long as you're happy with knowing that the position is a win, draw or loss, RTBW files are all y...
by ernest
Sat May 23, 2020 11:02 pm
Forum: Computer Chess Club: General Topics
Topic: Forbidden
Replies: 55
Views: 6911

Re: Forbidden

ernest wrote:
Fri May 22, 2020 8:58 pm
xr_a_y wrote:
Fri May 22, 2020 7:28 am
Seems that all IP from France are banned also,
It works again, now !!!
But obviously, for lots of users (France, Germany, ...) it still doesn't ! :(
by ernest
Fri May 22, 2020 8:58 pm
Forum: Computer Chess Club: General Topics
Topic: Forbidden
Replies: 55
Views: 6911

Re: Forbidden

xr_a_y wrote:
Fri May 22, 2020 7:28 am
Seems that all IP from France are banned also,
It works again, now !!!

Wonder why ! :o :D
by ernest
Thu Apr 30, 2020 12:44 am
Forum: Computer Chess Club: General Topics
Topic: LC0 30x384 (3350) + 2x RTX 2080 Ti vs Stockfish Dev 22.04.2020 + 6 cores + Contempt -75 should be a fair play/ratio
Replies: 26
Views: 3976

Re: LC0 30x384 (3350) + 2x RTX 2080 Ti vs Stockfish Dev 22.04.2020 + 6 cores + Contempt -75 should be a fair play/ratio

corres wrote:
Wed Apr 29, 2020 8:35 am
Some notes:
1, 40 games are too few to measure Elo.
Yes indeed !

The error margins (95%) are :
+17 =20 -3 approx. +/- 70 Elo

+9 =30 -1 approx. +/- 50 Elo

+10 =27 -3 approx. +/- 60 Elo

+6 =34 -0 approx. +/- 40 Elo
by ernest
Wed Mar 18, 2020 3:00 am
Forum: Computer Chess Club: General Topics
Topic: Speaking of TCEC, what is all this then?
Replies: 8
Views: 1846

Re: Speaking of TCEC, what is all this then?

Jouni wrote:
Tue Mar 17, 2020 6:36 pm
BTW they should use SF11
SF11 has bugs... (corrected in later devs)