Search found 4172 matches

by syzygy
Sat Jun 30, 2012 2:40 pm
Forum: Computer Chess Club: General Topics
Topic: Question about Critter 1.6 & Gaviota gtb's
Replies: 14
Views: 2934

Re: Question about Critter 1.6 & Gaviota gtb's

I would be surprised if HDD seek time beats USB2 access time. For tablebase access, seek time is what counts. USB access is 20-100x faster than HDD-access. Because of that, all TBs are very, very slow on HDDs! Yes, that's what I thought. Of course a real SSD would perform better than a USB stick. N...
by syzygy
Fri Jun 29, 2012 11:47 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: info about zappa on 512 cores ?
Replies: 23
Views: 3904

Re: transposition tables

Give N small hash tables of the same size with K entries, how do you hash a position with zobrist hash hash_key ? Assume the small hash tables are physically separate (on different processor if you will). If I understand you correctly, the N small hash tables form one large hash table. Each positio...
by syzygy
Fri Jun 29, 2012 8:49 pm
Forum: Computer Chess Club: General Topics
Topic: Question about Critter 1.6 & Gaviota gtb's
Replies: 14
Views: 2934

Re: Question about Critter 1.6 & Gaviota gtb's

flash memory certainly lessens wear and tear on the HDD. But what about access time? Using flash memory requires USB (usb2 in most cases) access time as opposed to HDD access time. I would be surprised if HDD seek time beats USB2 access time. For tablebase access, seek time is what counts. Of cours...
by syzygy
Mon Jun 25, 2012 7:02 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Texel recipe to fix TT draws scores
Replies: 8
Views: 1660

Re: Texel recipe to fix TT draws scores

If fixes the problem during the last 20 plies before the 50-move rule that hash grafts can cause the 50-move rule to be applied incorrectly. It doesn't in general fix draw by repetition problems though. It may "accidentally" fix some draw by repetition problems as a side effect of giving fewer hash...
by syzygy
Sun Jun 24, 2012 2:24 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: KQKP and KRKP endgames
Replies: 38
Views: 3544

Re: KQKP and KRKP endgames

In the eval, you only need suitable PSTs for solving KQK, KBNK, KRK, and you need nothing special to play KQKQ perfectly. So this can't be what Joona is talking about. I almost agree, except for the "PST" argument since mating the lone king usually also requires to approach the own king to the enem...
by syzygy
Sun Jun 24, 2012 12:39 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: KQKP and KRKP endgames
Replies: 38
Views: 3544

Re: KQKP and KRKP endgames

Could it be that Joona is talking about internal node recognisers that cut a whole subtree when a position with known outcome is reached? I suppose it could indeed be hard to statically determine with 100% certainty whether KQKQ is draw, but I don't see the need for the evaluation to evaluate KQKQ ...
by syzygy
Sun Jun 24, 2012 11:05 am
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: KQKP and KRKP endgames
Replies: 38
Views: 3544

Re: KQKP and KRKP endgames

Could it be that Joona is talking about internal node recognisers that cut a whole subtree when a position with known outcome is reached? I suppose it could indeed be hard to statically determine with 100% certainty whether KQKQ is draw, but I don't see the need for the evaluation to evaluate KQKQ 1...
by syzygy
Sat Jun 23, 2012 6:21 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: KQKP and KRKP endgames
Replies: 38
Views: 3544

Re: KQKP and KRKP endgames

Some positions predicted as a win, when there they are draw, could generate bad scenarios. But this is exactly the situation you get with a KQKP-specific rule that only predicts correct draws. When this rule does not give an answer, you fall back on the regular eval which surely will predict a win ...
by syzygy
Sat Jun 23, 2012 5:11 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: KQKP and KRKP endgames
Replies: 38
Views: 3544

Re: KQKP and KRKP endgames

In the given case of KQKP, why do you believe that it is better to accept some false draw predictions in exceptional cases, instead of restricting draw detection by 100% to those cases where exact knowledge is present? I am not saying that accepting some false draw predictions would cause severe tr...
by syzygy
Sat Jun 23, 2012 4:05 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: KQKP and KRKP endgames
Replies: 38
Views: 3544

Re: KQKP and KRKP endgames

If a simplistic case is applied, it should be better to include only basic cases in which you are 100% sure the position is predicted well (draws in this case), and let the other cases be found in search. This is what my KQKP predictor does. Without tablebases, my engine only recognises the draw at...