Search found 3980 matches

by jdart
Tue Jul 10, 2007 3:07 am
Forum: Computer Chess Club: General Topics
Topic: difficult endgame test
Replies: 22
Views: 7017

difficult endgame test

[D] 8/6p1/8/p1p1p1kP/1p1pP1P1/1P1P3K/1PP5/8 b - -

This is from Havasi-Peko, Budapest 1976 - it is one of the examples in Dvoretsky's endgame book.

bm is c4, which I have verified at long time control. Arasan gets this but not rapidly.

--Jon
by jdart
Tue Jun 12, 2007 3:56 am
Forum: Computer Chess Club: General Topics
Topic: "avoid move" positions
Replies: 2
Views: 1530

"avoid move" positions

A couple of these, from recent Arasan games:

[D] 8/7p/1p4p1/3P1n2/BK5k/2p2P1P/6p1/4R3 b - -

am Kxh3

[D] 4N3/5k2/P5p1/4p2p/3pPqp1/8/5PnP/R4R1K w - -

am Kxg2
by jdart
Sun Jun 10, 2007 7:16 pm
Forum: Computer Chess Club: General Topics
Topic: CCT 10 - Early discussion and date selection.
Replies: 13
Views: 5164

Re: CCT 10 - Early discussion and date selection.

Last time we discussed an entry fee it was an impediment to some. I will volunteer to sponsor up to 3 entrants who could not otherwise participate due to the entry fee.

--Jon
by jdart
Sun Jun 10, 2007 2:31 pm
Forum: Computer Chess Club: General Topics
Topic: CCT 10 - Early discussion and date selection.
Replies: 13
Views: 5164

Re: CCT 10 - Early discussion and date selection.

I may have a conflict Feb 23 but the others are ok.

--Jon
by jdart
Fri Jun 08, 2007 4:25 pm
Forum: Computer Chess Club: General Topics
Topic: Good Engine Test
Replies: 6
Views: 2575

Re: Good Engine Test

Cute. 99 seconds for Arasan, recent build, on a fairly slow machine: Fen: r1b2k1r/pp1q2pp/1p1p2n1/3p2BQ/8/8/1PP2PPP/R3R1K1 w - - 0 1 Bm: Ra3 no 1 0 -340 18 Qe2 no 1 0 -334 78 Qe2 no 2 0 -334 228 Qe2 no 2 0 -334 268 Qe2 no 2 0 -334 342 Qe2 no 2 0 -415 478 Rad1 Qf5 no 3 10 -409 2306 Rad1 Qf5 c3 no 3 1...
by jdart
Wed Jun 06, 2007 5:34 pm
Forum: Computer Chess Club: General Topics
Topic: position from ChessWar tournament
Replies: 3
Views: 1681

position from ChessWar tournament

This is from a game Rybka-Scorpio: [D] r1b1nrk1/pp1n1pbp/2p3p1/2p1P3/Nq2P3/5NPP/PP1B1PB1/R2Q1RK1 b - - am Qxe4. Scorpio played Qxe4 with an even score, but after Re1, its score dropped to -2. This is the full game: [Event "ChessWar X A 40m/40'"] [Site "DEVILLE1-5D9C28"] [Date "2007.05.27"] [Round "1...
by jdart
Mon May 28, 2007 4:04 am
Forum: Computer Chess Club: General Topics
Topic: SpiderChess 070525 is released....
Replies: 11
Views: 5762

Re: SpiderChess 070525 is released....

Since this doesn't come with an opening book and since I had a small book file handy in almost the right format, I made a minimal set of opening books. These are available from: http://www.arasanchess.org/SpiderBlack.bk http://www.arasanchess.org/SpiderWhite.bk No guarantees as to quality. The repet...
by jdart
Sat May 26, 2007 4:05 am
Forum: Computer Chess Club: General Topics
Topic: About 10% of the WCSAC EPD positions are broken.
Replies: 5
Views: 2735

Re: About 10% of the WCSAC EPD positions are broken.

The fact that Rybka gave the "correct" solution does not mean there are not other alternate solutions. This test suite in particular has a lot of positions where there's something like a mate in 4 and an alternative mate in 5. Many programs will find the mate in 5 first (because of move ordering) an...
by jdart
Sat May 26, 2007 3:01 am
Forum: Computer Chess Club: General Topics
Topic: About 10% of the WCSAC EPD positions are broken.
Replies: 5
Views: 2735

Re: About 10% of the WCSAC EPD positions are broken.

I think more than 10% are broken, over 50% I would guess. I stopped using this test suite a while ago.

As you say there are alternate mate sequences in many cases. But there are also positions where there is no clear best move, or the supplied solution is not best.

--Jon
by jdart
Fri May 25, 2007 4:43 pm
Forum: Computer Chess Club: General Topics
Topic: Eigenmann Endgame Test
Replies: 15
Views: 7013

Re: Eigenmann Endgame Test

>Do mean they will never get it or it takes 10 times longer?

Never. But I have tried scaling back null move and, while it "fixes" corner case like this, it caused overall performance to suffer.