Author: Robert Hyatt
Date: 11:30:22 06/09/04
Go up one level in this thread
On June 09, 2004 at 06:12:24, Stan Arts wrote:
>Yesterday, in the evening took my tabletop chesscomputer outside for a quick
>game.
>
>This position came on the board.
>
>[D]r4rk1/1Rp3pp/2pq4/p2p1p2/Q1nP4/2P1P3/P4PPP/2B2RK1 w - a6 0 18
>
>Here, it's easy as a human to see I can trap the rook with the knight, and then
>win the quality with my queen if white doesn't move it away first. Is there any
>program that can see that fast as well? Probably by evaluation, (For instance
>something like almost no mobility left and surrounded/paths stopped by opponent
>pieces fairly close by) because by calculation it's very deep and takes a long
>time.
>
>Neurosis does not have such code (yet) and needs depth 12 to see the rook will
>be in trouble after playing Rd1, and then finds Qb3 that probably delays the
>trouble till depth 13 or 14. The other program's I tried, seem to have the same
>trouble and need similair depths.
>
>Neurosis's output. (One hour on my slow PIII-550 48mb hash, speculative 1)
>
>Legal moves: 32 Static score: 47
>Ply Score Time Nodes Best move and expected line
>------------------------------------------------------------
>2 47 0 933 Qb3 c5
>3 48 0 1519 Qb3 c5 Bb2
>4 44 6 4309 Qb3 a4 Qc2 c5
>4 48 17 7369 Qc2 c5 Bb2 Nxb2
>5 39 39 20530 Qc2 Ra8-e8 Rd1 c5 Bb2
>6 32 121 70311 Qc2 c5 Bd2 cxd4 cxd4 c5
>7 23 429 240785 Qc2 c5 Re1 Qc6 Rb3 cxd4 cxd4
>7 24 511 291178 Qb3 < + > c5 Qb5 Rf6
>7 30 561 320674 Qb3 c5 Qb5 c6 Qxc5 Qxc5 dxc5
>8 17 1324 786007 Qb3 Ra6 Qc2 Rb6 Ra7 c5 Bd2 Rb2 Ra1
>8 18 1401 836391 Qc2 < + > c5 Re1 Qc6 Rb3 cxd4 cxd4
>8 18 1555 918079 Qc2 c5 Re1 Qc6 Bb2 Nd2 g4 Ra8-d8
>8 19 2236 1328063 Rd1 < + > Ra8-b8 Qb3 Rxb7 Qxb7 Rf7
>8 20 2340 1389702 Rd1 Ra8-b8 Qb3 a4 Qb1 Rxb7 Qxb7 Re8
>9 21 5663 3310854 Rd1 Ra8-b8 Rb3 Nb6 Qxa5 Nc4 Qc5 Qxc5 dxc5
>10 21 17780 10881418 Rd1 Ra8-b8 Qb3 Rxb7 Qxb7 Rd8 Bb2 c5 Rb1 Nxb2
>11 21 52542 33090279 Rd1 Ra8-b8 Qb3 Rxb7 Qxb7 Rd8 Qb1 Qd7 Qb3 c5 dxc5
>Bb2
>12 -31 163612 100323049 Rd1 < - > Nb6 Qb3 Rf7 Qa3 Qd8 Qc5 Rf6 Rxc7 Qxc7
>Qxb6 Qxb6
>12 -31 171181 104500716 Rd1 Nb6 Rxb6 cxb6 Ba3 c5 dxc5 bxc5 c4 Qb6 cxd5
>Ra8-d8
>12 -30 210222 125003013 Qb3 < + > Ra6 Qc2 Rb6 Ra7 c5 dxc5 Qxc5 f3 Qd3
>12 5 322072 194423290 Qb3 c5 Rd1 a4 Qb1 cxd4 cxd4 Na5 Rb5 c6 Qb4 Qxb4
>Total nodes:217936078 n/sec:60600 (Q-nodes:38% max depth:31)
>
>Greetings Stan
>
>[Date "2004.06.08"]
>[White "Saitek turbo advanced trainer mode d6 - 45 minutes game"]
>[Black "Me"]
>
>1. Nf3 Nf6 2. d4 d5 3. c4 e6 4. Nc3 Bb4 5. cxd5 exd5 6. Qa4+ Nc6 7. e3 Ne4
>8. Bb5 Bd7 9. Qb3 Nd6 10. Bxc6 Bxc3+ 11. bxc3 Bxc6 12. Ne5 O-O 13. Nxc6
>bxc6 14. O-O Nc4 15. Qa4 Qd6 16. Rb1 f5 17. Rb7 a5 18. Qc2 Nb6 19. Bb2 c5
>20. dxc5 Qxc5 21. c4 Qc6 22. Qc3 Rf7 23. Rxb6 cxb6 24. cxd5 Qxd5 25. a3 Rf6
>26. Rc1 Rg6 27. g3 Rf8 28. h3 Qd6 29. Kf1
>(my time was running out, and was time for coffee)
Crafty doesn't specifically understand this kind of trapped rook, but it does
think, from the get-go, that black is winning here:
7-> 0.29 -1.15 1. Qc2 Qd7 2. f3 Qc8 3. Rb3 Qe6 4.
Rb7
8 0.39 -1.15 1. Qc2 Qd7 2. f3 Qc8 3. Rb1 Qe6 4.
Rb7 Rfc8 <HT>
8-> 0.48 -1.15 1. Qc2 Qd7 2. f3 Qc8 3. Rb1 Qe6 4.
Rb7 Rfc8 <HT>
9 1.00 -1.11 1. Qc2 Ra6 2. e4 fxe4 3. g3 Rb6 4.
Bf4 Qf6 5. Rxc7
9 1.53 -1.01 1. f3 Ra6 2. e4 fxe4 3. fxe4 Rxf1+
4. Kxf1 Qxh2 5. exd5 Qh1+ 6. Kf2 Qh4+
7. Kf1 cxd5 8. Rxc7
9-> 1.53 -1.01 1. f3 Ra6 2. e4 fxe4 3. fxe4 Rxf1+
4. Kxf1 Qxh2 5. exd5 Qh1+ 6. Kf2 Qh4+
7. Kf1 cxd5 8. Rxc7 (s=2)
10 2.11 -1.23 1. f3 Ra6 2. Re1 Rb6 3. Ra7 Qf6 4.
Ba3 Nxa3 5. Qxa3 Qh4 6. Qc1
10 2.47 -1.13 1. Qc2 Ra6 2. g4 fxg4 3. Qe2 Rf3 4.
Rb8+ Kf7 5. Qd3 Qh6 6. Rb7
10-> 2.86 -1.13 1. Qc2 Ra6 2. g4 fxg4 3. Qe2 Rf3 4.
Rb8+ Kf7 5. Qd3 Qh6 6. Rb7 (s=2)
11 4.37 -1.28 1. Qc2 Ra6 2. Re1 Rb6 3. Rxb6 cxb6
4. f3 Qf6 5. e4 fxe4 6. fxe4 Qh4 7.
Re2
11-> 7.34 -1.28 1. Qc2 Ra6 2. Re1 Rb6 3. Rxb6 cxb6
4. f3 Qf6 5. e4 fxe4 6. fxe4 Qh4 7.
Re2 (s=3)
12 10.15 -1.32 1. Qc2 Ra6 2. g4 fxg4 3. Qe2 Rb6 4.
Rxb6 cxb6 5. Qxg4 Qf6 6. e4 dxe4 7.
Qxe4 a4 (s=2)
12 19.03 -1.21 1. Qd1 Qd7 2. Rb1 Ra6 3. f3 Rb6 4.
Rb3 Qd6 5. e4 fxe4 6. fxe4 dxe4 7.
Bg5
12-> 21.85 -1.21 1. Qd1 Qd7 2. Rb1 Ra6 3. f3 Rb6 4.
Rb3 Qd6 5. e4 fxe4 6. fxe4 dxe4 7.
Bg5
13 32.05 -1.20 1. Qd1 Qd7 2. Rb1 Ra6 3. f3 Rb6 4.
Rb3 Qe6 5. e4 fxe4 6. fxe4 Rxf1+ 7.
Qxf1 dxe4 8. Bf4 Qf7
13-> 37.23 -1.20 1. Qd1 Qd7 2. Rb1 Ra6 3. f3 Rb6 4.
Rb3 Qe6 5. e4 fxe4 6. fxe4 Rxf1+ 7.
Qxf1 dxe4 8. Bf4 Qf7
14 1:14 -1.20 1. Qd1 Qd7 2. Rb1 Ra6 3. f3 Rb6 4.
Rb3 Qe6 5. e4 fxe4 6. fxe4 Rxf1+ 7.
Qxf1 Qxe4 8. Rxb6 cxb6 9. Bg5
14-> 1:55 -1.20 1. Qd1 Qd7 2. Rb1 Ra6 3. f3 Rb6 4.
Rb3 Qe6 5. e4 fxe4 6. fxe4 Rxf1+ 7.
Qxf1 Qxe4 8. Rxb6 cxb6 9. Bg5 (s=2)
IE looks like black is going to win a pawn no matter what. Perhaps not worrying
about the rook is not a bad plan here as at best it loses the exchange...
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.