Author: Tim Foden
Date: 03:21:59 04/17/03
Go up one level in this thread
On April 15, 2003 at 17:23:09, Will Singleton wrote:
>[D]r1br2k1/p1q2ppp/2n2n2/2B5/P1BNp3/1QP1P3/5PPP/R4RK1 b - -
>
>Some programs never show an elevated score, others show +2 or so before figuring
>it out.
With Green Light's normal Q-search pruning algorithm:
8 0.811 -0.699 701927 Ne5 2. Qb4 Nfg4 3. g3 Nd3 4. Bxd3 exd3 5. Rab1 Bd7
8 1.002 -0.699 858591 Ne5 2. Qb4 Nfg4 3. g3 Nd3 4. Bxd3 exd3 5. Rab1 Bd7
9 1.763 -0.299 1546087 Ne5 {++} 2. Qb4 Nxc4 3. Qxc4 Rd5 4. Nb3 Rh5 5. g3
Be6 6. Bd6 Bxc4
9 3.165 +1.249 2674637 Ne5 2. Qb4 Nxc4 3. Nb5 Qc6 4. Qxc4 Rd5 5. Nxa7
Qxc5 6. Qxc5 Rxc5
9 3.355 +1.249 2849285 Ne5 2. Qb4 Nxc4 3. Nb5 Qc6 4. Qxc4 Rd5 5. Nxa7
Qxc5 6. Qxc5 Rxc5
10 6.229 +1.359 5424666 Ne5 2. Qb4 Nxc4 3. Nb5 Qc6 4. Nxa7 Qe6 5. Nc6 Rd2
6. a5
10 6.610 +1.359 5792384 Ne5 2. Qb4 Nxc4 3. Nb5 Qc6 4. Nxa7 Qe6 5. Nc6 Rd2
6. a5
11 10.605 +0.959 9194565 Ne5 {--} 2. Qb4 Nxc4 3. Nb5 Qc6 4. Qxc4 Rd5 5.
Rad1 Rxc5 6. Rd8+ Ne8 7. Rxe8+ Qxe8 8. Qxc5
11 14.441 -0.029 12922k Ne5 2. Qb4 Nxc4 3. Nb5 Qc6 4. Qxc4 Nd7 5. Rfd1 Bb7
6. Rxd7 Qxd7 7. Nxa7
11 17.856 -0.029 15855k Ne5 2. Qb4 Nxc4 3. Nb5 Qc6 4. Qxc4 Nd7 5. Rfd1 Bb7
6. Rxd7 Qxd7 7. Nxa7
With all q-search pruning switched off:
7 0.561 -0.811 541626 Ne5 2. Qb4 Nfg4 3. g3 Rb8 4. Bb5 Nd3
7 0.752 -0.811 695868 Ne5 2. Qb4 Nfg4 3. g3 Rb8 4. Bb5 Nd3
8 1.182 -0.411 1140402 Ne5 {++} 2. Qb4 Nxc4 3. Qxc4 Rd5 4. Nb5 Qxc5 5.
Qxc5 Rxc5 6. Nxa7
8 1.533 +0.116 1510398 Ne5 2. Qb4 Nxc4 3. Qxc4 Rd5 4. Nb3 Rb8 5. Rad1
Rxd1 6. Rxd1 Rxb3 7. Qxb3 Qxc5
8 1.733 +0.116 1697069 Ne5 2. Qb4 Nxc4 3. Qxc4 Rd5 4. Nb3 Rb8 5. Rad1
Rxd1 6. Rxd1 Rxb3 7. Qxb3 Qxc5
9 2.764 +0.116 2640575 Ne5 2. Qb4 Nxc4 3. Qxc4 Rd5 4. Nb3 Rb8 5. Rad1
Rxd1 6. Rxd1 Rxb3 7. Qxb3 Qxc5
9 3.445 +0.116 3294822 Ne5 2. Qb4 Nxc4 3. Qxc4 Rd5 4. Nb3 Rb8 5. Rad1
Rxd1 6. Rxd1 Rxb3 7. Qxb3 Qxc5
10 6.490 +0.516 6168784 Ne5 {++} 2. Qb4 Nxc4 3. Qxc4 Rd5 4. Nb3 Rh5 5. g3
Be6 6. Bb6 Qxc4
10 11.587 +1.359 10906k Ne5 2. Qb4 Nxc4 3. Nb5 Qc6 4. Nxa7 Qe6 5. Nc6 Rd2
6. a5
10 12.388 +1.359 11669k Ne5 2. Qb4 Nxc4 3. Nb5 Qc6 4. Nxa7 Qe6 5. Nc6 Rd2
6. a5
11 16.915 +0.959 15981k Ne5 {--} 2. Qb4 Nxc4 3. Nb5 Qc6 4. Qxc4 Rd5 5.
Rad1 Rxc5 6. Rd8+ Ne8 7. Rxe8+ Qxe8 8. Qxc5
11 22.363 +0.600 21211k Ne5 2. Qb4 Nxc4 3. Nb5 Qc6 4. Qxc4 Nd7 5. Rad1 Bb7
6. Nd6 Qxc5 7. Qxf7+
11 26.519 +0.600 24926k Ne5 2. Qb4 Nxc4 3. Nb5 Qc6 4. Qxc4 Nd7 5. Rad1 Bb7
6. Nd6 Qxc5 7. Qxf7+
12 46.868 +0.200 43648k Ne5 {--} 2. Qb4 Nxc4 3. Nb5 Qc6 4. Qxc4 Nd7 5.
Rfd1 Qxc5 6. Qxc5 Rb8
12 59.526 -0.081 55525k Ne5 2. Qb4 Nxc4 3. Nb5 Qc6 4. Qxc4 Nd7 5. Rfd1 Ba6
6. Rd5 Rdc8 7. Rxd7 Qxd7 8. Qxe4 Bxb5 9. axb5
Qxb5 10. Bxa7
12 1:17.91 -0.081 71525k Ne5 2. Qb4 Nxc4 3. Nb5 Qc6 4. Qxc4 Nd7 5. Rfd1 Ba6
6. Rd5 Rdc8 7. Rxd7 Qxd7 8. Qxe4 Bxb5 9. axb5
Qxb5 10. Bxa7
So this seems to say that the pruning in the q-search is not the problem here.
It may be that there is a combination of things that cause this problem to be
exhibited. E.G. GL doesn't see any checkmates in the q-search. So if it did
see checkmates, and didn't prune, maybe that would make the difference?
Maybe I'll give it a try.
Cheers, Tim.
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.