Discussion of anything and everything relating to chess playing software and machines.
Moderator: Ras
lucasart
Posts: 3243Joined: Mon May 31, 2010 1:29 pmFull name: lucasart
Post by lucasart Sun Dec 18, 2011 11:03 pm
zullil wrote: Not what I'm seeing
That's because you're looking at the PV. It's sometimes buggy, and anyway irrelevant. From the starting position, see if it finds the right move. Then play it, and repeat the operation
This is what happens in a real game.
By the way, you just pointed out a bug in my program. The following pathetic PV line:
Code: Select all
info score cp 26 depth 25 nodes 12679634 time 8962 pv Nf6e4 Ke3e2 Kc6d5 Ke2d1 Ne4f6 Kd1e2 Kd5e4 Be1f2 Nf6d5 Bf2e1 Nd5e3 Be1d2 Ne3d5 Bd2e1 Nd5e3 Be1d2 Ne3d5 Bd2e1 Nd5e3 Be1d2 Ne3d5 Bd2e1 Nd5e3 Be1d2 Ne3d5 Bd2e1 Nd5e3 Be1d2 Ne3d5 Bd2e1 Nd5e3 Be1d2 Ne3d5 Bd2e1 Nd5e3 Be1d2 Ne3d5 Bd2e1 Nd5e3 Be1d2 Ne3d5 Bd2e1 Nd5e3 Be1d2 Ne3d5 Bd2e1 Nd5e3 Be1d2 Ne3d5 Bd2e1 Nd5e3 Be1d2 Ne3d5 Bd2e1 Nd5e3 Be1d2 Ne3d5 Bd2e1 Nd5e3 Be1d2 Ne3d5 Bd2e1 Nd5e3 Be1d2 Ne3d5 Bd2e1 Nd5e3 Be1d2 Ne3d5 Bd2e1 Nd5e3 Be1d2 Ne3d5 Bd2e1 Nd5e3 Be1d2 Ne3d5 Bd2e1 Nd5e3 Be1d2 Ne3d5 Bd2e1 Nd5e3 Be1d2 Ne3d5 Bd2e1 Nd5e3 Be1d2 Ne3d5 Bd2e1 Nd5e3 Be1d2 Ne3d5 Bd2e1 Nd5e3 Be1d2 Ne3d5 Bd2e1 Nd5e3 Be1d2 Ne3d5 Bd2e1 Nd5e3 Be1d2 Ne3d5 Bd2e1 Nd5e3 Be1d2 Ne3d5 Bd2e1 Nd5e3 Be1d2 Ne3d5 Bd2e1 Nd5e3 Be1d2 Ne3d5 Bd2e1 Nd5e3 Be1d2 Ne3d5 Bd2e1 Nd5e3 Be1d2 Ne3d5 Bd2e1 Nd5e3 Be1d2
come to think of it, it's a "normal" consequence of how I compute the PV. I just retrieve it from the transpotition table, only stopping when the move is illegal, but not testing whether a draw (here a 3 move draw) has been reached.
I'll fix it for the next release
Thanks!
Houdini
Posts: 1471Joined: Tue Mar 16, 2010 12:00 am
Post by Houdini Sun Dec 18, 2011 11:42 pm
ernest wrote: rabbits wrote: It certainly does Robert. I might have to get myself a faster computer
No need... Just let your test last longer.
Actually, with my Houdinis (1.5a, 2.0c) I dont'get Robert's result, and stay miserably with (-0.38)
You must run to around search depth 40 to find the solution.
My Houdini 2.0c is running now and shows the fail high at search depth 39, I'll post the PV once it's available.
Robert
zullil
Posts: 6442Joined: Tue Jan 09, 2007 12:31 amLocation: PA USAFull name: Louis Zulli
Post by zullil Sun Dec 18, 2011 11:44 pm
lucasart wrote: zullil wrote: Not what I'm seeing
That's because you're looking at the PV. It's sometimes buggy, and anyway irrelevant. From the starting position, see if it finds the right move. Then play it, and repeat the operation
This is what happens in a real game.
Yes, this is a valid point. It's enough to make good moves, one after another. Will give DoubleCheck another try.
lucasart
Posts: 3243Joined: Mon May 31, 2010 1:29 pmFull name: lucasart
Post by lucasart Sun Dec 18, 2011 11:55 pm
zullil wrote: lucasart wrote: zullil wrote: Not what I'm seeing
That's because you're looking at the PV. It's sometimes buggy, and anyway irrelevant. From the starting position, see if it finds the right move. Then play it, and repeat the operation
This is what happens in a real game.
Yes, this is a valid point. It's enough to make good moves, one after another. Will give DoubleCheck another try.
and you may find that other engines play the correct moves too.
anyway, thanks for pointing out the bug on repetition in PV line, even if it was not intended
zullil
Posts: 6442Joined: Tue Jan 09, 2007 12:31 amLocation: PA USAFull name: Louis Zulli
Post by zullil Mon Dec 19, 2011 12:12 am
How about from here, which should be much easier?
lucasart
Posts: 3243Joined: Mon May 31, 2010 1:29 pmFull name: lucasart
Post by lucasart Mon Dec 19, 2011 12:39 am
zullil wrote: How about from here, which should be much easier?
Indeed, DC 2.3 can't find Na3! here, even after a long time.
Anyway, still a good surprise that my 2300 elo patzer played the first 7 moves correctly.
In fact even StockFish can't find Na3 in a "reasonable time" here, so it's not as easy as you think
zullil
Posts: 6442Joined: Tue Jan 09, 2007 12:31 amLocation: PA USAFull name: Louis Zulli
Post by zullil Mon Dec 19, 2011 12:50 am
lucasart wrote: zullil wrote: How about from here, which should be much easier?
Indeed, DC 2.3 can't find Na3! here, even after a long time.
Anyway, still a good surprise that my 2300 elo patzer played the first 7 moves correctly.
In fact even StockFish can't find Na3 in a "reasonable time" here, so it's not as easy as you think
Yes, I let your engine run to depth 29, but decided it wasn't going to happen. Stockfish found Na3 at depth 57. Critter succeeded almost instantly:
Code: Select all
Critter Chess v1.2.1 64-bit SSE4
By Richard Vida, Slovakia
GTB Init OK
Hash table: 67108864 entries of 16 bytes = 1024 MB total
Eval cache: 32768 entries of 8 bytes = 256 KB total
Pawn hash : 65536 entries of 32 bytes = 2 MB total
8 cpu(s) detected
Opening book: "book.cbk" - 53436 entries [native]
setoption name Hash value 2048
Hash table: 134217728 entries of 16 bytes = 2048 MB total
setboard 8/8/4p3/1n1k1p2/p1pP1Pp1/2P1K1P1/P2B4/8 b - - 14 8
go infinite
2/ 2 00:00 32 0 +0.33 Nd6 Bc1 Ne4
3/ 4 00:00 76 0 +0.33 Nd6 Bc1 Ne4
4/ 4 00:00 248 0 +0.50 Nd6 Be1 Ne4 a3
5/ 5 00:00 341 0 +0.22 Nd6 Be1 Ne4 a3 Kd6
6/ 6 00:00 475 0 +0.44 Nd6 Be1 Ne4 a3 Kd6 Ke2
7/ 7 00:00 643 0 +0.24 Nd6 Be1 Ne4 a3 Kd6 Ke2 Ke7 Ke3
8/ 9+ 00:00 988 988000 +0.32 Nd6 Be1 Ne4 a3 Kd6 Ke2 Ke7 Ke3
8/ 9+ 00:00 1052 1052000 +0.41 Nd6 Be1 Ne4 a3 Kd6 Ke2 Ke7 Ke3
8/ 9 00:00 1188 1188000 +0.44 Nd6 Be1 a3 Bd2 Ne4 Be1 Kd6 Ke2
9/ 9- 00:00 1969 1969000 +0.35 Nd6 Be1 a3 Bd2 Ne4 Be1 Kd6 Ke2
9/ 9- 00:00 2228 2228000 +0.26 Nd6 Be1 a3 Bd2 Ne4 Be1 Kd6 Ke2
9/ 9 00:00 2432 2432000 +0.24 Nd6 Be1 a3 Bd2 Ne4 Be1 Kd6 Ke2 Ke7 Ke3
10/11 00:00 3050 1525000 +0.24 Nd6 Be1 a3 Bd2 Ne4 Be1 Kd6 Ke2 Ke7 Ke3
11/15 00:00 6986 2328666 +0.28 Nd6 Be1 a3 Kf2 Ne4+ Ke2 Kd6 Ke3 Ke7 Ke2 Kf6 Ke3
12/15 00:00 9878 1975600 +0.28 Nd6 Be1 a3 Kf2 Ne4+ Ke2 Kd6 Ke3 Ke7 Ke2 Kf6 Ke3
13/15 00:00 13278 1896857 +0.28 Nd6 Be1 a3 Kf2 Ne4+ Ke2 Kd6 Ke3 Ke7 Ke2 Kd7 Ke3 Kd6 Ke2 Ke7 Ke3 Kd7 Ke2 Kd6
14/16 00:00 19247 2138555 +0.28 Nd6 Be1 a3 Kf2 Ne4+ Ke2 Kd6 Ke3 Ke7 Ke2 Kd7 Ke3 Kd6 Ke2 Ke7 Ke3 Kd7 Ke2 Kd6
15/18 00:00 28823 2217153 +0.28 Nd6 Be1 a3 Kf2 Ne4+ Ke2 Kd6 Ke3 Ke7 Ke2 Kd7 Ke3 Kd6 Ke2 Ke7 Ke3 Kd7 Ke2 Kd6
16/21 00:00 45259 2514388 +0.28 Nd6 Be1 a3 Kf2 Ne4+ Ke2 Kd6 Ke3 Ke7 Ke2 Kd7 Ke3 Kd6 Ke2 Ke7 Ke3 Kd7 Ke2 Kd6
17/21 00:00 66538 2661520 +0.28 Nd6 Be1 a3 Kf2 Ne4+ Ke2 Kd6 Kf1 Ke7 Ke2 Kd7 Ke3 Kd6 Ke2 Ke7 Ke3 Kd7 Ke2 Kd6
18/24 00:00 120725 2944512 +0.28 Nd6 Be1 a3 Kf2 Ne4+ Ke3 Kd6 Ke2 Kc6 Ke3 Kb5 Ke2 Kb6 Ke3 Kc6 Ke2 Kb5 Ke3 Kb6 Ke2 Kc7 Ke3 Kd7 Ke2 Ke7 Ke3 Kf7 Ke2 Kf6 Ke3
19/28 00:00 201056 3191365 +0.28 Nd6 Be1 a3 Kf2 Ne4+ Ke3 Kd6 Ke2 Kc6 Ke3 Kb5 Ke2 Kb6 Ke3 Kc6 Ke2 Kb5 Ke3 Kb6 Ke2 Kc7 Ke3 Kd7 Ke2 Ke7 Ke3 Kf7 Ke2 Kf6 Ke3
20/29 00:00 383923 3427883 +0.28 Nd6 Be1 a3 Kf2 Ne4+ Ke3 Kd6 Ke2 Kc6 Ke3 Kb5 Ke2 Kb6 Ke3 Kc6 Ke2 Kd6 Kf1 Ke7 Ke2 Kd7 Kf1 Ke8 Ke2 Ke7 Ke3 Kf7 Ke2 Kf6 Ke3
21/31 00:00 626104 3640139 +0.28 Nd6 Be1 a3 Kf2 Ne4+ Ke3 Kc6 Ke2 Kd6 Ke3 Kd5 Ke2 Kc6 Ke3 Kb5 Ke2 Kb6 Ke3 Kc7 Ke2 Kd7 Kf1 Ke7 Ke2 Kf7 Ke3 Ke8 Ke2 Ke7 Ke3 Kf7 Ke2 Kf6 Ke3
22/34 00:00 1272170 4064440 +0.28 Nd6 Be1 a3 Kf2 Ne4+ Ke3 Kc6 Ke2 Kd6 Ke3 Kd5 Ke2 Kc6 Ke3 Kd6 Ke2 Kc7 Ke3 Kd7 Ke2 Ke7 Ke3 Kd7
23/37 00:00 2102377 4334797 +0.28 Nd6 Be1 a3 Kf2 Ne4+ Ke3 Kc6 Ke2 Kd6 Ke3 Kd5 Ke2 Kc6 Ke3 Kd7 Ke2 Kc7 Ke3 Kb6 Ke2 Kb5 Ke3 Kb6
24/43 00:00 3641640 4662791 +0.28 Nd6 Be1 a3 Kf2 Ne4+ Ke3 Kc6 Ke2 Kd6 Ke3 Kd5 Ke2 Kc6 Ke3 Kd6 Ke2 Nf6 Kf1 Kc6 Kf2 Kb5 Ke2 Kc6 Kf1 Kb5 Ke2
24/43+ 00:01 5115900 4821771 +0.37 Na3 Bc1 Nb1 Bb2 a3 Ba1 Kd6 Ke2 Kc6 Kd1 Kd5 Ke1 Ke4 Ke2 Nd2 Kxd2 Kf3 Kd1 Kxg3 Ke2 Kh2 Ke3 g3 Kf3 g2
24/43+ 00:01 5411359 4717836 +0.45 Na3 Bc1 Nb1 Bb2 a3 Ba1 Kd6 Ke2 Kc6 Kd1 Kd5 Ke1 Ke4 Ke2 Nd2 Kxd2 Kf3 Kd1 Kxg3 Ke2 Kh2 Ke3 g3 Kf3 g2
24/43+ 00:01 5887985 4618027 +0.58 Na3 Bc1 Nb1 Bb2 a3 Ba1 Kd6 Ke2 Kc6 Kd1 Kd5 Ke1 Ke4 Ke2 Nd2 Kxd2 Kf3 Kd1 Kxg3 Ke2 Kh2 Ke3 g3 Kf3 g2
24/47+ 00:01 6582457 4587078 +0.77 Na3 Bc1 Nb1 Bb2 a3 Ba1 Kd6 Ke2 Kc6 Kd1 Kd5 Ke1 Ke4 Ke2 Nd2 Kxd2 Kf3 Kd1 Kxg3 Ke2 Kh2 Ke3 g3 Kf3 g2
24/49+ 00:01 8266629 4696948 +1.06 Na3 Bc1 Nb1 Bb2 a3 Ba1 Kd6 Ke2 Kc6 Kd1 Kd5 Ke1 Ke4 Ke2 Nd2 Kxd2 Kf3 Kd1 Kxg3 Ke2 Kh2 Ke3 g3 Kf3 g2
24/50+ 00:02 10793211 4923910 +1.48 Na3 Bc1 Nb1 Bb2 a3 Ba1 Kd6 Ke2 Kc6 Kd1 Kd5 Ke1 Ke4 Ke2 Nd2 Kxd2 Kf3 Kd1 Kxg3 Ke2 Kh2 Ke3 g3 Kf3 g2
24/80+ 00:03 22573937 6184640 +2.12 Na3 Bc1 Nb1 Bb2 a3 Ba1 Kd6 Ke2 Kc6 Kd1 Kd5 Ke1 Ke4 Ke2 Nd2 Kxd2 Kf3 Kd1 Kxg3 Ke2 Kh2 Ke3 g3 Kf3 g2
24/85+ 00:07 54088299 6981838 +3.07 Na3 Bc1 Nb1 Bb2 a3 Ba1 Kd6 Ke2 Kc6 Kd1 Kd5 Kc2 Ke4 Kxb1 Kf3 Bb2 axb2 Kxb2 Kxg3 a4 Kxf4 a5 g3 a6 g2 a7 Ke3 a8=Q g1=Q Qc6 Qf2+ Ka3 Kd3 Qxe6 Kxc3 Ka4 Qc2+ Kb5 Kxd4 Qf6+ Ke3 Qg5+ f4 Qc5+ Ke4 Qc6+ Kf5 Qc5+ Kg4 Qg1+ Kf3 Qh1+ Kg3 Qg1+ Kh4 Qd4 Kg5 Qg1+ Kh5 Qc5+ Kg6 Qg1+ Kf6 Qd4+ Ke6 Qxf4 c3 Qe3+ Kf6
24/85+ 00:18 144694947 7785576 +4.50 Na3 Bc1 Nb1 Bb2 a3 Ba1 Kd6 Ke2 Kc6 Kd1 Kd5 Kc2 Ke4 Kxb1 Kf3 Bb2 axb2 Kxb2 Kxg3 a4 Kxf4 a5 g3 a6 g2 a7 Ke3 a8=Q g1=Q Qc6 Qf2+ Ka3 Kd3 Qxe6 Kxc3 Ka4 Qc2+ Kb5 Kxd4 Qf6+ Ke3 Qg5+ f4 Qc5+ Ke4 Qc6+ Kf5 Qc5+ Kg4 Qg1+ Kf3 Qh1+ Kg3 Qg1+ Kh4 Qd4 Kg5 Qg1+ Kh5 Qc5+ Kh4 Qd4
24/85 00:22 172687440 7754263 +4.59 Na3 Bc1 Nb1 Bb2 a3 Ba1 Kd6 Ke2 Kc6 Kd1 Kd5 Kc2 Ke4 Kxb1 Kf3 Bb2 axb2 Kxb2 Kxg3 a4 Kxf4 a5 g3 a6 g2 a7 g1=Q a8=Q Qd1 Qb8+ Kf3 d5 Qd2+ Kb1 Qxd5 Qh8 f4 Qh1+ Ke3 Qg1+ Kd3 Qf1+ Kxc3 Qc1+ Kd4 Qd2+ Ke5 Qc3+ Ke4 Qe1+ Kf5 Qh4 f3 Qh5+ Kf4 Qh6+ Qg5 Qh2+ Ke3 Qd6 f2 Qxe6+ Kd3
25/85- 00:22 173731151 7727910 +4.30 Na3 Bc1 Nb1 Bb2 a3 Ba1 Kd6 Ke2 Kc6 Kd1 Kd5 Kc2 Ke4 Kxb1 Kf3 Bb2 axb2 Kxb2 Kxg3 a4 Kxf4 a5 g3 a6 g2 a7 g1=Q a8=Q Qd1 Qb8+ Kf3 d5 Qd2+ Kb1 Qxd5 Qh8 f4 Qh1+ Ke3 Qg1+ Kd3 Qf1+ Kxc3 Qc1+ Kd4 Qd2+ Ke5 Qc3+ Ke4 Qe1+ Kf5 Qh4 f3 Qh5+ Kf4 Qh6+ Qg5 Qh2+ Ke3 Qd6 f2 Qxe6+ Kd3
25/85- 00:22 175230643 7700076 +4.02 Na3 Bc1 Nb1 Bb2 a3 Ba1 Kd6 Ke2 Kc6 Kd1 Kd5 Kc2 Ke4 Kxb1 Kf3 Bb2 axb2 Kxb2 Kxg3 a4 Kxf4 a5 g3 a6 g2 a7 g1=Q a8=Q Qd1 Qb8+ Kf3 d5 Qd2+ Kb1 Qxd5 Qh8 f4 Qh1+ Ke3 Qg1+ Kd3 Qf1+ Kxc3 Qc1+ Kd4 Qd2+ Ke5 Qc3+ Ke4 Qe1+ Kf5 Qh4 f3 Qh5+ Kf4 Qh6+ Qg5 Qh2+ Ke3 Qd6 f2 Qxe6+ Kd3
25/85 00:23 182920924 7624564 +4.57 Na3 Bc1 Nb1 Bb2 a3 Ba1 Kd6 Ke2 Kc6 Kd1 Kd5 Kc2 Ke4 Kxb1 Kf3 Bb2 axb2 Kxb2 Kxg3 a4 Kxf4 a5 g3 a6 g2 a7 g1=Q a8=Q Qd1 Qb8+ Kf3 d5 Qd2+ Kb1 Qxd5 Qh8 f4 Qh1+ Ke3 Qg1+ Kd3 Qf1+ Kxc3 Qc1+ Kb3 Qb2+ Ka4 Qc2+ Kb4 Qb2+ Kc5 Qa3+ Kd4 Kc1 f3 Qa7+ Ke4 Qf2 Qd4 Qh4+ Kd3 Qh7+ Ke2 Qc2+ Ke3
26/85- 00:24 183262675 7598270 +4.28 Na3 Bc1 Nb1 Bb2 a3 Ba1 Kd6 Ke2 Kc6 Kd1 Kd5 Kc2 Ke4 Kxb1 Kf3 Bb2 axb2 Kxb2 Kxg3 a4 Kxf4 a5 g3 a6 g2 a7 g1=Q a8=Q Qd1 Qb8+ Kf3 d5 Qd2+ Kb1 Qxd5 Qh8 f4 Qh1+ Ke3 Qg1+ Kd3 Qf1+ Kxc3 Qc1+ Kb3 Qb2+ Ka4 Qc2+ Kb4 Qb2+ Kc5 Qa3+ Kd4 Kc1 f3 Qa7+ Ke4 Qf2 Qd4 Qh4+ Kd3 Qh7+ Ke2 Qc2+ Ke3
26/85 00:25 191107665 7521850 +4.56 Na3 Bc1 Nb1 Bb2 a3 Ba1 Kd6 Ke2 Kc6 Kd1 Kd5 Kc2 Ke4 Kxb1 Kf3 Bb2 axb2 Kxb2 Kxg3 a4 Kxf4 a5 g3 a6 g2 a7 g1=Q a8=Q Qd1 Qb8+ Kf3 d5 Qd2+ Kb1 Qxd5 Qh8 f4 Qh1+ Ke2 Qh2+ Ke3 Qg1+ Kd3 Kb2 Qb5+ Kc1 Qe5 Qd1+ Kxc3 Qd2+ Kb3 Qc2+ Kb4 Qb1+ Kc5 Qg6 f3 Qg1+ Kc6 Qg4 Qe3+ Kc2 Kd5 Qh5+ Kd4 Qh8+ e5 Qd8+ Ke4 Qh4+ Qf4 Qh7+ Ke3
27/85 00:26 197552582 7508364 +4.56 Na3 Bc1 Nb1 Bb2 a3 Ba1 Kd6 Ke2 Kc6 Kd1 Kd5 Kc2 Ke4 Kxb1 Kf3 Bb2 axb2 a4 Kxg3 a5 Kxf4 a6 g3 a7 g2 a8=Q g1=Q+ Kxb2 Qd1 Qb8+ Kf3 d5 Qd2+ Kb1 Qxd5 Qh8 f4 Qh1+ Ke3 Qg1+ Kd3 Kb2 Qb5+ Kc1 Qe5 Qd1+ Kxc3 Qd2+ Kb3 Qc2+ Kb4 Qb1+ Kc5 Qg6 f3 Qg1+ Kc6 Qg4 Qe3+ Kc2 Kd5 Qh5+ Kd4 Qh8+ e5 Qd8+ Ke4 Qh4+ Qf4 Qh7+ Ke3
28/85 00:29 215534925 7412557 +4.49 Na3 Bc1 Nb1 Bb2 a3 Ba1 Kd6 Ke2 Kc6 Kd1 Kd5 Kc2 Ke4 Kxb1 Kf3 Bb2 axb2 Kxb2 Kxg3 a4 Kxf4 a5 g3 a6 g2 a7 g1=Q a8=Q Qd1 Qb8+ Kf3 d5 Qd2+ Kb1 Qxd5 Qh8 f4 Qh1+ Ke3 Qg1+ Kd3 Qf1+ Kxc3 Qc1+ Kd4 Qd2+ Ke5 Qc3+ Ke4 Qe1+ Kf5 Qh4 f3 Qh5+ Kf4 Qh6+ Qg5 Qh2+ Ke3 Qd6 f2 Qxe6+ Kd3
29/85 00:36 277101994 7544296 +4.59 Na3 Bc1 Nb1 Bb2 a3 Ba1 Kd6 Ke2 Kc6 Kd1 Kd5 Kc2 Ke4 Kxb1 Kf3 Bb2 axb2 Kxb2 Kxg3 a4 Kxf4 a5 g3 a6 g2 a7 g1=Q a8=Q Qd1 Qb8+ Kf3 d5 Qd2+ Kb1 Qxd5 Qh8 f4 Qh1+ Ke3 Qg1+ Kd3 Qf1+ Kxc3 Qc1+ Kb3 Qb2+ Ka4 Qc2+ Kb4 Qb2+ Kc5 Qa3+ Kd4 Qb2+ Ke3 Qc1+ Ke2 Qb2+ Kf3 Kc1 Ke3 Qc3+ Ke4 Qe1+ Kf5 Qh4 f3 Qh7+ Kf6 Qh8+ Kf7 Qh4 e5 Qg5 Qe4 Qh5+ Kf6 Qh6+ Kf5 Qh5+ Kf4 Qh2+
30/85 00:53 430510175 8006214 +4.75 Na3 Bc1 Nb1 Bb2 a3 Ba1 Kd6 Ke2 Kc6 Kd1 Kd5 Kc2 Ke4 Kxb1 Kf3 Bb2 axb2 Kxb2 Kxg3 a4 Kxf4 a5 g3 a6 g2 a7 g1=Q a8=Q Qd1 Qb8+ Kf3 d5 Qd2+ Kb1 Qxd5 Qh8 f4 Qh1+ Ke3 Qg1+ Kd3 Qd1+ Kxc3 Qc1+ Kd4 Qd2+ Ke5 Qc3+ Ke4 Qe1+ Kf5 Qh4 f3 Qh5+ Kf4 Qh6+ Qg5 Qh1 Qb5+ Kc2 f2 Qh4+
?
bestmove Na3 ponder Bc1
time: 81876 nodes: 701156944 evals: 455177089 knps: 8563
phash: 80.47% evalcache: 23.77% lazy: 22.86% eg_recog: 26926823
splits: 705931
quit
rabbits
Posts: 276Joined: Wed Mar 08, 2006 10:14 pmLocation: Sydney, Australia
Post by rabbits Mon Dec 19, 2011 4:50 am
Yes Lucas that is weird. My programmes only find the correct winning line once I walk them through the initial moves by Black's King and Knight which enable Black's Knight to attack white's pawns more effectively and safely
rabbits
Posts: 276Joined: Wed Mar 08, 2006 10:14 pmLocation: Sydney, Australia
Post by rabbits Mon Dec 19, 2011 4:59 am
Lucas I responded to your comment about Double Check's ability to find the winning line in more detail in a message posted below.Sorry about the confusion.
Robert Flesher
Posts: 1290Joined: Tue Aug 18, 2009 3:06 am
Post by Robert Flesher Mon Dec 19, 2011 6:40 am
Houdini wrote: ernest wrote: rabbits wrote: It certainly does Robert. I might have to get myself a faster computer
No need... Just let your test last longer.
Actually, with my Houdinis (1.5a, 2.0c) I dont'get Robert's result, and stay miserably with (-0.38)
You must run to around search depth 40 to find the solution.
My Houdini 2.0c is running now and shows the fail high at search depth 39, I'll post the PV once it's available.
Robert
Heya Robert, how did I originally get a solution so fast? Is it because I first played over the moves/solution doing my own analysis, but at the same time running Houdini in infinite analysis mode? Did Houdini remember/preserve this information in its hash tables, and therefore find the win so fast when I first ran it?
This is what happens in a real game.
