Interesting blog post about Komodo...

Discussion of anything and everything relating to chess playing software and machines.

Moderators: hgm, Rebel, chrisw

tomgdrums
Posts: 736
Joined: Wed Dec 02, 2009 9:48 am

Interesting blog post about Komodo...

Post by tomgdrums »

This correspondence chess player doesn't seem to trust Komodo based on this and a previous blog post. I wonder if there is a bug in Komodo or something odd in his machine?


http://tartajubow.blogspot.com/2011/10/ ... rning.html


Edit: For the record his blog posts have made me think twice about Komodo for a little while.
Robert Flesher
Posts: 1280
Joined: Tue Aug 18, 2009 3:06 am

Re: Interesting blog post about Komodo...

Post by Robert Flesher »

tomgdrums wrote:This correspondence chess player doesn't seem to trust Komodo based on this and a previous blog post. I wonder if there is a bug in Komodo or something odd in his machine?


http://tartajubow.blogspot.com/2011/10/ ... rning.html


Edit: For the record his blog posts have made me think twice about Komodo for a little while.

[d]2kr2r1/1q1n1p2/ppbRp3/2p1P1p1/8/2N1Q3/PPPN1PP1/3R2K1 b - - 0 0
viriathus
Posts: 32
Joined: Wed Dec 16, 2009 12:35 am

Re: Interesting blog post about Komodo...

Post by viriathus »

[d]2kr2r1/1q1n1p2/ppbRp3/2p1P1p1/8/2N1Q3/PPPN1PP1/3R2K1 b - - 0 1
It's obvious that White stands better and Houdini 1.5 x64 recommends 1...Qc7 with White's advantage at about 1/3 of a Pawn. Wanting to check this with another engine, I set up the position and left Komodo3 ponder it while I went off to do something else. 5-1/2 hours later I returned to discover Komodo had analyzed the position 13 moves deep and was showing the position as equal. It's recommendation began with 1...Bxg2 2.Nde4. 1...Bxg2?? is a horrible move because after 2.f3 Black loses a piece. Houdini evaluates the position after 2.f3 as White having an advantage of over 2P's; other engines agree.

Just to make sure Komodo wasn't seeing something in the position other engines were missing, I did some further analysis using Houdini, Komodo and Critter and managed to come up with a line where Black did get 3 P's for the pieces, but it hardly mattered because all the engines, including Komodo, agreed that Black was quite lost.
I test with Komodo 3 on Linux and Komodo choose Qc7:

1 +0.14 1.... Bxg2 (0.00)
2 +0.45 1.... Bxg2 2.Nce4 Bxe4 3.Nxe4 (0.00)
3 +0.28 1.... Bxg2 2.f3 g4 3.Kxg2 gxf3+ 4.Kf1 (0.00)
4 +0.15 1.... Bxg2 2.f3 Bh3 3.Nde4 Kb8 4.Nxg5 (0.00)
5 -0.01 1.... Bxg2 (0.00)
5 -0.11 1.... Bxg2 2.f3 Bh3 3.Nde4 Kb8 4.Nf6 (0.00)
6 +0.08 1.... Bxg2 2.Nde4 (0.01)
6 +0.11 1.... Bxg2 2.Nde4 Bxe4 3.Nxe4 Rh8 4.f3 Kb8 (0.01)
7 +0.09 1.... Bxg2 2.Nde4 Bxe4 3.Nxe4 Rh8 4.f3 Kb8 5.Nxg5 Rhg8 6.c4 (0.01)
8 +0.25 1.... Bxg2 2.f3 (0.02)
8 +0.64 1.... Bxg2 2.f3 Bh3 3.Nc4 g4 4.Nxb6+ Nxb6 5.Qxc5+ Qc7 6.Qxb6 gxf3+ (0.02)
9 +0.41 1.... Bxg2 2.f3 Bh3 3.Nc4 g4 4.Nxb6+ Nxb6 5.Qxc5+ Qc7 6.Rxd8+ Rxd8 7.Rxd8+ Kxd8 8.Qf8+ (0.04)
10 +0.80 1.... Bxg2 2.Nc4 (0.08)
10 +0.79 1.... Bxg2 (0.13)
10 +0.41 1.... Bxg2 2.f3 Bh3 3.Nc4 g4 4.Nxb6+ Nxb6 5.Qxc5+ Qc7 6.Rxd8+ Rxd8 7.Rxd8+ Kxd8 8.Qf8+ Kd7 (0.17)
11 +0.38 1.... Bxg2 2.Nc4 Bc6 3.Rxe6 Rh8 4.Nd6+ Kb8 5.f3 Qa8 6.Nxf7 Bxf3 7.Rxb6+ Ka7 8.Nxd8 Rh1+ 9.Kf2 (0.23)
12 +0.54 1.... Bxg2 2.Nc4 (0.30)
12 +0.53 1.... Bxg2 (0.41)
12 +0.53 1.... Bxg2 2.Nc4 Bc6 3.Rxe6 Rh8 4.Nd6+ Kb8 5.f3 Qa8 6.Nxf7 Bxf3 7.Nxd8 Rh1+ 8.Kf2 Bxd1 9.Nc6+ Kc7 10.Qd3 Qf8+ 11.Kg2 c4 (0.42)
13 +0.37 1.... Bxg2 (0.82)
13 +0.24 1.... Bxg2 2.Nc4 Kc7 3.Nxb6 Nxb6 4.Rxd8 Rxd8 5.Rxd8 Kxd8 6.Qxg5+ Kd7 7.Qxg2 Qxg2+ 8.Kxg2 Nc4 9.Kf3 Kc6 10.Ne4 Nxe5+ 11.Kf4 Nc4 12.b3 Na3 (1.20)
14 +0.46 1.... Bxg2 2.Nde4 (1.65)
14 +0.54 1.... Bxg2 2.Nde4 Bxe4 3.Nxe4 Qc7 4.Nxg5 Nb8 5.Rxd8+ Rxd8 6.Rxd8+ Qxd8 7.Nxf7 Qg8+ 8.Ng5 Qg6 9.Qf4 Nd7 10.Kg2 Qxc2 11.Nxe6 Qe2 12.Qf3 Qxf3+ 13.Kxf3 Nxe5+ 14.Ke4 Nc4 15.b3 Nd6+ (2.98)
15 +0.73 1.... Bxg2 2.Nde4 Bxe4 3.Nxe4 Qc7 4.Nxg5 Qb7 5.f3 Nb8 6.Kf2 Nc6 7.Ne4 Kb8 8.Rxd8+ Rxd8 9.Rxd8+ Nxd8 10.Nd6 Qd7 11.Qf4 Nc6 12.Qxf7 Nxe5 13.Qxd7 Nxd7 (4.80)
15 +0.60 1.... Qc7 2.Nc4 Kb8 3.Ne4 Bd5 4.R1xd5 exd5 5.Rxd5 Rde8 6.Ned6 Re7 7.a4 Qc6 8.Rd1 g4 9.Qf4 Qxa4 10.Qe4 Nf8 11.Qf5 (6.29)
16 +0.78 1.... Qc7 2.Nc4 (8.37)
16 +0.77 1.... Qc7 (12.22)
16 +0.83 1.... Qc7 2.Nc4 Kb8 3.b3 Rgf8 4.Ne4 Bd5 5.Rd3 Bxc4 6.bxc4 Nxe5 7.Nxc5 bxc5 8.Qxe5 Rxd6 9.Rxd6 Rg8 10.a4 g4 11.Rb6+ Kc8 12.Rxe6 Qxe5 13.Rxe5 Rd8 (15.20)
17 +0.99 1.... Qc7 2.Nc4 (21.20)
17 +0.69 1.... Bxg2 2.Nde4 Bxe4 3.Nxe4 Qc7 4.Nxg5 Qb7 5.f3 Rxg5+ 6.Qxg5 Qxf3 7.R6d3 Qe4 8.Qg2 Qxg2+ 9.Kxg2 Rg8+ 10.Rg3 Rxg3+ 11.Kxg3 Nxe5 12.Kf4 Nd7 13.Rg1 Kc7 14.Rg7 f5 15.Rg1 Kc6 16.Re1 (23.08)
18 +0.87 1.... Bxg2 2.f3 (27.75)
18 +1.40 1.... Bxg2 2.f3 (38.95)
18 +0.86 1.... Qc7 (42.02)
18 +0.85 1.... Qc7 2.Nc4 Kb8 3.b3 Rh8 4.Ne4 Bxe4 5.Qxe4 b5 6.Rxa6 Qb7 7.Qxb7+ Kxb7 8.Rad6 bxc4 9.Rxd7+ Rxd7 10.Rxd7+ Kc6 11.Rxf7 Ra8 12.a4 cxb3 13.cxb3 Kd5 14.Rg7 Rb8 15.Rd7+ Kxe5 16.Rd3 Ke4 17.Re3+ Kd4 18.Kf1 e5 (56.19)
19 +0.89 1.... Qc7 2.Nc4 Kb8 3.b3 Rh8 4.Ne4 Bxe4 5.Qxe4 b5 6.Rxa6 Qb7 7.Qxb7+ Kxb7 8.Rad6 bxc4 9.Rxd7+ Rxd7 10.Rxd7+ Kc6 11.Rxf7 Ra8 12.a4 cxb3 13.cxb3 Kd5 14.Rg7 Rb8 15.Rd7+ Kxe5 16.Rd3 c4 17.bxc4 Ra8 18.Ra3 Kd4 19.a5 Kxc4 (62.44)
20 +0.89 1.... Qc7 2.Nc4 Kb8 3.b3 Rh8 4.Ne4 Bxe4 5.Qxe4 b5 6.Rxa6 Qb7 7.Qxb7+ Kxb7 8.Rad6 bxc4 9.Rxd7+ Rxd7 10.Rxd7+ Kc6 11.Rxf7 Ra8 12.a4 cxb3 13.cxb3 Kd5 14.Rg7 Rb8 15.Rd7+ Kxe5 16.Rd3 c4 17.bxc4 Ra8 18.Ra3 Kd4 (73.69)
21 +0.85 1.... Qc7 2.Nc4 Kb8 3.b3 Rh8 4.Ne4 Bxe4 5.Qxe4 b5 6.Rxa6 Qb7 7.Qxb7+ Kxb7 8.Rad6 bxc4 9.Rxd7+ Rxd7 10.Rxd7+ Kc6 11.Rxf7 Ra8 12.bxc4 Ra4 13.Re7 Rxc4 14.Rxe6+ Kd5 15.Rd6+ Kxe5 16.Rd2 Ra4 17.c3 Kf6 18.Kf1 Ra3 19.c4 Ra4 20.Rd6+ Ke5 21.Rd5+ (88.12)
22 +0.87 1.... Qc7 2.Nc4 Kb8 3.a4 Rh8 4.Ne4 Rh4 5.Qxg5 Rxe4 6.Rxc6 f6 7.Rxc7 fxg5 8.Rb7+ Kxb7 9.Nd6+ Kc7 10.Nxe4 Nxe5 11.Rxd8 Kxd8 12.Kh2 Ke7 13.Kg3 Nc6 14.Kg4 Nb4 15.a5 bxa5 16.Kxg5 Nxc2 17.g4 c4 18.Nd2 a4 19.Nxc4 (125.48)
23 +1.02 1.... Qc7 2.Nc4 Kb8 3.a4 Rh8 4.Ne4 Rh4 5.Qxg5 Rxe4 6.Rxc6 f6 7.Rxc7 fxg5 8.Rb7+ Kxb7 9.Nd6+ Kc7 10.Nxe4 Nxe5 11.Rxd8 Kxd8 12.Kh2 Nc4 13.b3 Na3 14.c4 Nc2 15.Kg3 Nd4 16.Nd2 Ke7 17.Kg4 Kf6 18.a5 bxa5 19.Ne4+ Ke7 20.Nxc5 Kf6 21.f3 Kg6 (187.48)
Robert Flesher
Posts: 1280
Joined: Tue Aug 18, 2009 3:06 am

Re: Interesting blog post about Komodo...

Post by Robert Flesher »

New game
2kr2r1/1q1n1p2/ppbRp3/2p1P1p1/8/2N1Q3/PPPN1PP1/3R2K1 b - - 0 1

Analysis by Komodo64 3:

1...Qc7 2.Nc4 Kb8 3.R6d2 Rh8 4.Nd6 f5
+/= (0.54) Depth: 6 00:00:00 2kN
1...Qc7 2.Nc4 Kb8 3.R6d2 Rh8 4.Nd6 f5
+/= (0.54) Depth: 6 00:00:00 3kN
1...Qc7 2.Nc4 Kb8 3.R6d2 Rh8 4.Nd6 f5 5.Nf7
+/= (0.54) Depth: 7 00:00:00 3kN
1...Qc7 2.Nc4 Kb8 3.R6d2 Rh8 4.Nd6 f5 5.Nf7
+/= (0.54) Depth: 7 00:00:00 4kN
1...Qc7 2.Nc4 Kb8 3.R6d2 Rh8 4.Nd6 f5 5.Nf7 f4
+/= (0.54) Depth: 8 00:00:00 5kN
1...Qc7 2.Nc4 Kb8 3.R6d2 Rh8 4.Nd6 f5 5.Nf7 f4
+/= (0.54) Depth: 8 00:00:00 6kN
1...Qc7 2.Nc4 Kb8 3.R6d2 Rh8 4.Nd6 f5 5.Nf7 f4 6.Qe2
+/= (0.54) Depth: 9 00:00:00 9kN
1...Qc7 2.Nc4 Kb8 3.R6d2 Rh8 4.Nd6 f5 5.Nf7 f4 6.Qe2
+/= (0.54) Depth: 9 00:00:00 10kN
1...Qc7 2.Nc4
+/= (0.70) Depth: 10 00:00:00 20kN
1...Qc7 2.Nc4 Kb8 3.R6d2 Rh8 4.Nd6 Rh7 5.Qxg5 Rdh8 6.f3 Rh5
+/- (0.79) Depth: 10 00:00:00 59kN
1...Qc7 2.Nc4 Kb8 3.R6d2 Rh8 4.Nd6 Rh7 5.Qxg5 Rdh8 6.f3 Rh5
+/- (0.79) Depth: 10 00:00:00 75kN
1...Qc7
+/= (0.63) Depth: 11 00:00:00 95kN
1...Qc7 2.Nc4 Kb8 3.R6d3 Rh8 4.Nd6 f5 5.Nf7 c4 6.Nxd8 cxd3 7.Nxc6+ Qxc6 8.Rxd3 f4 9.Rd6 fxe3
+/= (0.60) Depth: 11 00:00:00 177kN
1...Qc7 2.Nc4 Kb8 3.R6d3 Rh8 4.Nd6 f5 5.Nf7 c4 6.Nxd8 cxd3 7.Nxc6+ Qxc6 8.Rxd3 f4 9.Rd6 fxe3
+/= (0.60) Depth: 11 00:00:00 201kN
1...Qc7 2.Nf1 Kb8 3.Ng3 c4 4.a4 a5 5.Nge4 Nxe5 6.Nf6 Rxd6 7.Rxd6 Qxd6 8.Qxb6+ Kc8 9.Nxg8
+/= (0.57) Depth: 12 00:00:00 424kN
1...Qc7 2.Nf1 Kb8 3.Ng3 c4 4.a4 a5 5.Nge4 Nxe5 6.Nf6 Rxd6 7.Rxd6 Qxd6 8.Qxb6+ Kc8 9.Nxg8
+/= (0.57) Depth: 12 00:00:00 444kN
1...Qc7 2.Nf1 Qb7 3.Ng3 Bxg2 4.Nce4 Bxe4 5.Nxe4 Qc7 6.Nxg5 Qb7 7.f3 Rxg5+ 8.Qxg5 Qxf3 9.R6d3 Qe2 10.Qg2
+/= (0.64) Depth: 13 00:00:00 657kN
1...Qc7 2.Nf1 Qb7 3.Ng3 Bxg2 4.Nce4 Bxe4 5.Nxe4 Qc7 6.Nxg5 Qb7 7.f3 Rxg5+ 8.Qxg5 Qxf3 9.R6d3 Qe2 10.Qg2
+/= (0.64) Depth: 13 00:00:00 779kN
1...Qc7 2.Nf1 Qb7 3.Ng3 Bxg2 4.Nce4 Bxe4 5.Nxe4 Qc7 6.Nxg5 Qb7 7.f3 Rxg5+ 8.Qxg5 Qxf3 9.R6d3 Qe2 10.Qg2 Qxg2+ 11.Kxg2 Rg8+ 12.Rg3
+/= (0.64) Depth: 14 00:00:00 1014kN
1...Qc7 2.Nf1 Qb7 3.Ng3 Bxg2 4.Nce4 Bxe4 5.Nxe4 Qc7 6.Nxg5 Qb7 7.f3 Rxg5+ 8.Qxg5 Qxf3 9.R6d3 Qe2 10.Qg2 Qxg2+ 11.Kxg2 Rg8+ 12.Rg3
+/= (0.64) Depth: 14 00:00:01 1137kN
1...Qc7 2.Nf1 Qb7 3.Ng3 Bxg2 4.Nce4 Bxe4 5.Nxe4 Qc7 6.Nxg5 Qb7 7.f3 Rxg5+ 8.Qxg5 Qxf3 9.R6d3 Qe2 10.Qg2 Qxg2+ 11.Kxg2 Rg8+ 12.Rg3 Rxg3+ 13.Kxg3
+/= (0.64) Depth: 15 00:00:01 1517kN
1...Qc7 2.Nf1 Qb7 3.Ng3 Bxg2 4.Nce4 Bxe4 5.Nxe4 Qc7 6.Nxg5 Qb7 7.f3 Rxg5+ 8.Qxg5 Qxf3 9.R6d3 Qe2 10.Qg2 Qxg2+ 11.Kxg2 Rg8+ 12.Rg3 Rxg3+ 13.Kxg3
+/= (0.64) Depth: 15 00:00:01 1783kN
1...Qc7 2.Nc4
+/- (0.80) Depth: 16 00:00:03 3413kN
1...Qc7 2.Nc4 Kb8 3.b3 Rh8 4.Ne4 b5 5.Nxg5 bxc4 6.Nxf7 Rdg8 7.Nxh8 Bxg2 8.Nf7 Bf3+ 9.Kf1 Bxd1 10.Rxd1 Nb6 11.Ng5 Nd5 12.Qg3 Qe7
+/- (0.93) Depth: 16 00:00:06 6970kN
1...Qc7 2.Nc4 Kb8 3.b3 Rh8 4.Ne4 b5 5.Nxg5 bxc4 6.Nxf7 Rdg8 7.Nxh8 Bxg2 8.Nf7 Bf3+ 9.Kf1 Bxd1 10.Rxd1 Nb6 11.Ng5 Nd5 12.Qg3 Qe7
+/- (0.93) Depth: 16 00:00:07 7878kN
1...Qc7
+/- (0.77) Depth: 17 00:00:09 10021kN
1...Qc7 2.Nc4
+/- (0.78) Depth: 17 00:00:10 11351kN
1...Qc7 2.Nc4 Kb8 3.b3 Ka7 4.a4 Rh8 5.f3 g4 6.fxg4 Rdg8 7.Qd2 Rd8 8.Ne4 Bxe4 9.Rxd7 Rxd7 10.Qxd7 Kb8 11.Rd6 Qxd7 12.Rxd7 Bxc2 13.Nxb6
+/- (0.79) Depth: 17 00:00:12 13538kN
1...Qc7 2.Nc4 Kb8 3.b3 Ka7 4.a4 Rh8 5.f3 g4 6.fxg4 Rdg8 7.Qd2 Rd8 8.Ne4 Bxe4 9.Rxd7 Rxd7 10.Qxd7 Kb8 11.Rd6 Qxd7 12.Rxd7 Bxc2 13.Nxb6
+/- (0.79) Depth: 17 00:00:13 14366kN
1...Qc7 2.Nc4 Kb8 3.b3 Ba8 4.a4 b5 5.axb5 axb5 6.Nb2 Qb7 7.Qg3 Rh8 8.f3 c4 9.bxc4 Qa7+ 10.R1d4 bxc4 11.Qxg5 Bxf3 12.Qe3 Ba8 13.Nxc4 Rhg8
+/- (0.97) Depth: 18 00:00:21 23469kN
1...Qc7 2.Nc4 Kb8 3.b3 Ba8 4.a4 b5 5.axb5 axb5 6.Nb2 Qb7 7.Qg3 Rh8 8.f3 c4 9.bxc4 Qa7+ 10.R1d4 bxc4 11.Qxg5 Bxf3 12.Qe3 Ba8 13.Nxc4 Rhg8
+/- (0.97) Depth: 18 00:00:25 28599kN
1...Qc7
+/- (0.81) Depth: 19 00:00:30 33897kN
1...Qc7 2.Nc4 Kb8 3.b3 Rh8 4.Ne4 Bxe4 5.Qxe4 b5 6.Rxa6 Qb7 7.Qxb7+ Kxb7 8.Rad6 bxc4 9.Rxd7+ Rxd7 10.Rxd7+ Kc6 11.Rxf7 Kd5 12.a4 cxb3 13.cxb3 Kxe5 14.Rg7 Kf6 15.Rc7 Rb8 16.a5 Rxb3 17.Rxc5 Rb1+ 18.Kh2
+/- (0.77) Depth: 19 00:00:32 35658kN
1...Qc7 2.Nc4 Kb8 3.b3 Rh8 4.Ne4 Bxe4 5.Qxe4 b5 6.Rxa6 Qb7 7.Qxb7+ Kxb7 8.Rad6 bxc4 9.Rxd7+ Rxd7 10.Rxd7+ Kc6 11.Rxf7 Kd5 12.a4 cxb3 13.cxb3 Kxe5 14.Rg7 Kf6 15.Rc7 Rb8 16.a5 Rxb3 17.Rxc5 Rb1+ 18.Kh2
+/- (0.77) Depth: 19 00:00:35 39568kN
1...Qc7 2.Nc4 Kb8 3.b3 Rh8 4.Ne4 Bxe4 5.Qxe4 b5 6.Rxa6 Qb7 7.Qxb7+ Kxb7 8.Rad6 bxc4 9.Rxd7+ Rxd7 10.Rxd7+ Kc6 11.Rxf7 Ra8 12.a4 cxb3 13.cxb3 g4 14.Kh2 Rb8 15.Re7 Kd5 16.Rd7+ Ke4 17.Kg3 Rxb3+ 18.Kxg4 Rb2 19.Kg3 Kxe5 20.Rd3
+/- (0.75) Depth: 20 00:00:41 46473kN
1...Qc7 2.Nc4 Kb8 3.b3 Rh8 4.Ne4 Bxe4 5.Qxe4 b5 6.Rxa6 Qb7 7.Qxb7+ Kxb7 8.Rad6 bxc4 9.Rxd7+ Rxd7 10.Rxd7+ Kc6 11.Rxf7 Ra8 12.a4 cxb3 13.cxb3 g4 14.Kh2 Rb8 15.Re7 Kd5 16.Rd7+ Ke4 17.Kg3 Rxb3+ 18.Kxg4 Rb2 19.Kg3 Kxe5 20.Rd3
+/- (0.75) Depth: 20 00:00:45 51185kN
1...Qc7 2.Nc4
+/- (0.91) Depth: 21 00:01:02 70978kN
1...Qc7
+/- (0.90) Depth: 21 00:01:21 92424kN
1...Qc7 2.Nc4 Kb8 3.b3 Rh8 4.Ne4 Bxe4 5.Qxe4 b5 6.Rxa6 Qb7 7.Qxb7+ Kxb7 8.Rad6 bxc4 9.Rxd7+ Rxd7 10.Rxd7+ Kc6 11.Rxf7 Ra8 12.a4 cxb3 13.cxb3 g4 14.Re7 Kd5 15.Rb7 Kxe5 16.Kh2 c4 17.Kg3 cxb3 18.Rb5+ Kf6 19.a5 b2 20.Rxb2 Rxa5 21.Kxg4 Rg5+ 22.Kf3
+/- (0.73) Depth: 21 00:01:31 105mN
1...Qc7 2.Nc4 Kb8 3.b3 Rh8 4.Ne4 Bxe4 5.Qxe4 b5 6.Rxa6 Qb7 7.Qxb7+ Kxb7 8.Rad6 bxc4 9.Rxd7+ Rxd7 10.Rxd7+ Kc6 11.Rxf7 Ra8 12.a4 cxb3 13.cxb3 g4 14.Re7 Kd5 15.Rb7 Kxe5 16.Kh2 c4 17.Kg3 cxb3 18.Rb5+ Kf6 19.a5 b2 20.Rxb2 Rxa5 21.Kxg4 Rg5+ 22.Kf3
+/- (0.73) Depth: 21 00:01:31 105mN
1...Qc7 2.Nc4 Kb8 3.b3 Rh8 4.Ne4 Bxe4 5.Qxe4 b5 6.Rxa6 Qb7 7.Qxb7+ Kxb7 8.Rad6 bxc4 9.Rxd7+ Rxd7 10.Rxd7+ Kc6 11.Rxf7 Ra8 12.a4 cxb3 13.cxb3 g4 14.Re7 Kd5 15.Rb7 Kxe5 16.Kh2 c4 17.Kg3 cxb3 18.Rb5+ Kf6 19.a5 b2 20.Rxb2 Rxa5 21.Kxg4 Rg5+ 22.Kf3 Rd5
+/- (0.73) Depth: 22 00:01:38 113mN
1...Qc7 2.Nc4 Kb8 3.b3 Rh8 4.Ne4 Bxe4 5.Qxe4 b5 6.Rxa6 Qb7 7.Qxb7+ Kxb7 8.Rad6 bxc4 9.Rxd7+ Rxd7 10.Rxd7+ Kc6 11.Rxf7 Ra8 12.a4 cxb3 13.cxb3 g4 14.Re7 Kd5 15.Rb7 Kxe5 16.Kh2 c4 17.Kg3 cxb3 18.Rb5+ Kf6 19.a5 b2 20.Rxb2 Rxa5 21.Kxg4 Rg5+ 22.Kf3 Rd5
+/- (0.73) Depth: 22 00:01:50 127mN
1...Qc7 2.Nc4 Kb8 3.b3 Rh8 4.Ne4 Bxe4 5.Qxe4 b5 6.Rxa6 Qb7 7.Qxb7+ Kxb7 8.Rad6 bxc4 9.Rxd7+ Rxd7 10.Rxd7+ Kc6 11.Rxf7 Ra8 12.a4 cxb3 13.cxb3 g4 14.Re7 Kd5 15.Rb7 Kxe5 16.Kh2 c4 17.Kg3 cxb3 18.Rb5+ Kf6 19.a5 b2 20.Rxb2 Rxa5 21.Kxg4 Rg5+ 22.Kf3 Rd5
+/- (0.73) Depth: 23 00:02:10 152mN
1...Qc7 2.Nc4 Kb8 3.b3 Rh8 4.Ne4 Bxe4 5.Qxe4 b5 6.Rxa6 Qb7 7.Qxb7+ Kxb7 8.Rad6 bxc4 9.Rxd7+ Rxd7 10.Rxd7+ Kc6 11.Rxf7 Ra8 12.a4 cxb3 13.cxb3 g4 14.Re7 Kd5 15.Rb7 Kxe5 16.Kh2 c4 17.Kg3 cxb3 18.Rb5+ Kf6 19.a5 b2 20.Rxb2 Rxa5 21.Kxg4 Rg5+ 22.Kf3 Rd5
+/- (0.73) Depth: 23 00:02:22 165mN
1...Qc7 2.Nc4
+/- (0.89) Depth: 24 00:02:54 202mN
1...Qc7
+/- (0.88) Depth: 24 00:04:05 286mN

(, Microsoft 10.10.2011)


If black is already in hot water, then this may be a bad test position. Black could be already lost. However, my version on my pc avoids Bxg2 pretty fast.
tomgdrums
Posts: 736
Joined: Wed Dec 02, 2009 9:48 am

Re: Interesting blog post about Komodo...

Post by tomgdrums »

It might be his particular machine or set up then.
rlsuth
Posts: 322
Joined: Wed Mar 08, 2006 9:37 pm

Re: Interesting blog post about Komodo...

Post by rlsuth »

I put it into my PC and it swapped over to Qc7 at a depth of 16 ply. That took about 16 seconds.
Albert Silver
Posts: 3019
Joined: Wed Mar 08, 2006 9:57 pm
Location: Rio de Janeiro, Brazil

Re: Interesting blog post about Komodo...

Post by Albert Silver »

tomgdrums wrote:This correspondence chess player doesn't seem to trust Komodo based on this and a previous blog post. I wonder if there is a bug in Komodo or something odd in his machine?


http://tartajubow.blogspot.com/2011/10/ ... rning.html


Edit: For the record his blog posts have made me think twice about Komodo for a little while.
For the record, he wrote "5-1/2 hours later I returned to discover Komodo had analyzed the position 13 moves deep", and you think Komodo is fishy, not the comment that it took 5 and a half hours to reach ply 13? I reached that depth in ZERO seconds and found ...Qc7 in four.


Analysis by Komodo64 3 SSE:

1...Bxg2
+/= (0.37) Depth: 13 00:00:00 941kN
1...Bxg2 2.Nc4 Kc7 3.Nxb6 Nxb6 4.Rxd8 Rxd8 5.Rxd8 Kxd8 6.Qxg5+ Kd7 7.Qxg2 Qxg2+ 8.Kxg2 Nc4 9.Kf3 Kc6 10.Ne4 Nxe5+ 11.Kf4 Nc4 12.b3 Na3
= (0.24) Depth: 13 00:00:00 1398kN
1...Bxg2 2.Nc4 Kc7 3.Nxb6 Nxb6 4.Rxd8 Rxd8 5.Rxd8 Kxd8 6.Qxg5+ Kd7 7.Qxg2 Qxg2+ 8.Kxg2 Nc4 9.Kf3 Kc6 10.Ne4 Nxe5+ 11.Kf4 Nc4 12.b3 Na3
= (0.24) Depth: 13 00:00:01 1491kN
1...Bxg2 2.Nde4
+/= (0.46) Depth: 14 00:00:01 1924kN
1...Bxg2 2.Nde4 Bxe4 3.Nxe4 Qc7 4.Nxg5 Nb8 5.Rxd8+ Rxd8 6.Rxd8+ Qxd8 7.Nxf7 Qg8+ 8.Ng5 Qg6 9.Qf4 Nd7 10.Kg2 Qxc2 11.Nxe6 Qe2 12.Qf3 Qxf3+ 13.Kxf3 Nxe5+ 14.Ke4 Nc4 15.b3 Nd6+
+/= (0.54) Depth: 14 00:00:02 3414kN
1...Bxg2 2.Nde4 Bxe4 3.Nxe4 Qc7 4.Nxg5 Nb8 5.Rxd8+ Rxd8 6.Rxd8+ Qxd8 7.Nxf7 Qg8+ 8.Ng5 Qg6 9.Qf4 Nd7 10.Kg2 Qxc2 11.Nxe6 Qe2 12.Qf3 Qxf3+ 13.Kxf3 Nxe5+ 14.Ke4 Nc4 15.b3 Nd6+
+/= (0.54) Depth: 14 00:00:02 3709kN
1...Bxg2 2.Nde4 Bxe4 3.Nxe4 Qc7 4.Nxg5 Qb7 5.f3 Nb8 6.Kf2 Nc6 7.Ne4 Kb8 8.Rxd8+ Rxd8 9.Rxd8+ Nxd8 10.Nd6 Qd7 11.Qf4 Nc6 12.Qxf7 Nxe5 13.Qxd7 Nxd7
+/- (0.73) Depth: 15 00:00:03 5521kN
1...Qc7 2.Nc4 Kb8 3.Ne4 Bd5 4.R1xd5 exd5 5.Rxd5 Rde8 6.Ned6 Re7 7.a4 Qc6 8.Rd1 g4 9.Qf4 Qxa4 10.Qe4 Nf8 11.Qf5
+/= (0.60) Depth: 15 00:00:04 7116kN
1...Qc7 2.Nc4 Kb8 3.Ne4 Bd5 4.R1xd5 exd5 5.Rxd5 Rde8 6.Ned6 Re7 7.a4 Qc6 8.Rd1 g4 9.Qf4 Qxa4 10.Qe4 Nf8 11.Qf5
+/= (0.60) Depth: 15 00:00:04 7240kN
1...Qc7 2.Nc4
+/- (0.78) Depth: 16 00:00:05 9324kN
1...Qc7
+/- (0.77) Depth: 16 00:00:08 13696kN
"Tactics are the bricks and sticks that make up a game, but positional play is the architectural blueprint."
Dave Mitchell
Posts: 115
Joined: Thu Nov 25, 2010 12:16 pm

Re: Interesting blog post about Komodo...

Post by Dave Mitchell »

Obviously, there was something using the computer cycles, besides Komodo, during that 5 hour period. Other software, a virus, a heat shutdown, or power down settings.

My vote would be for a dirty and hot system. The moment that Komodo starts turning up the temps, it shut down -- cools off -- starts up and shuts down again, etc. Then a power saving option kicked in and did the rest. Just a guess though.
User avatar
marcelk
Posts: 348
Joined: Sat Feb 27, 2010 12:21 am

Re: Interesting blog post about Komodo...

Post by marcelk »

Albert Silver wrote:For the record, he wrote "5-1/2 hours later I returned to discover Komodo had analyzed the position 13 moves deep", and you think Komodo is fishy, not the comment that it took 5 and a half hours to reach ply 13? I reached that depth in ZERO seconds and found ...Qc7 in four.
13 moves is 26 ply. Not always unreasonable after 5 1/2 hours depending on the system. The blogger claims that the engine switches back to Bxg2 in that time.
Albert Silver
Posts: 3019
Joined: Wed Mar 08, 2006 9:57 pm
Location: Rio de Janeiro, Brazil

Re: Interesting blog post about Komodo...

Post by Albert Silver »

marcelk wrote:
Albert Silver wrote:For the record, he wrote "5-1/2 hours later I returned to discover Komodo had analyzed the position 13 moves deep", and you think Komodo is fishy, not the comment that it took 5 and a half hours to reach ply 13? I reached that depth in ZERO seconds and found ...Qc7 in four.
13 moves is 26 ply. Not always unreasonable after 5 1/2 hours depending on the system. The blogger claims that the engine switches back to Bxg2 in that time.
Point taken, however he also wrote "Houdini 1.5 x64 recommends 1...Qc7 with White's advantage at about 1/3 of a Pawn."

Yet, when I ran it, not only is the eval nowhere near 1/3 of a pawn as he stated, but it actually switches to Kb8, meaning that Qc7 is not some universal choice either.

Analysis by Houdini 1.5a x64 Gaviota:

1...Qc7 2.Nc4 Kb8 3.a4 Rh8 4.Ne4 b5 5.axb5 axb5 6.Nxg5 bxc4 7.Nxf7 Rdg8 8.Nxh8 Bxg2 9.Nf7 Bf3+ 10.Kf1 Bxd1 11.Rxd1 Qb7 12.Qh3 Rf8 13.Qh5 Qb5 14.c3 Qb7 15.Rd6 Ka7 16.Qg6 Qh1+ 17.Ke2 Qf3+ 18.Kxf3 Nxe5+ 19.Kg3 Nxg6 20.Ng5 Nf4
+/- (0.77) Depth: 24/66 00:01:56 1245mN
1...Qc7 2.Nc4
+/- (0.86 --) Depth: 25/70 00:03:10 2073mN
1...Qc7 2.Nc4
+/- (1.04 --) Depth: 25/70 00:07:34 5048mN
1...Qc7 2.Nc4 Kb7 3.a4 Kb8 4.b3 g4 5.Ne4 b5 6.Nb2 Rh8 7.axb5 axb5 8.Nxc5 Nxc5 9.Rxd8+ Rxd8 10.Rxd8+ Qxd8 11.Qxc5 Kb7 12.Nd3 Qd5 13.Qxd5 Bxd5 14.Kh2 Kc6 15.Kg3 Be4 16.Nb4+ Kd7 17.Kxg4 Bxg2 18.f3 Kc7 19.Nd3 Kc6
+/- (1.03) Depth: 25/73 00:16:58 11466mN
1...Qc7 2.Nc4 Kb7 3.a3 Kb8 4.a4 g4 5.Qf4 Rgf8 6.b3 g3 7.f3 b5 8.Ne3 Nb6 9.axb5 axb5 10.Ne2 Nc8 11.Rxd8 Rxd8 12.Rxd8 Qxd8 13.Nxg3 Nb6 14.Qxf7 Qd4 15.Qg8+ Kc7 16.Qg7+
+/- (1.04) Depth: 26/73 00:20:55 14164mN
1...Qc7
+/- (0.95 ++) Depth: 27/73 00:26:42 18211mN
1...Qc7 2.Nc4 Kb7 3.a3 Kb8 4.a4 g4 5.Qf4 g3 6.fxg3 Rgf8 7.Ne4 Bxa4 8.Nf6 Bc6 9.Qd2 Kc8 10.Nxd7 Rxd7 11.Rxd7 Bxd7 12.Qxd7+ Qxd7 13.Nxb6+ Kc7 14.Nxd7 Rd8 15.Kf2 Rxd7 16.Rxd7+ Kxd7 17.Ke3 Kc6 18.Ke4 a5 19.g4 a4 20.g5
+/- (1.09) Depth: 27/73 00:36:46 24994mN
1...Kb8 2.Nde4 Rc8 3.Nxg5 Nf8 4.Nce4 Ng6 5.R6d2 Nh4 6.f3 Rg7 7.Qf4 Nxf3+ 8.gxf3 Bxe4 9.Rd7 Rc7 10.Rd8+ Rc8 11.R1d7 Rxd8 12.Rxd8+ Ka7 13.fxe4 Qe7 14.Qf6 Rxg5+ 15.Kf2 Qxf6+ 16.exf6 Kb7 17.Rd7+ Kc6 18.Rxf7 Rg6 19.Ke3 e5 20.b3 Rg3+ 21.Kf2 Rg6 22.a4 Kd6
+/- (1.08) Depth: 27/73 00:46:26 31494mN
"Tactics are the bricks and sticks that make up a game, but positional play is the architectural blueprint."