10 beautiful testpositions!!

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

Moderator: Ras

kgburcham
Posts: 2016
Joined: Sun Feb 17, 2008 4:19 pm

Re: 10 beautiful testpositions!!

Post by kgburcham »

Houdini making heat

Image
Dirt
Posts: 2851
Joined: Wed Mar 08, 2006 10:01 pm
Location: Irvine, CA, USA

Re: 10 beautiful testpositions!!

Post by Dirt »

Ignacio wrote:What was the last white´s move?.
Bh7? Followed by ...g6.
User avatar
George Tsavdaris
Posts: 1627
Joined: Thu Mar 09, 2006 12:35 pm

Re: 10 beautiful testpositions!!

Post by George Tsavdaris »

Dirt wrote:
Ignacio wrote:What was the last white´s move?.
Bh7? Followed by ...g6.
It could be, but in that case the draw is over for white.... :roll:
After his son's birth they've asked him:
"Is it a boy or girl?"
YES! He replied.....
User avatar
George Tsavdaris
Posts: 1627
Joined: Thu Mar 09, 2006 12:35 pm

Re: 10 beautiful testpositions!!

Post by George Tsavdaris »

kgburcham wrote: with a little faster
What exactly?

[d] r3r2b/1bpq4/2n1pPk1/1n1pP1p1/1p1P4/2P4Q/PP2NP1P/R4RK1 w - - 0 1

Houdini 1.5 x64:

1.Nf4+
= (0.06 ++) Depth: 20/61 00:00:08 276mN
1.Nf4+
= (0.19 ++) Depth: 20/61 00:00:09 304mN
1.Nf4+
+/= (0.48 ++) Depth: 20/68 00:00:12 376mN
1.Nf4+ gxf4 2.Kh1 Nxe5 3.dxe5 d4+ 4.f3 Bxf3+
+/- (1.07) Depth: 20/77 00:00:18 585mN
1.Nf4+
+/- (1.16 ++) Depth: 21/77 00:00:27 846mN
1.Nf4+
+/- (1.34 ++) Depth: 21/79 00:00:35 1111mN
1.Nf4+ gxf4 2.Kh1 Nxe5 3.dxe5 d4+ 4.f3 Bxf3+
+/- (1.27) Depth: 21/79 00:00:44 1427mN
1.Nf4+ gxf4 2.Kh1 Nxe5 3.dxe5 d4+ 4.f3 Bxf3+
+/- (1.35) Depth: 22/79 00:01:06 2237mN
1.Nf4+ gxf4 2.Kh1 Nxe5 3.dxe5 d4+ 4.f3 Bxf3+
+/- (1.30) Depth: 23/79 00:01:38 3395mN
Thanks but what about the initial position?
Does it find 1.Bg6+!! ?
[d]r3r2b/1bpq1k2/2n1pP2/1n1pP1p1/1p1P4/2PB4/PP1QNP1P/R4RK1 w - - 0 1

How Houdini on this very fast hardware fares?
(With clear hash of course.)
After his son's birth they've asked him:
"Is it a boy or girl?"
YES! He replied.....
User avatar
George Tsavdaris
Posts: 1627
Joined: Thu Mar 09, 2006 12:35 pm

Re: 10 beautiful testpositions!!

Post by George Tsavdaris »

GARGAMEL wrote:Hello

so, now the correct position......

NN - NN, Teststellung, Df6
r2q1r2/1b2bpkp/n3p1p1/2ppP1P1/p6R/1PN1BQR1/NPP2P1P/4K3 w - - 0 1

Analysis by Houdini 1.5 x64:
1.Tgh3 h5 2.gxh6+ Kh8 3.Txa4 Dd7 4.Lf4 Lc6 5.De2 Lxa4 6.bxa4 Sb4 7.b3 Sc6 8.Kf1 Sd4 9.Dd2
= (-0.05) Tiefe: 6/16 00:00:00 5kN
-/+ (-0.81) Tiefe: 10/28 00:00:00 316kN
1.bxa4 Sc7 2.Ld2 Th8 3.b4 Kg8 4.bxc5 Sa6 5.Le3 h5 6.Sb5 Sxc5 7.Lxc5 Lxc5 8.Sac3 Le7 9.De3 Tc8 10.Kf1 Kg7 11.Kg1
=/+ (-0.70) Tiefe: 10/28 00:00:00 446kN
= (-0.16) Tiefe: 25/64 00:14:38 11913mN
1.Df6!!+ Lxf6 2.gxf6+ Dxf6 3.Lh6+ Kg8 4.exf6 d4 5.Lg7 h5 6.Tgh3 Kh7 7.Se2 e5 8.Lxf8 Txf8 9.bxa4 Lc8 10.Tg3 Lf5 11.Kd2 Kh6 12.f4 Sc7 13.Tg5 Sd5 14.Txf5 gxf5 15.fxe5 Ta8 16.Sac1 Txa4 17.Sd3 Kg5 18.Th3 Se3 19.Tg3+ Kh6 20.Tg7 Ta1 21.Sef4 c4 22.Se1 Sg4 23.Txf7 c3+
= (0.00) Tiefe: 25/66 00:18:59 15624mN
+- (5.02) Tiefe: 29/77 01:47:48 94346mN

regards Anton
Thanks.
So in the modified position it does 19 minutes while in the original(from the game) one it does 3 minutes.

6.3 times longer! Thank you for the analysis, i tried to make the position harder but i didn't think it would become so much harder. :D

And what hardware did you use for the analysis?
Last edited by George Tsavdaris on Sat Dec 25, 2010 11:45 am, edited 1 time in total.
After his son's birth they've asked him:
"Is it a boy or girl?"
YES! He replied.....
User avatar
George Tsavdaris
Posts: 1627
Joined: Thu Mar 09, 2006 12:35 pm

Re: 10 beautiful testpositions!!

Post by George Tsavdaris »

Robert Flesher wrote:
George Tsavdaris wrote:
zullil wrote:
George Tsavdaris wrote:
zullil wrote:Strange. Crafty finds it in 5 minutes:
Why do you think it's strange?
I guess I've come to expect that Houdini solves all problems very quickly. But having now tried several other engines, I am coming to the conclusion that Crafty was either especially smart or especially SMP-lucky. :)

[EDIT] See this: http://www.talkchess.com/forum/viewtopi ... 25&t=37265
He analyzed wrong position. See my reply there.
Perhaps of course Houdini on my position also solves it fast in his hardware, but we will see.

His hardware cannot be that fast. My i7 920 solves it alot faster.


FEN: r2q1r2/1b2bpkp/n3p1p1/2ppP1P1/p6R/1PN1BQR1/NPP2P1P/4K3 w - - 0 1

Houdini_15_x64:
24/68 08:21 3,886,174,626 7,742,000 -0.10 b3xa4 Qd8d7 Be3d2 Rf8h8 b2b3 Kg7g8 Qf3f4 Na6c7 Ke1f1 d5d4 Nc3e4 Nc7d5 Qf4g4 Qd7c7 Na2c1 Qc7xe5 Nc1d3 Qe5c7 Bd2f4 Nd5xf4 Qg4xf4 Qc7xf4 Rh4xf4 Bb7xe4 Rf4xe4 h7h6 g5xh6 Rh8xh6 Kf1g2 Kg8g7 Rg3h3 Rh6xh3 Kg2xh3 c5c4 b3xc4 Ra8xa4 Re4xd4 Be7f6
24/68+ 09:17 4,366,588,336 7,834,000 -0.03 Qf3f6+
24/68+ 09:48 4,640,956,711 7,880,000 +1.40 Qf3f6+
24/73 12:23 5,941,404,997 7,986,000 +4.52 Qf3f6+ Be7xf6 g5xf6+ Qd8xf6 e5xf6+ Kg7xf6 Nc3xa4 h7h5 Na4xc5 Na6xc5 Be3xc5 Rf8e8 Rh4f4+ Kf6g7 Rg3f3 f7f5 Rf4b4 Bb7c8 Rb4a4 Ra8b8 Na2c3 g6g5 b3b4 Kg7f6 Rf3d3 Bc8b7 b4b5 Rb8c8 b2b4
Thanks for the analysis, but what hardware did you use for the analysis?
After his son's birth they've asked him:
"Is it a boy or girl?"
YES! He replied.....
Uri Blass
Posts: 10890
Joined: Thu Mar 09, 2006 12:37 am
Location: Tel-Aviv Israel

Re: 10 beautiful testpositions!!

Post by Uri Blass »

I am disappointed by houdini.

I read that rybka cluster solved position number 1 and 2 relatively fast(near 5 minutes)

I hoped to see houdini1.5 4 cores 32 bits solving it in less than an hour and it did not do it.


FEN: 2K3k1/1p6/R3p1p1/1rB1P1P1/8/8/1Pb5/8 w - - 0 1

Houdini_15_w32:
4/11 00:00 1.114 42.000 -0.87 Ra6xe6 Rb5xc5+ Kc8xb7 Bc2f5 Kb7b6
5/14 00:00 1.878 69.000 -0.97 Ra6xe6 Rb5xc5+ Kc8xb7 Bc2f5 Kb7b6 Rc5c2
6/16 00:00 5.701 158.000 -1.52 Ra6xe6 Rb5xc5+ Kc8d8 Rc5d5+ Kd8c7 Kg8f7 Re6f6+ Kf7e7
7/17 00:00 10.398 81.000 -2.15 Ra6a8 Rb5xc5+ Kc8xb7+ Kg8f7 Kb7b6 Rc5xe5 Ra8a7+ Kf7e8 b2b4 Re5xg5 b4b5 Bc2e4
8/17 00:00 11.989 90.000 -2.15 Ra6a8 Rb5xc5+ Kc8xb7+ Kg8f7 Kb7b6 Rc5xe5 Ra8a7+ Kf7e8 b2b4 Re5xg5 b4b5 Bc2e4
9/20 00:00 31.118 219.000 -2.44 Ra6a8 Rb5xc5+ Kc8xb7+ Kg8f7 b2b4 Rc5xe5 Ra8a5 Bc2e4+ Kb7b6 Re5f5 Ra5a7+ Kf7e8 b4b5 Rf5xg5 Ra7c7 Rg5f5
10/21 00:00 48.561 321.000 -2.44 Ra6a8 Rb5xc5+ Kc8xb7+ Kg8f7 b2b4 Rc5xe5 Ra8a5 Bc2e4+ Kb7b6 Re5f5 Ra5a7+ Kf7e8 b4b5 Rf5xg5 Ra7c7 Rg5f5
11/24- 00:00 87.115 521.000 -2.56 Ra6a8 Rb5xc5+
11/24 00:00 136.832 735.000 -2.43 Ra6a8 Rb5xc5+ Kc8xb7+ Kg8f7 b2b4 Bc2e4+ Kb7b6 Rc5xe5 Ra8a7+ Kf7e8 Ra7a5 Re5f5 Ra5c5 e6e5 Rc5c7 Rf5xg5 b4b5 Rg5g3
12/25- 00:00 161.443 823.000 -2.55 Ra6a8 Rb5xc5+
12/27- 00:00 234.669 1.052.000 -2.79 Ra6a8 Rb5xc5+
12/33 00:00 476.628 1.134.000 -2.87 Ra6a8 Rb5xc5+ Kc8xb7+ Kg8f7 b2b4 Bc2e4+ Kb7b6 Rc5xe5 Ra8a7+ Kf7e8 Ra7a5 Re5f5 Kb6c7 Ke8e7 Ra5c5 e6e5 b4b5 Ke7e6 Rc5c4 Be4d5 Rc4c3 e5e4 Rc3e3 Ke6e5 b5b6 Rf5f7+ Kc7c8 Bd5b7+ Kc8b8
13/33 00:00 617.974 1.346.000 -2.87 Ra6a8 Rb5xc5+ Kc8xb7+ Kg8f7 b2b4 Bc2e4+ Kb7b6 Rc5xe5 Ra8a7+ Kf7e8 Ra7a5 Re5f5 Kb6c7 Ke8e7 Ra5c5 e6e5 b4b5 Ke7e6 Rc5c4 Be4d5 Rc4c3 e5e4 Rc3e3 Ke6e5 b5b6 Rf5f7+ Kc7c8 Bd5b7+ Kc8b8
14/33 00:01 877.525 1.497.000 -2.84 Ra6a8 Rb5xc5+ Kc8xb7+ Kg8f7 Ra8a7 Kf7e8 b2b4 Bc2e4+ Kb7b8 Rc5xe5 Ra7a5 Re5f5 Kb8c8 Ke8e7 Kc8c7 Rf5d5 b4b5 Rd5xg5 Ra5a4 Rg5c5+ Kc7b6 Rc5e5 Kb6c7 g6g5 b5b6 Re5c5+ Kc7b8 Be4c6 Ra4a2
15/33 00:01 1.115.319 1.652.000 -2.90 Ra6a8 Rb5xc5+ Kc8xb7+ Kg8f7 Ra8a7 Kf7e8 b2b4 Bc2e4+ Kb7b6 Rc5xe5 Ra7a5 Re5f5 Ra5c5 Ke8d7 b4b5 Kd7d6 Rc5c4 Be4d5 Rc4c8 Rf5xg5 Rc8d8+ Kd6e5 Kb6c5 Rg5g3 b5b6 g6g5
16/33 00:01 1.829.193 2.048.000 -2.90 Ra6a8 Rb5xc5+ Kc8xb7+ Kg8f7 Ra8a7 Kf7e8 b2b4 Bc2e4+ Kb7b6 Rc5xe5 Ra7a5 Re5f5 Ra5c5 Ke8d7 b4b5 Kd7d6 Rc5c4 Be4d5 Rc4c8 Rf5xg5 Rc8d8+ Kd6e5 Kb6c5 Rg5g3 b5b6 g6g5
17/33 00:01 2.197.682 2.184.000 -2.90 Ra6a8 Rb5xc5+ Kc8xb7+ Kg8f7 Ra8a7 Kf7e8 b2b4 Bc2e4+ Kb7b6 Rc5xe5 Ra7a5 Re5f5 Ra5c5 Ke8d7 b4b5 Kd7d6 Rc5c4 Be4d5 Rc4c8 Rf5xg5 Rc8d8+ Kd6e5 Kb6c5 Rg5g3 b5b6 g6g5
18/38- 00:01 3.744.893 2.767.000 -3.04 Ra6a8 Rb5xc5+
18/41 00:02 6.367.966 3.190.000 -3.03 Ra6a8 Rb5xc5+ Kc8xb7+ Kg8f7 Ra8a7 Kf7e8 b2b4 Bc2e4+ Kb7b6 Rc5xe5 Ra7a5 Re5f5 b4b5 Rf5xg5 Kb6c7 Rg5d5 Ra5a4 Rd5d7+ Kc7b6 Be4d5 Kb6c5 Rd7c7+ Kc5d6 Rc7b7 Kd6c5 Ke8e7 b5b6 g6g5 Ra4g4 Ke7f6 Rg4g1 Kf6f5 Rg1f1+ Kf5e5 Rf1e1+ Bd5e4
19/41 00:02 7.052.429 3.192.000 -3.03 Ra6a8 Rb5xc5+ Kc8xb7+ Kg8f7 Ra8a7 Kf7e8 b2b4 Bc2e4+ Kb7b6 Rc5xe5 Ra7a5 Re5f5 b4b5 Rf5xg5 Kb6c7 Rg5d5 Ra5a4 Rd5d7+ Kc7b6 Be4d5 Kb6c5 Rd7c7+ Kc5d6 Rc7b7 Kd6c5 Ke8e7 b5b6 g6g5 Ra4g4 Ke7f6 Rg4g1 Kf6f5 Rg1f1+ Kf5e5 Rf1e1+ Bd5e4
20/46- 00:03 11.921.713 3.662.000 -3.18 Ra6a8 Rb5xc5+
20/49- 00:05 22.111.188 4.044.000 -3.48 Ra6a8 Rb5xc5+
20/53 00:10 44.274.635 4.610.000 -3.54 Ra6a8 Rb5xc5+ Kc8xb7+ Kg8f7 Kb7b6 Rc5xe5 Ra8c8 Bc2e4 Rc8c5 Re5f5 Rc5c4 Be4d3 Rc4c7+ Kf7e8 Kb6c6 Rf5xg5 Kc6d6 Rg5b5 Kd6xe6 Rb5b6+ Ke6d5 Rb6xb2 Kd5e5 Bd3f5 Rc7a7 Rb2f2 Ke5f6 Rf2d2 Ra7a6 Ke8d7 Kf6g5 Kd7e7
21/53 00:13 61.531.516 4.818.000 -3.61 Ra6a8 Rb5xc5+ Kc8xb7+ Kg8f7 Kb7b6 Rc5xe5 Ra8c8 Bc2e4 Rc8c5 Re5f5 Rc5c4 Be4d3 Rc4c7+ Kf7e8 Kb6c6 Rf5xg5 Kc6d6 Rg5b5 Kd6xe6 Rb5b6+ Ke6e5 Rb6xb2 Rc7a7 Bd3f5 Ke5f6 Rb2d2 Ra7e7+ Ke8d8 Re7a7 Rd2d6+ Kf6g5 Kd8e8 Ra7a3 Ke8d7 Ra3a7+ Kd7e6
22/53- 00:22 112.496.335 5.079.000 -3.78 Ra6a8 Rb5xc5+
22/53 00:25 125.789.277 5.120.000 -3.73 Ra6b6 Rb5xc5+ Kc8d7 Rc5xe5 Rb6xb7 Bc2e4 Rb7c7 Re5xg5 Kd7xe6 Rg5b5 Ke6f6 Rb5xb2 Rc7g7+ Kg8h8 Rg7e7 Rb2f2+ Kf6g5 Be4f5 Kg5f6 Kh8g8 Re7g7+ Kg8f8 Rg7f7+ Kf8e8 Rf7e7+ Ke8d8 Re7a7 Rf2d2 Ra7h7 Rd2d7 Rh7h4 Rd7d6+ Kf6g5 Kd8e7 Rh4b4 Rd6d7 Rb4b5 Rd7d4 Rb5b7+ Ke7e6 Rb7b6+ Ke6e5
23/53 00:34 175.638.455 5.224.000 -3.74 Ra6b6 Rb5xc5+ Kc8d7 Rc5xe5 Rb6xb7 Bc2e4 Rb7c7 Re5xg5 Kd7xe6 Rg5b5 Ke6f6 Rb5xb2 Rc7g7+ Kg8h8 Rg7e7 Rb2f2+ Kf6g5 Be4f5 Kg5f6 Kh8g8 Re7g7+ Kg8f8 Rg7f7+ Kf8e8 Rf7a7 Rf2e2 Ra7a8+ Ke8d7 Ra8a7+ Kd7c6 Ra7a6+ Kc6d5 Kf6g5 Re2g2+ Kg5f6 Rg2f2 Kf6g5 Rf2e2 Ra6a5+ Kd5d4 Ra5a6 Kd4e5 Ra6a5+ Ke5d6 Ra5a6+ Kd6d5 Ra6a5+
24/53 00:46 242.072.919 5.317.000 -3.74 Ra6b6 Rb5xc5+ Kc8d7 Rc5xe5 Rb6xb7 Bc2e4 Rb7c7 Re5xg5 Kd7xe6 Rg5b5 Ke6f6 Rb5xb2 Rc7g7+ Kg8h8 Rg7e7 Rb2f2+ Kf6g5 Be4f5 Kg5f6 Kh8g8 Re7g7+ Kg8f8 Rg7f7+ Kf8e8 Rf7a7 Rf2e2 Ra7a6 Ke8d7 Ra6a7+ Kd7c6 Kf6g5 Re2e5 Ra7a6+ Kc6b5 Ra6a7 Kb5b6 Ra7a3 Bf5e4+ Kg5f4
25/53 01:07 361.963.002 5.388.000 -3.80 Ra6b6 Rb5xc5+ Kc8d7 Rc5xe5 Rb6xb7 Bc2e4 Rb7c7 Re5xg5 Kd7xe6 Rg5b5 b2b4 Rb5xb4 Ke6f6 Rb4b6+ Kf6e5 Be4f5 Rc7a7 Kg8f8 Ke5f4 Rb6e6 Kf4g5 Re6e2 Ra7h7 Re2e7 Rh7h2 Kf8g7 Rh2b2 Bf5d3 Rb2b3 Bd3c4 Rb3b6 Re7e5+ Kg5f4 Re5f5+ Kf4g4 Bc4d5 Rb6d6 Rf5h5 Kg4f4 Kg7f7 Kf4g4 Bd5e6+ Kg4f4 Rh5h4+ Kf4e5
26/53 01:54 626.615.325 5.478.000 -3.84 Ra6b6 Rb5xc5+ Kc8d7 Rc5xe5 Rb6xb7 Bc2e4 Rb7c7 Re5xg5 Kd7d6 Rg5d5+ Kd6xe6 Rd5b5 b2b4 Rb5xb4 Ke6f6 Rb4b6+ Kf6e5 Be4f5 Rc7a7 Kg8f8 Ke5f4 Rb6e6 Kf4g5 Re6e7 Ra7a6 Kf8g7 Ra6b6 Re7f7 Rb6c6 Bf5d3 Rc6e6 Rf7f5+ Kg5h4 Kg7f7
27/58 03:23 1.125.844.493 5.535.000 -3.92 Ra6b6 Rb5xc5+ Kc8d7 Rc5xe5 Rb6xb7 Bc2e4 Rb7a7 Re5xg5 Kd7xe6 Rg5b5 Ra7c7 Rb5xb2 Ke6e5 Be4f5 Ke5f6 Rb2e2 Kf6g5 Kg8f8 Rc7a7 Re2e7 Ra7a6 Kf8g7 Ra6b6 Bf5e4 Rb6d6 Re7f7 Kg5g4 Rf7f6 Rd6d7+ Kg7h6 Rd7e7 Be4d5 Re7e3 Kh6g7
28/60 05:17 1.759.630.918 5.558.000 -3.92 Ra6b6 Rb5xc5+ Kc8d7 Rc5xe5 Rb6xb7 Bc2e4 Rb7a7 Re5xg5 Kd7xe6 Rg5b5 Ra7c7 Rb5xb2 Ke6e5 Be4f5 Ke5f6 Rb2e2 Kf6g5 Kg8f8 Rc7a7 Re2e7 Ra7a6 Kf8g7 Ra6b6 Bf5e4 Rb6a6 Re7f7 Kg5h4 Rf7f6 Ra6a5 Rf6f5 Ra5a7+ Kg7h6 Kh4g4 Be4d5 Ra7a6 Kh6g7 Ra6d6 Kg7f7 Kg4g3 Rf5e5 Kg3f4 Re5f5+
29/60- 08:41 2.916.791,761 5.598.000 -4.10 Ra6b6 Rb5xc5+
29/65 15:50 5.375.261,601 5.659.000 -4.17 Ra6b6 Rb5xc5+ Kc8d7 Rc5xe5 Rb6xb7 Bc2e4 Rb7a7 Re5xg5 Kd7xe6 Rg5b5 Ra7a4 Be4d5+ Ke6f6 Rb5b6+ Kf6e5 Bd5f7 b2b4 Bf7b3 Ra4a7 Rb6xb4 Ke5f6 Rb4b6+ Kf6e5 Rb6b5+ Ke5f4 g6g5+ Kf4g3 Bb3f7 Ra7a1 Kg8g7 Ra1a6 Rb5e5 Ra6a3 Kg7f6 Ra3f3+ Re5f5 Rf3e3 Rf5f4 Re3d3 Kf6e5 Rd3e3+ Ke5f5 Re3d3 Rf4g4+ Kg3f3 Rg4e4
30/65 25:52 8.834.247,827 5.691.000 -4.19 Ra6b6 Rb5xc5+ Kc8d7 Rc5xe5 Rb6xb7 Bc2e4 Rb7a7 Re5xg5 Kd7xe6 Rg5b5 Ra7c7 Rb5xb2 Ke6e5 Be4b1 Rc7a7 Bb1a2 Ke5f4 Ba2f7 Ra7a5 Kg8g7 Ra5a6 Rb2b5 Kf4g3 Bf7d5 Kg3h4 Kg7f7 Kh4g4 Bd5e6+ Kg4f3 Kf7f6 Kf3e3 Rb5b4 Ra6a5 g6g5 Ke3f2 Rb4b2+ Kf2g3 Rb2a2 Ra5xa2 Be6xa2 Kg3g4 Ba2e6+ Kg4f3 Kf6e5 Kf3g3

13 million nodes 37 minutes 10 seconds no solution by houdini and I believe that it also cannot solve it in depth 31 inspite of using the gaviota tablebases because the evaluation goes down too slowly.

houdini seems to be able to solve position number 2 but it needs more than 2 hours for it(considering Re6 for more than 5 million nodes is a good sign).


FEN: 2bq1b2/1p3k2/1r1p1p1p/pNpP1PpP/P1P3P1/6B1/1P5Q/2K1R3 w - - 0 1

Houdini_15_w32:
4/17 00:00 1.052 1.052.000 +1.72 Qh2e2 Bf8e7 Kc1b1 Bc8d7
5/17 00:00 1.501 750.000 +1.42 Qh2e2 Bf8e7 Kc1b1 Bc8d7 Kb1a2 Bd7xb5 a4xb5
5/17 00:00 2.649 883.000 +1.88 Qh2d2 Bf8e7 Qd2xa5 Kf7f8 Kc1b1 Bc8d7
6/23 00:00 5.458 909.000 +1.56 Qh2d2 Bc8d7 Qd2xa5 Bd7xb5 a4xb5 Bf8e7 Qa5c3 Kf7g7 Kc1b1
7/23 00:00 6.924 989.000 +1.56 Qh2d2 Bc8d7 Qd2xa5 Bd7xb5 a4xb5 Bf8e7 Qa5c3 Kf7g7 Kc1b1
8/23 00:00 9.447 1.180.000 +1.66 Qh2d2 Bc8d7 Qd2xa5 Bd7xb5 a4xb5 Bf8e7 Qa5c3 Kf7g8 Kc1b1
9/23+ 00:00 13.497 1.349.000 +1.76 Qh2d2
9/23 00:00 33.470 1.761.000 +1.50 Qh2d2 Rb6a6 Qd2e2 Bf8e7 Kc1b1 Ra6b6 Qe2e3 Bc8d7 Kb1a2 Bd7xb5 a4xb5
10/23+ 00:00 43.179 1.799.000 +1.61 Qh2d2
10/23+ 00:00 45.870 1.764.000 +1.82 Qh2d2
10/23 00:00 53.315 1.904.000 +1.50 Qh2d2 Rb6a6 Qd2e2 Bf8e7 Kc1b1 Ra6b6 Qe2d2 Rb6a6 Qd2e3 Bc8d7 Bg3h2 Kf7f8 Kb1a2 Bd7xb5 a4xb5
11/23 00:01 94.311 2.143.000 +1.45 Qh2d2 Rb6a6 Qd2c3 Bf8e7 Kc1b1 Bc8d7 Bg3h2 Kf7g7 Qc3e3 Kg7f8 Kb1a2 Bd7xb5 a4xb5 Ra6b6
12/23 00:01 191.326 2.620.000 +1.43 Qh2d2 Rb6a6 Re1e2 Bf8e7 Qd2e3 Kf7f8 Bg3e1 b7b6 Kc1b1 Bc8d7 Kb1a2 Bd7xb5 a4xb5 Ra6a8
13/24- 00:01 349.226 3.036.000 +1.32 Qh2d2 Rb6a6
13/24+ 00:01 400.082 3.101.000 +1.53 b2b3
13/24 00:01 424.406 3.143.000 +1.40 b2b3 Bf8e7 Kc1b2 Rb6a6 Kb2a3 Ra6b6 Qh2d2 Bc8d7 Qd2xa5 Bd7xb5 c4xb5 Kf7f8 Qa5c3 Kf8g7
14/27+ 00:01 542.129 3.115.000 +1.50 b2b3
14/29 00:01 563.768 3.114.000 +1.50 b2b3 Bf8e7 Kc1b2 Rb6a6 Kb2a3 Ra6b6 Qh2d2 Bc8d7 Qd2xa5 Bd7xb5 c4xb5 Kf7f8 Qa5c3 Qd8d7 Qc3e3 Be7d8 Qe3e6
15/30 00:01 848.580 3.289.000 +1.50 b2b3 Bf8e7 Kc1b2 Rb6a6 Kb2a1 Ra6b6 Ka1a2 Kf7g8 Ka2a3 Kg8f7 Qh2d2 Bc8d7 Qd2xa5 Bd7xb5 c4xb5 Kf7f8 Qa5c3 Qd8d7 Qc3e3 Be7d8 Qe3e6
16/30 00:01 1.361.044 3.311.000 +1.42 b2b3 Bf8e7 Kc1b2 Rb6a6 Kb2a1 Ra6b6 Ka1a2 Kf7g8 Qh2d2 Bc8d7 Ka2b2 Rb6a6 Qd2e2 Kg8f7 Qe2e4 Kf7f8 Bg3h2 Bd7xb5 a4xb5 Ra6b6
17/35+ 00:01 2.653.804 3.487.000 +1.53 b2b3
17/35 00:01 2.854.237 3.447.000 +1.47 b2b3 Bf8e7 Kc1b2 Rb6a6 Qh2e2 Bc8d7 Nb5c7 Ra6b6 Nc7e6 Qd8e8 Ne6g7 Qe8d8 Kb2a2 Rb6b4 Ng7e6 Bd7xe6 f5xe6+ Kf7g8 Qe2d3 Qd8f8
18/37 00:02 6.558.984 4.048.000 +1.57 b2b3 Bf8e7 Kc1b2 Rb6a6 Qh2e2 Kf7f8 Qe2h2 Kf8f7 Qh2e2
19/37 00:02 8.840.449 4.166.000 +1.54 b2b3 Bf8e7 Kc1b2 Rb6a6 Qh2e2 Bc8d7 Nb5c7 Ra6b6 Nc7e6 Qd8e8 Ne6g7 Qe8d8 Kb2a2 Rb6b4 Ng7e6 Bd7xe6 f5xe6+ Kf7g8 Qe2d3 Qd8f8 Re1f1 Qf8g7 Bg3e1
20/48 00:04 19.595.391 4.546.000 +1.55 b2b3 Rb6a6 Kc1b2 Qd8b6 Re1e4 Qb6d8 Re4e3 Bf8e7 Re3e1 Ra6b6 Qh2d2 Rb6a6 Qd2e3 Bc8d7 Nb5c7 Ra6b6 Nc7e6 Bd7xe6 f5xe6+ Kf7g8 Qe3d3 Qd8e8 Re1f1 Rb6b4 Bg3e1
21/48- 00:07 35.121.946 4.686.000 +1.44 b2b3 Rb6a6
21/57 00:16 72.549.475 4.651.000 +1.54 Kc1b1 Bf8e7 Kb1a2 Rb6a6 b2b3 Be7f8 Ka2a3 Kf7g7 Ka3b2 Kg7g8 Re1e3 Kg8f7 Kb2a2 Bf8e7 Re3e1 Be7f8
22/57 00:21 96.621.733 4.702.000 +1.50 Kc1b1 Bf8e7 b2b3 Rb6a6 Kb1a1 Kf7f8 Ka1a2 Kf8f7 Ka2b2 Be7f8 Kb2a3 Kf7g7 Ka3a2 Ra6b6 Re1e3 Rb6a6 Ka2b2 Kg7f7 Re3e1 Qd8b6 Re1e4 Bc8d7 Re4e3 Kf7g7 Kb2a3 Kg7f7 Re3e4 Ra6a8 Qh2e2 Kf7g8 Bg3e1 Bd7xb5 a4xb5
23/57 00:33 155.538.353 4.780.000 +1.50 Kc1b1 Rb6a6 Kb1a2 Qd8b6 Re1e2 Qb6d8 b2b3 Bf8e7 Re2e1 Be7f8 Ka2b2 Qd8b6 Re1e4 Bc8d7 Re4e3 Kf7g7 Kb2a3 Kg7f7 Re3e4 Ra6a8 Qh2e2 Kf7g8 Bg3e1 Bd7xb5 a4xb5
24/57 00:45 217.465.658 4.798.000 +1.45 Kc1b1 Rb6a6 Kb1a2 Qd8b6 Re1e2 Bc8d7 Re2e3 Kf7g8 Qh2e2 Qb6d8 Re3e4 Kg8f7 Qe2h2 Kf7g8 Re4e1 Bf8e7 Qh2e2 Be7f8 Qe2e3 Kg8f7 Qe3d3 Kf7g8 Re1e3 Kg8g7 Qd3d2 Bf8e7 Qd2e2 Kg7f8 Nb5c7 Qd8xc7 Re3xe7 Kf8g8 Bg3e1 Ra6a8 Qe2e6+ Bd7xe6 Re7xc7 Be6f7 Rc7xb7 Kg8g7 Rb7b6 Ra8e8 Be1xa5
24/57 01:24 414.351.043 4.905.000 +1.50 Re1e2 Bf8e7 b2b3 Rb6a6 Kc1b2 Kf7f8 Re2e1 Kf8f7 Re1e4 Kf7f8 Kb2a2 Kf8g7 Re4e1 Be7f8 Ka2b2 Bc8d7 Kb2a3 Bd7xb5 a4xb5 Ra6a8 Re1e6 Qd8c7 Qh2e2
25/57 01:42 498.195.607 4.882.000 +1.51 Re1e2 Bf8e7 b2b3 Rb6a6 Kc1b2 Kf7f8 Re2e3 Kf8f7 Re3e4 Kf7f8 Re4e2 Kf8g7 Re2e1 Kg7f7 Qh2e2 Be7f8 Bg3f2 Bc8d7 Bf2e3 Bf8e7 Be3d2 Bd7xb5 a4xb5 Ra6a8 Qe2e4 a5a4
26/57 03:01 890.478.543 4.925.000 +1.51 Re1e2 Bf8e7 Re2e3 Rb6a6 b2b3 Kf7f8 Kc1b2 Kf8f7 Kb2a3 Kf7g7 Re3e2 Kg7f8 Ka3b2 Kf8f7 Re2e3 Kf7f8 Kb2b1 Kf8f7 Kb1a2 Ra6b6 Qh2e2 Kf7f8 Qe2e1 Bc8d7 Qe1xa5 Kf8g7 Qa5e1
27/57 04:17 1.272.673.108 4.951.000 +1.46 Re1e2 Bf8e7 Re2e3 Rb6a6 b2b3 Be7f8 Kc1b2 Qd8b6 Re3e4 Bc8d7 Re4e1 Kf7g7 Re1e3 Bd7xb5 a4xb5 Ra6a8 Re3e6 a5a4 Bg3xd6 a4a3+ Kb2a2 Bf8xd6 Qh2xd6 Qb6xd6 Re6xd6
28/62 08:14 2.476.327,027 5.011.000 +1.46 Re1e2 Bf8e7 Re2e3 Rb6a6 b2b3 Be7f8 Kc1b2 Qd8b6 Qh2c2 Qb6d8 Re3e1 Bc8d7 Qc2h2 Qd8b6 Re1e2 Kf7g8 Kb2a2 Kg8f7 Re2e4 Kf7g7 Ka2b2 Kg7g8 Kb2c2 Kg8g7 Qh2e2 Qb6d8 Kc2c3 Bd7xb5 a4xb5
29/67 12:26 3.764.327,937 5.042.000 +1.46 Re1e2 Bf8e7 Re2e3 Rb6a6 b2b3 Be7f8 Kc1b2 Qd8b6 Qh2c2 Qb6d8 Re3e1 Bc8d7 Qc2h2 Qd8b6 Re1e2 Bd7c8 Re2e4 Bc8d7 Re4e1 Kf7g7 Re1e3 Kg7f7 Kb2a3 Kf7g7 Qh2e2 Bd7xb5 a4xb5 Ra6a8 Re3e6 a5a4
30/67 19:56 6.033.454,494 5.044.000 +1.46 Re1e2 Bf8e7 Re2e3 Rb6a6 b2b3 Be7f8 Kc1b2 Qd8b6 Qh2c2 Qb6d8 Re3e1 Bc8d7 Qc2h2 Qd8b6 Re1e2 Kf7g7 Re2e4 Kg7g8 Kb2a2 Qb6d8 Re4e2 Qd8b6 Re2e1 Kg8g7 Re1e3 Kg7f7 Qh2e2 Qb6d8 Ka2b2 Kf7g8 Nb5c3 Kg8g7 Qe2h2 Kg7g8 Kb2a2 Kg8f7 Re3e1 Qd8c7 Qh2b2 Qc7d8 Nc3b5 Bf8e7 Qb2h2 Qd8b8 Ka2b2 Be7f8 Kb2a3 Kf7g7 Re1e2 Kg7f7 Ka3b2 Kf7g7 Re2e3 Kg7g8 Kb2a2 Kg8f7 Re3e2 Kf7g7 Qh2f2 Bd7xb5 a4xb5
31/67 30:22 9.274.681,181 5.088.000 +1.46 Re1e2 Bf8e7 Re2e3 Rb6a6 b2b3 Be7f8 Kc1b2 Bf8e7 Kb2a3 Be7f8 Ka3a2 Kf7g7 Re3e2 Kg7f7 Re2e1 Qd8b6 Qh2c2 Qb6d8 Ka2b2 Bc8d7 Qc2h2 Bd7c8 Re1e4 Qd8b6 Qh2c2 Qb6d8 Re4e3 Bc8d7 Qc2e4 Qd8e7 Qe4d3 Qe7d8 Qd3e2 Bf8e7 Qe2h2 Bd7c8 Kb2a3
32/68 48:38 14.822.401,941 5.079.000 +1.46 Re1e2 Bf8e7 Re2e3 Rb6a6 b2b3 Be7f8 Kc1b2 Bf8e7 Kb2a3 Be7f8 Ka3a2 Kf7g7 Re3e2 Kg7f7 Re2e1 Bf8e7 Re1e3 Be7f8 Ka2a3 Qd8b6 Qh2e2 Qb6d8 Qe2f3 Bc8d7 Qf3e4 Qd8b6 Qe4h1 Bf8e7 Qh1e1 Qb6d8 Nb5c7 Ra6a7 Nc7e6 Bd7xe6 Re3xe6
33/70 1:21:43 24.916.031,911 5.081.000 +1.46 Re1e2 Bf8e7 Re2e3 Rb6a6 b2b3 Be7f8 Kc1b2 Bf8e7 Kb2a3 Be7f8 Ka3a2 Ra6b6 Nb5c3 Rb6a6 Ka2b2 Bc8d7 Nc3b5 Qd8b6 Kb2a3 Kf7g7 Qh2e2 Qb6d8 Re3e4 Kg7f7 Qe2h2 Qd8b6 Re4e1 Kf7g8 Qh2b2 Qb6d8 Qb2e2 Kg8g7 Nb5c3 Qd8b6 Ka3a2 Qb6d8 Qe2e3 Kg7f7 Nc3b5 Kf7g7 Qe3e4 Bd7xb5 a4xb5 Ra6a8 Ka2a3 Qd8c7 Qe4e6 Ra8d8


searching more than 40,000,000 knodes when the last 9,000,000 knodes is for the right move but still houdini does not choose Re6

cluster Rybka is clearly faster by a big margin(even if you use houdini on the fastest hardware that it can use and I guess houdini may need 1 hour instead of more than 2 hours on 8 cpu and 64 bits)

http://rybkaforum.net/cgi-bin/rybkaforu ... ?tid=19948
Uri Blass
Posts: 10890
Joined: Thu Mar 09, 2006 12:37 am
Location: Tel-Aviv Israel

Re: 10 beautiful testpositions!!

Post by Uri Blass »

finally houdini solve position number 2
Here is the next line of the analysis.

It may be interesting to see result of rybka4 4 cpu for comparison:
33/84+ 2:44:47 50.491.538,048 5.106.000 +1.57 Re1e6
User avatar
Laskos
Posts: 10948
Joined: Wed Jul 26, 2006 10:21 pm
Full name: Kai Laskos

Re: 10 beautiful testpositions!!

Post by Laskos »

Uri Blass wrote:finally houdini solve position number 2
Here is the next line of the analysis.

It may be interesting to see result of rybka4 4 cpu for comparison:
33/84+ 2:44:47 50.491.538,048 5.106.000 +1.57 Re1e6
Uri, do you think all 10 positions here are relevant for the engine strength? That sort of things happen in less than 1% of the games. Actually, "overtesting" on those things while developing an engine could harm her strength.

Kai
Robert Flesher
Posts: 1287
Joined: Tue Aug 18, 2009 3:06 am

Re: 10 beautiful testpositions!!

Post by Robert Flesher »

Laskos wrote:
Uri Blass wrote:finally houdini solve position number 2
Here is the next line of the analysis.

It may be interesting to see result of rybka4 4 cpu for comparison:
33/84+ 2:44:47 50.491.538,048 5.106.000 +1.57 Re1e6
Uri, do you think all 10 positions here are relevant for the engine strength? That sort of things happen in less than 1% of the games. Actually, "overtesting" on those things while developing an engine could harm her strength.

Kai

The problem I see with this postion is Re6 although a very good move, most likely the best, is not the only path to victory. White is winning after many different moves. But, black can defend for a long while, unless I am missing something. I will try to run Houdini on my machine and see what happens.