Test Position - Divide and conquer

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

Moderators: hgm, Rebel, chrisw

Albert Silver
Posts: 3019
Joined: Wed Mar 08, 2006 9:57 pm
Location: Rio de Janeiro, Brazil

Test Position - Divide and conquer

Post by Albert Silver »

This position came up in the high-level match hosted by Martin.

[D]1r2r1k1/1bp1qppn/1p1p3p/p7/P1PPp2n/BBP1P1NP/4QPP1/1R2R1K1 b - - 0 24

Black played 1...Nf3+ and went on to win.
"Tactics are the bricks and sticks that make up a game, but positional play is the architectural blueprint."
zullil
Posts: 6442
Joined: Tue Jan 09, 2007 12:31 am
Location: PA USA
Full name: Louis Zulli

Re: Test Position - Divide and conquer

Post by zullil »

Albert Silver wrote:This position came up in the high-level match hosted by Martin.

[D]1r2r1k1/1bp1qppn/1p1p3p/p7/P1PPp2n/BBP1P1NP/4QPP1/1R2R1K1 b - - 0 24

Black played 1...Nf3+ and went on to win.
Spark-0.4 (8 threads) switches to Nf3+ at depth 26 but doesn't see it as winning (in fact, score is 0), even at depth 30:

Code: Select all

info depth 26 seldepth 66 score cp 0 time 1400061 nodes 22082959491 nps 15772855 hashfull 1000 pv h4f3 g2f3 e4f3 e2f1 e7h4 d4d5 h7g5 g3f5 h4h5 f5g3 h5h4

Code: Select all

info depth 30 seldepth 68 score cp 0 time 3473323 nodes 53679093815 nps 15454679 hashfull 1000 pv h4f3 g2f3 e4f3 e2f1 e7h4 d4d5 h7g5 g3f5 h4h5 f5g3 h5h4
Albert Silver
Posts: 3019
Joined: Wed Mar 08, 2006 9:57 pm
Location: Rio de Janeiro, Brazil

Re: Test Position - Divide and conquer

Post by Albert Silver »

zullil wrote:
Albert Silver wrote:This position came up in the high-level match hosted by Martin.

[D]1r2r1k1/1bp1qppn/1p1p3p/p7/P1PPp2n/BBP1P1NP/4QPP1/1R2R1K1 b - - 0 24

Black played 1...Nf3+ and went on to win.
Spark-0.4 (8 threads) switches to Nf3+ at depth 26 but doesn't see it as winning (in fact, score is 0), even at depth 30:

Code: Select all

info depth 26 seldepth 66 score cp 0 time 1400061 nodes 22082959491 nps 15772855 hashfull 1000 pv h4f3 g2f3 e4f3 e2f1 e7h4 d4d5 h7g5 g3f5 h4h5 f5g3 h5h4

Code: Select all

info depth 30 seldepth 68 score cp 0 time 3473323 nodes 53679093815 nps 15454679 hashfull 1000 pv h4f3 g2f3 e4f3 e2f1 e7h4 d4d5 h7g5 g3f5 h4h5 f5g3 h5h4
I wouldn't worry. Rybka didn't see it as a win either, just better, but after playing over the game, it is clear that this longterm attacking move leaves White in a hopeless position.
"Tactics are the bricks and sticks that make up a game, but positional play is the architectural blueprint."
User avatar
Eelco de Groot
Posts: 4565
Joined: Sun Mar 12, 2006 2:40 am
Full name:   

Re: Test Position - Divide and conquer

Post by Eelco de Groot »

Rainbow Serpent doesn't detect any long term advantages with Nf3+, this move is also not very high in the move list but all that does not say very much. There is a preference for 1... g6 which is maybe a more cautious move? A pity it is only found after 18 minutes and a very long think at iteration 17, maybe it is related to recent changes in build 59 in the PV search, don't know, I would have to compare with other builds, this is the first test for build 059.

Some analysis with Rybka 4 would be interesting, for example what is the score for Nf3+, and what would be the PV?
Maybe the plexiglas computer from KG could tell us more?

Regards, Eelco

[D]1r2r1k1/1bp1qppn/1p1p3p/p7/P1PPp2n/BBP1P1NP/4QPP1/1R2R1K1 b - -

Engine: Rainbow Serpent 1.7.1s(dc) Build 059 (Athlon 2009 MHz, 256 MB)
by Tord Romstad, Marco Costalba, Joona Kiiski

1.00 0:01 -0.56 1...Ng5 (449) 0

2.01 0:01 -0.36 1...Ng5 2.Qg4 (795) 0

3.01 0:01 -0.08 1...Ng5 2.Qg4 Qf6 3.Bxd6 (4.324) 2

4.01 0:01 +0.08 1...Ng5 2.c5 bxc5 3.Qh5 (18.851) 9

5.01 0:01 -0.61 1...Ng5 2.c5 bxc5 3.Qh5 Nhf3+ 4.gxf3 Nxf3+
5.Kg2 cxd4 6.cxd4 Nxe1+ 7.Rxe1 (28.620) 14

5.09 0:02 -0.48 1...Bc6 2.Qh5 Qg5 3.Bc2 Qxh5 4.Nxh5 Re6
5.Nf4 (149.536) 64

5.17 0:02 -0.08 1...Nf6 2.Qd2 Qd7 3.Kf1 Nf5 4.Nxf5 Qxf5 (156.699) 67

6.01 0:02 +0.16 1...Nf6 2.c5 dxc5 3.dxc5 Bd5 4.c4 Bc6 (167.202) 70

7.01 0:02 -0.08 1...Nf6 2.c5 dxc5 3.Qb5 Bd5 4.c4 c6 (197.591) 79

8.01 0:03 0.00 1...Nf6 2.c5 dxc5 3.Qa2 Qd7 4.dxc5 Bd5
5.cxb6 Bxb3 6.Qxb3 Rxb6 (613.833) 165

9.01 0:04 0.00 1...Nf6 2.c5 dxc5 3.Qa2 Qd7 4.dxc5 Bd5
5.cxb6 Bxb3 6.Qxb3 Rxb6 (744.985) 181

10.01 0:10 +0.08++ 1...Nf6 2.Qd1 Qd7 3.Bb2 Bc8 (3.119.669) 291

10.01 0:11 +0.16++ 1...Nf6 2.c5 dxc5 3.Qa2 Red8 4.dxc5 bxc5
5.Bc4 Bc6 (3.318.559) 294

10.01 0:12 +0.12 1...Nf6 2.Qd1 Bc8 3.Qe2 Ba6 4.Qf1 Qe6
5.Bb2 Qd7 6.Rbd1 d5 (3.582.092) 296

11.01 0:12 +0.12 1...Nf6 2.Qd1 Bc8 3.Qe2 Ba6 4.Qf1 Qe6
5.Bb2 Qd7 6.Rbd1 d5 (3.803.729) 299

12.01 0:18 +0.16 1...Nf6 2.c5 dxc5 3.Qa2 Red8 4.dxc5 bxc5
5.Ne2 Ba6 6.Nf4 g5 7.Nd5 Nxd5 8.Bxd5 Rxb1
9.Rxb1 (5.891.860) 319

13.01 0:20 +0.08-- 1...Nf6 2.c5 dxc5 3.dxc5 Bd5 4.Bxd5 Nxd5
5.Qc4 c6 6.Rec1 (6.754.473) 324

13.01 0:27 0.00-- 1...Nf6 2.c5 dxc5 3.dxc5 Bd5 4.Qc2 Bxb3
5.Qxb3 Qe5 6.Rbd1 Re6 7.Rd4 (9.050.058) 334

13.01 0:36 +0.24++ 1...Nf6 2.c5 dxc5 3.dxc5 Bd5 4.Qc2 Bxb3
5.Qxb3 Qe5 6.Rbd1 bxc5 7.Qc4 Qg5 (12.305.021) 341

13.01 0:38 +0.08-- 1...Nf6 2.c5 dxc5 3.dxc5 Bd5 4.Qc2 Bxb3
5.Qxb3 Qe5 6.Rbd1 bxc5 7.Qc4 Qg5
8.Rb1 Rxb1 9.Rxb1 (13.296.530) 343

13.01 0:41 +0.24 1...Nf6 2.c5 dxc5 3.dxc5 Bd5 4.Bxd5 Nxd5
5.Qc4 Qe5 6.Rbd1 c6 7.Rd4 b5 8.axb5 Rxb5 (14.166.021) 344

14.01 0:42 +0.04-- 1...Nf6 2.c5 dxc5 3.dxc5 Bd5 4.Bxd5 Nxd5
5.Qc4 Qe5 6.Rbd1 c6 7.Rd4 b5 8.axb5 Rxb5
9.Rxe4 (14.630.052) 345

14.01 0:52 +0.24 1...Nf6 2.c5 dxc5 3.dxc5 bxc5 4.Red1 Bc6
5.Qa2 Rb6 6.Rd4 Nd7 7.Rdd1 Ne5 8.Kh1 Nd3
9.Nh5 Nf5 (18.412.088) 348

15.01 0:54 +0.16-- 1...Nf6 2.c5 bxc5 3.Qa2 Qd7 4.dxc5 d5 (19.108.765) 349

15.01 1:02 +0.08-- 1...Nf6 2.c5 dxc5 3.dxc5 bxc5 4.Rb2 Bd5
5.Bxd5 Nxd5 6.Rxb8 Rxb8 7.Qa6 Nxc3
8.Qxa5 (21.703.408) 348

15.01 1:08 0.00 1...Nf6 2.c5 dxc5 3.dxc5 bxc5 4.Rb2 Bc6
5.Reb1 Rb6 6.Qc4 Bd5 7.Qxc5 Qxc5
8.Bxc5 Rxb3 9.Rxb3 Bxb3 10.Rxb3 Nd7
11.Ba3 Nb6 12.Bd6 Nd5 (24.000.773) 350

16.01 1:34 +0.08 1...Nf6 2.c5 dxc5 3.dxc5 bxc5 4.Rb2 Bd5
5.Bxd5 Nxd5 6.Rxb8 Rxb8 7.Qa6 Qe5
8.Qxa5 c4 9.Qa7 Rd8 10.Qd4 Qxd4
11.cxd4 Ra8 12.Nxe4 Rxa4 13.Bc5 (33.324.990) 353

17.01 2:14 -0.08 1...Nf6 2.c5 dxc5 3.dxc5 bxc5 4.Bc2 Qe6
5.Qb5 Qa2 6.Qb2 Qxb2 7.Rxb2 Bc6
8.Rxb8 Rxb8 9.Bxc5 Rb2 10.Rc1 Kh7
11.Be7 Kg6 12.Bd8 Nf5 13.Nxf5 Kxf5
14.Bxc7 Nd5 (48.445.631) 359

17.03 4:33 0.00 1...Ng5 2.Qh5 Bc8 3.Bd1 Qf6 4.Bg4 Ba6
5.Be2 Bc8 6.Bg4 (100.160.767) 366


17.05 18:10 +0.48++ 1...g6 2.c5 dxc5 3.dxc5 Bc8 4.Qa2 Qf6
5.cxb6 Ng5 6.Bc5 Ngf3+ 7.gxf3 Qxf3
8.Bxf7+ Kh7 (312.332.037) 286

17.05 20:27 +0.52 1...g6 2.c5 dxc5 3.dxc5 Bc8 4.Qb5 Bxh3
5.gxh3 bxc5 6.Qc4 Qd7 7.Qxf7+ Qxf7
8.Bxf7+ Kxf7 9.Red1 Ng5 10.Kf1 Nhf3
11.Rxb8 Rxb8 12.h4 Nxh4 13.Rd7+ Ke6
14.Rxc7 Rb1+ (357.542.296) 291

18.01 21:34 +0.44 1...g6 2.c5 dxc5 3.dxc5 Bc8 4.Qb5 Bxh3
5.gxh3 bxc5 6.Bxc5 Rxb5 7.Bxe7 Nf3+
8.Kf1 Rb6 9.Bc5 Rxb3 10.Rxb3 Nd2+
11.Ke2 Nxb3 12.Ba7 c5 13.c4 Re7
14.Rb1 Rxa7 (383.126.048) 296

19.01 27:47 +0.60 1...g6 2.c5 dxc5 3.dxc5 Bc8 4.Qb5 Bxh3
5.gxh3 bxc5 6.Bxc5 Rxb5 7.Bxe7 Nf3+
8.Kf1 Rb6 9.Bc5 Rxb3 10.Rxb3 Nd2+
11.Ke2 Nxb3 12.Ba7 c5 13.c4 Re7
14.Rb1 Rxa7 (519.810.607) 311

20.01 35:49 +0.20-- 1...g6 2.Qf1 Bc8 3.Red1 (698.599.407) 325

20.01 71:26 +0.48 1...g6 2.Qf1 Bc8 3.Red1 Ba6 4.Ne2 Qg5
5.Kh1 Nf6 6.Ba2 Kg7 7.Bb3 Ra8 8.Nf4 Kg8
9.Qg1 Rab8 10.g4 Nf3 11.Qg3 Re7 (1.486.386.432) 346

21.01 80:18 +0.48 1...g6 2.Qf1 Bc8 3.Red1 Ba6 4.Kh1 Qf6
5.Ne2 Qg5 6.Nf4 Nf6 7.Qg1 Nh5 8.Nxh5 Qxh5
9.Qh2 Qg5 10.Re1 Kg7 11.Qg3 Qxg3
12.fxg3 Nf5 13.g4 Nh4 14.Rf1 f5 (1.686.124.971) 349

22.01 110:38 +0.52 1...g6 2.Qf1 Bc8 3.Red1 Ba6 4.Kh1 Qf6
5.Ne2 Qg5 6.Nf4 Nf6 7.Qg1 Nh5 8.c5 dxc5
9.Nxh5 gxh5 10.Bc2 Qg7 11.Qh2 cxd4
12.Rxd4 f5 13.Rbd1 Kh7 14.Rd7 (2.278.676.289) 343


best move: g7-g6 time: 114:32.891 min n/s: 345.144 nodes: 2.372.130.178
Debugging is twice as hard as writing the code in the first
place. Therefore, if you write the code as cleverly as possible, you
are, by definition, not smart enough to debug it.
-- Brian W. Kernighan
Sean Evans
Posts: 1777
Joined: Thu Jun 05, 2008 10:58 pm
Location: Canada

Re: Test Position - Divide and conquer

Post by Sean Evans »

Hi, looking at that position with my human eye, I do not see a winning position for Black. The horizon much be quite a few moves to see a potential advantage, ever so slight that it might be.

Cordially,

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

Re: Test Position - Divide and conquer

Post by kgburcham »

another fun test position---not clear early.
I can play this out later but I think Albert already has.

[D] 1r2r1k1/1bp1qppn/1p1p3p/p7/P1PPp2n/BBP1P1NP/4QPP1/1R2R1K1 b - -

Deep Rybka 4 x64

20.01 1:26 -0.08 1...Ng5 2.Qh5 Bc8 3.Bd1 Kh7 4.Kh1 Nxg2 5.Kxg2 g6 6.Qh4 Nf3 (53.301.525) 617
20.02 2:39 -0.09 1...Nf6 2.Red1 Ba6 3.Qa2 Rbd8 4.Qd2 Qd7 5.Qa2 Qe6 6.c5 d5 (94.329.123) 591
20.21 3:45 -0.22++ 1...Nf3+ 2.gxf3 exf3 3.Qf1 Qh4 4.Red1 Bc8 5.d5 Bxh3 6.Rd4 Bg4 (141.088.094) 626
20.01 6:46 -0.37++ 1...Nf3+ 2.gxf3 exf3 3.Qf1 Qh4 4.Red1 Bc8 5.d5 Bxh3 6.Rd4 Bg4 (244.966.210) 602
20.01 11:03 -0.77++ 1...Nf3+ 2.gxf3 exf3 3.Qf1 Qh4 4.Red1 Bc8 5.d5 Re4 6.Bc2 Bxh3 (413.414.602) 623
20.01 11:45 -0.71 1...Nf3+ 2.gxf3 exf3 3.Qf1 Qh4 4.Red1 Bc8 5.d5 Re4 6.Bc2 Bxh3 (440.886.869) 624
21.01 17:25 -0.83++ 1...Nf3+ 2.gxf3 exf3 3.Qf1 Qh4 4.Red1 Bc8 5.d5 Re4 6.Bc2 Bxh3 (682.574.947) 652
21.01 22:28 -0.88 1...Nf3+ 2.gxf3 exf3 3.Qf1 Qh4 4.Red1 Bc8 5.d5 Re4 6.Bc2 Bxh3 (902.982.197) 669
22.01 30:31 -0.90 1...Nf3+ 2.gxf3 exf3 3.Qf1 Qh4 4.d5 Bc8 5.Bc2 g6 6.Red1 Bxh3 (1.255.232.087) 685
23.01 33:39 -0.90 1...Nf3+ 2.gxf3 exf3 3.Qf1 Qh4 4.d5 Bc8 5.Bc2 g6 6.Rbd1 Bxh3 (1.384.750.412) 685
23.01 53:13 -0.90 1...Nf3+ 2.gxf3 exf3 3.Qf1 Qh4 4.d5 Bc8 5.Bc2 g6 6.Rbd1 Bxh3 7.Rd4 Bg4 8.Rxg4 Qxg4 9.Rd1 h5 10.Rd4 Qg5 11.Kh2 Nf6 12.Bc1 Ng4+ 13.Kh3 Qe7 14.Rxg4 hxg4+ (1.501.108.633) 470
User avatar
Eelco de Groot
Posts: 4565
Joined: Sun Mar 12, 2006 2:40 am
Full name:   

Re: Test Position - Divide and conquer

Post by Eelco de Groot »

Thanks for the analysis KG, it seems your Xeon already anticipated and started analyzing, before I even asked! That reminds me of Orac in the TV series "Blake's Seven" also in a transparent case. In one episode it seemingly anticipated far in advance the destruction of Blake's spaceship but it turned out this was all part of its own plan.

Image

Orac quote:"I am shutting down. I have much to do. You have engaged my circuits on your petty affairs for far too long."
("Redemption")
User avatar
Eelco de Groot
Posts: 4565
Joined: Sun Mar 12, 2006 2:40 am
Full name:   

Re: Test Position - Divide and conquer

Post by Eelco de Groot »

Rainbow Serpent still saw nothing after 1... Nf3+

[D]1r2r1k1/1bp1qppn/1p1p3p/p7/P1PPp3/BBP1PnNP/4QPP1/1R2R1K1 w - -

Engine: Rainbow Serpent 1.7.1s(dc) Build 059 (Athlon 2009 MHz, 256 MB)
by Tord Romstad, Marco Costalba, Joona Kiiski

1.00 0:00 +4.16 2.gxf3 (45) 0

2.00 0:00 +4.00 2.gxf3 exf3 3.Qd2 (141) 0

3.00 0:01 +3.30 2.gxf3 Ng5 3.Nf5 Nxf3+ 4.Qxf3 exf3
5.Nxe7+ Rxe7 (1.073) 0

4.01 0:02 +3.31 2.gxf3 Ng5 3.Nf5 Nxf3+ 4.Qxf3 exf3
5.Nxe7+ Rxe7 (1.807) 0

5.01 0:06 +2.86 2.gxf3 Ng5 3.Nxe4 Nxh3+ 4.Kh1 Bxe4
5.fxe4 Qxe4+ 6.f3 (9.285) 1

6.01 0:07 +1.93-- 2.gxf3 Ng5 3.Nxe4 Nxh3+ 4.Kh1 Qh4
5.d5 Nf4+ (12.467) 1

6.01 0:10 +2.42 2.gxf3 Ng5 3.f4 Nf3+ 4.Kg2 Nxe1+
5.Rxe1 Qd7 6.Bc2 (18.692) 1

7.01 0:27 +1.85 2.gxf3 Ng5 3.f4 Nf3+ 4.Kg2 Nh4+
5.Kh1 Nf3 6.f5 Qh4 7.Qf1 Nxe1 8.Rxe1 (56.366) 2

8.01 0:33 +1.85 2.gxf3 Ng5 3.f4 Nf3+ 4.Kg2 Nh4+
5.Kh1 Nf3 6.f5 Qh4 7.Qf1 Nxe1 8.Rxe1 (76.661) 2

9.01 0:52 +2.18 2.gxf3 Ng5 3.f4 Nf3+ 4.Kg2 Nh4+
5.Kh1 Nf3 6.Bc2 Qh4 7.Kg2 Bc8 8.Rh1 Bd7
9.Bxe4 (145.869) 2

10.01 1:06 +2.38 2.gxf3 Ng5 3.f4 Nf3+ 4.Kg2 Nh4+
5.Kf1 Nf3 6.Bc2 Qh4 7.Kg2 Bc8 8.Rh1 Bd7
9.f5 (321.290) 4

11.01 1:10 +2.18 2.gxf3 Ng5 3.f4 Nf3+ 4.Kg2 Nh4+
5.Kh1 Nf3 6.Bc2 Bc8 7.Bxe4 Nxe1
8.Rxe1 Qh4 9.Qc2 Bxh3 10.Kg1 (700.117) 9

12.01 1:12 +2.38 2.gxf3 Ng5 3.f4 Nf3+ 4.Kg2 Nh4+
5.Kh1 Nf3 6.Bc2 Bc8 7.Bxe4 Nxe1
8.Rxe1 Qh4 9.Qc2 Bxh3 10.Kg1 Bc8
11.Bc6 Bh3 12.Bd5 (1.127.100) 15

13.01 1:17 +2.29 2.gxf3 Ng5 3.fxe4 Bxe4 4.Nxe4 Qxe4
5.Qg4 Nf3+ 6.Kh1 Qxg4 7.hxg4 Nxe1
8.Rxe1 g6 9.Kg2 Kg7 10.Kf3 Rh8
11.Bc2 h5 12.g5 Rbe8 13.Rh1 f6
14.Kf4 fxg5+ 15.Kxg5 (3.168.992) 40

14.01 1:22 +2.34 2.gxf3 Ng5 3.fxe4 Bxe4 4.Nxe4 Qxe4
5.Qg4 Nf3+ 6.Kh1 Nxe1+ 7.Qxe4 Rxe4
8.Rxe1 Kf8 9.Kg2 g6 10.Bc2 Re6
11.Kf1 Ree8 12.Ke2 Ke7 13.Kd2 Kd7
14.Rg1 Kc6 (5.288.236) 64

15.01 1:31 +2.14-- 2.gxf3 exf3 3.Qf1 Qh4 4.c5 dxc5
5.Bc2 Ng5 6.Bf5 Bc6 7.Bg4 g6 8.dxc5 bxc5 (9.012.216) 98

15.01 1:34 +1.93-- 2.gxf3 exf3 3.Qf1 Qh4 4.c5 dxc5
5.Bc2 Ng5 6.Bf5 Bc6 7.Bg4 g6 8.dxc5 bxc5
9.Rxb8 Rxb8 10.Qa6 Nxh3+ 11.Bxh3 Qxh3
12.Qxc6 Qg2+ (9.868.608) 104

15.01 1:38 +1.53-- 2.gxf3 exf3 3.Qf1 Qh4 4.c5 dxc5
5.Bc2 Ng5 6.Bf5 Bc6 7.Bg4 g6 8.dxc5 bxc5
9.Rxb8 Rxb8 10.Qa6 Nxh3+ 11.Bxh3 Qxh3
12.Qxc6 Qg2+ (11.447.406) 116

15.01 1:48 +0.72-- 2.gxf3 exf3 3.Qf1 Qh4 4.c5 dxc5
5.Bc2 Ng5 6.Bf5 Bc6 7.Bg4 g6 8.dxc5 bxc5
9.Rxb8 Rxb8 10.Qa6 Nxh3+ 11.Bxh3 Qxh3
12.Qxc6 Qg2+ (15.011.672) 137

15.01 2:10 +0.36 2.gxf3 exf3 3.Qf1 Qh4 4.Bc2 g6 5.Rb5 Ng5
6.Rxg5 hxg5 7.Bd1 Kg7 8.c5 bxc5
9.dxc5 Bc8 10.Qd3 Bxh3 11.Bxf3 (21.962.204) 168

16.01 2:22 +0.44 2.gxf3 exf3 3.Qf1 Qh4 4.Bc2 g6 5.Rb5 Ng5
6.Rxg5 hxg5 7.Bd1 Kh8 8.c5 dxc5
9.dxc5 Bc8 10.cxb6 Rxb6 11.Bxf3 Bxh3
12.Qd3 (25.959.321) 182

17.01 2:51 0.00 2.gxf3 exf3 3.Qf1 Qh4 4.d5 Ng5 5.Nf5 Qe4
6.Ng3 Qh4 (36.448.583) 212

18.01 3:19 0.00 2.gxf3 exf3 3.Qf1 Qh4 4.d5 Ng5 5.Nf5 Qe4
6.Ng3 Qh4 (45.849.473) 229

19.01 4:20 0.00 2.gxf3 exf3 3.Qf1 Qh4 4.d5 Ng5 5.Nf5 Qe4
6.Ng3 Qh4 (65.705.139) 252

20.01 6:09 0.00 2.gxf3 exf3 3.Qf1 Qh4 4.d5 Ng5 5.Nf5 Qe4
6.Ng3 Qh4 (100.920.690) 273

21.01 8:47 0.00 2.gxf3 exf3 3.Qf1 Qh4 4.d5 Ng5 5.Nf5 Qe4
6.Ng3 Qh4 (156.400.554) 296

22.01 10:05 0.00 2.gxf3 exf3 3.Qf1 Qh4 4.d5 Ng5 5.Nf5 Qe4
6.Ng3 Qh4 (184.805.529) 305

23.01 16:53 0.00 2.gxf3 exf3 3.Qf1 Qh4 4.d5 Ng5 5.Nf5 Qe4
6.Ng3 Qh4 (328.412.629) 323

24.01 48:13 0.00 2.gxf3 exf3 3.Qf1 Qh4 4.d5 Ng5 5.Nf5 Qe4
6.Ng3 Qh4 (1.003.161.716) 346

25.01 137:27 0.00 2.gxf3 exf3 3.Qf1 Qh4 4.d5 Ng5 5.Nf5 Qe4
6.Ng3 Qh4 (2.826.134.473) 342

best move: g2xf3 time: 156:31.125 min n/s: 347.276 nodes: 3.261.210.282
Debugging is twice as hard as writing the code in the first
place. Therefore, if you write the code as cleverly as possible, you
are, by definition, not smart enough to debug it.
-- Brian W. Kernighan
User avatar
Eelco de Groot
Posts: 4565
Joined: Sun Mar 12, 2006 2:40 am
Full name:   

Re: Test Position - Divide and conquer

Post by Eelco de Groot »

But after taking the PV to the position after 4.d5 finally Serpent sees the light

[FEN "1r2r1k1/1bp1qppn/1p1p3p/p7/P1PPp2n/BBP1P1NP/4QPP1/1R2R1K1 b - -"]

1... Nf3+ 2. gxf3 exf3 3. Qf1 Qh4 4. d5 *


[D]1r2r1k1/1bp2ppn/1p1p3p/p2P4/P1P4q/BBP1PpNP/5P2/1R2RQK1 b - -

Engine: Rainbow Serpent 1.7.1s(dc) Build 059 (Athlon 2009 MHz, 256 MB)
by Tord Romstad, Marco Costalba, Joona Kiiski

1.00 0:00 -3.39 4...Ng5 (1.093)

2.00 0:00 -1.17 4...Ng5 5.Kh1 Nxh3 6.Nf5 Nxf2+ 7.Kg1 Qg4+
8.Kxf2 Qxf5 (1.584)

3.00 0:00 -1.17 4...Ng5 5.Kh1 Nxh3 6.Nf5 Nxf2+ 7.Kg1 Qg4+
8.Kxf2 Qxf5 (5.021) 313


3.00 0:00 +0.04 4...Bc8 5.Bc2 Ng5 6.Bf5 Bxf5 7.Nxf5 Nxh3+
8.Kh2 (11.163) 360

4.00 0:00 +0.04 4...Bc8 5.Bc2 Ng5 6.Bf5 Bxf5 7.Nxf5 Nxh3+
8.Kh2 (12.633) 407

5.00 0:00 +0.04 4...Bc8 5.Bc2 Ng5 6.Bf5 Bxf5 7.Nxf5 Nxh3+
8.Kh2 (15.373) 327

6.00 0:00 +0.04 4...Bc8 5.Bc2 Ng5 6.Bf5 Bxf5 7.Nxf5 Nxh3+
8.Kh2 (18.649) 296

7.00 0:00 -0.04-- 4...Bc8 5.Bc2 Ng5 6.Bf5 Bxf5 7.Nxf5 Nxh3+
8.Kh2 (22.064) 282


7.00 0:00 0.00 4...Ng5 5.Nf5 Qe4 6.Ng3 Qh4 (29.955) 318

8.00 0:00 0.00 4...Ng5 5.Nf5 Qe4 6.Ng3 Qh4 (32.821) 349

9.00 0:00 0.00 4...Ng5 5.Nf5 Qe4 6.Ng3 Qh4 (39.285) 360

10.00 0:00 0.00 4...Ng5 5.Nf5 Qe4 6.Ng3 Qh4 (57.390) 367

11.00 0:00 0.00 4...Ng5 5.Nf5 Qe4 6.Ng3 Qh4 (126.597) 385

12.00 0:00 0.00 4...Ng5 5.Nf5 Qe4 6.Ng3 Qh4 (190.179) 380

13.01 0:01 0.00 4...Ng5 5.Nf5 Qe4 6.Ng3 Qh4 (430.420) 372

14.01 0:01 0.00 4...Ng5 5.Nf5 Qe4 6.Ng3 Qh4 (656.373) 378

15.01 0:03 0.00 4...Ng5 5.Nf5 Qe4 6.Ng3 Qh4 (1.203.069) 379

16.01 0:04 0.00 4...Ng5 5.Nf5 Qe4 6.Ng3 Qh4 (1.682.337) 377

17.01 1:00 0.00 4...Ng5 5.Nf5 Qe4 6.Ng3 Qh4 (20.808.487) 346

18.01 1:07 0.00 4...Ng5 5.Nf5 Qe4 6.Ng3 Qh4 (23.418.308) 349

19.01 1:43 0.00 4...Ng5 5.Nf5 Qe4 6.Ng3 Qh4 (36.108.344) 348

20.01 3:27 0.00 4...Ng5 5.Nf5 Qe4 6.Ng3 Qh4 (74.051.183) 356


20.03 24:54 +0.08++ 4...Bc8 5.Red1 Ng5 6.Rd4 Nxh3+ 7.Kh2 Qf6
8.Bc2 Re5 9.c5 dxc5 10.Bxc5 Bd7
11.Kh1 Qg5 12.Be4 Bg4 13.Bd3 Qh4
14.Rxg4 Qxg4 (545.288.509) 364

20.03 30:11 +0.16++ 4...Bc8 5.Red1 Ng5 6.Rd4 Nxh3+ 7.Kh2 Qf6
8.Bc2 Re5 9.c5 dxc5 10.Bxc5 Bd7
11.Ba3 h5 12.Qd3 g6 13.Qf1 Bg4
14.Rc4 Rxd5 15.Rxc7 Qh4 (655.392.537) 361

20.03 30:34 +0.32++ 4...Bc8 5.Red1 Ng5 6.Rd4 Nxh3+ 7.Kh2 Qf6
8.Bc2 Re5 9.c5 dxc5 10.Bxc5 Bd7
11.Ba3 h5 12.Qd3 g6 13.Qf1 Bg4
14.Rc4 Rxd5 15.Rxc7 Qh4 (663.392.074) 361

20.03 38:06 +0.64++ 4...Bc8 5.Red1 Ng5 6.Rd4 Nxh3+ 7.Kh2 Qf6
8.Bc2 Re5 9.c5 dxc5 10.Bxc5 h5
11.Qd3 g6 12.d6 Rxc5 13.dxc7 Rxc7
14.Rd8+ Kh7 15.Nxh5 Qh4 (817.591.228) 357

20.03 50:20 +0.68 4...Bc8 5.Bc2 g6 6.Rbd1 Bxh3 7.Rd4 Bg4
8.Bc1 Ng5 9.Rxg4 Qxg4 10.Rd1 f5
11.Rd4 Ne4 12.Kh2 Qh4+ 13.Kg1 h5
14.Bxe4 fxe4 15.Bd2 Kf7 16.Qb1 Qh3
17.Qf1 Qxf1+ (1.052.728.334) 348

21.01 54:33 +1.20 4...Bc8 5.Bc2 g6 6.Rbd1 Bxh3 7.Rd4 Bg4
8.Bc1 Ng5 9.Rxg4 Qxg4 10.Rd1 f5
11.Rd4 Ne4 12.Kh2 Qh4+ 13.Kg1 h5
14.Ba3 Qg5 15.Bxe4 Rxe4 16.Rxe4 fxe4
17.Qh3 Qe7 (1.139.642.563) 348

22.01 72:19 +1.81 4...Bc8 5.Bc2 g6 6.Rbd1 Bxh3 7.Rd4 Bg4
8.Rxg4 Qxg4 9.Rd1 h5 10.Rd4 Qg5
11.Rf4 h4 12.Rxf3 hxg3 13.Rxg3 Qh4
14.Qb1 Nf8 15.Qb5 Re5 16.Qc6 Rh5
17.Kf1 Qh1+ (1.542.097.136) 355
Debugging is twice as hard as writing the code in the first
place. Therefore, if you write the code as cleverly as possible, you
are, by definition, not smart enough to debug it.
-- Brian W. Kernighan
User avatar
Eelco de Groot
Posts: 4565
Joined: Sun Mar 12, 2006 2:40 am
Full name:   

Re: Test Position - Divide and conquer

Post by Eelco de Groot »

4 more lines of analysis:

23.01 106:25 +1.65 4...Bc8 5.Bc2 g6 6.Rbd1 Bxh3 7.Rd4 Bg4
8.Rxg4 Qxg4 9.Rd1 h5 10.Rd4 Qg5
11.Rf4 h4 12.Rxf3 hxg3 13.Rxg3 Qh4
14.Rh3 Qf6 15.f4 Qg7 16.Bb2 Nf6
17.Qg2 Kf8 (2.349.687.635) 367

24.01 142:41 +1.73 4...Bc8 5.Bc2 g6 6.Rbd1 Bxh3 7.Rd4 Bg4
8.Rxg4 Qxg4 9.Rd1 h5 10.Rd4 Qg5
11.Rf4 h4 12.Rxf3 hxg3 13.Rxg3 Qf6
14.f4 Qh4 15.Qg2 Nf6 16.Kf1 Nh5
17.Rh3 Nxf4 (3.243.942.606) 378

25.01 229:57 +1.89 4...Bc8 5.Bc2 g6 6.Rbd1 Bxh3 7.Rd4 Bg4
8.Rxg4 Qxg4 9.Rd1 h5 10.Rd4 Qg5
11.Rf4 h4 12.Rxf3 hxg3 13.Rxg3 Qf6
14.f4 Qh4 15.Qg2 Nf6 16.Kf1 Nh5
17.Rh3 Nxf4 (5.393.825.302) 390

26.01 461:12 +1.77 4...Bc8 5.Bc2 g6 6.Rbd1 Bxh3 7.Rd4 Bg4
8.Rxg4 Qxg4 9.Rd1 h5 10.Rd4 Qg5
11.Rf4 h4 12.Rxf3 hxg3 13.Rxg3 Qf6
14.f4 Qh4 15.Qg2 Nf6 16.Kf1 Nh5
17.Rh3 Nxf4 (11.145.670.129) 402
Debugging is twice as hard as writing the code in the first
place. Therefore, if you write the code as cleverly as possible, you
are, by definition, not smart enough to debug it.
-- Brian W. Kernighan