[d]1rr3k1/1p1bnpbp/p2pp1p1/2p5/Pq1nPPP1/2NPB1NP/1PPQ2BK/1R3R2 b - - 4 16
What do your favorite engines choose here? Is f5 necessary?
Test Position
Moderator: Ras
-
zullil
- Posts: 6442
- Joined: Tue Jan 09, 2007 12:31 am
- Location: PA USA
- Full name: Louis Zulli
-
zamar
- Posts: 613
- Joined: Sun Jan 18, 2009 7:03 am
Re: Test Position
Here is what current stockfish devel version thinks about the position:
1 thread, core i7:
2 +1.09 00:00 2846 Bxa4 Nxa4 Qxa4
3 +0.93 00:00 4589 Bxa4 Nxa4 Qxa4 b3
4 +1.05 00:00 6444 Bxa4 Nxa4 Qxa4 c3 Nb3
5 +0.93 00:00 8005 Bxa4 Nxa4 Qxa4 c3 Nb3 Qc2
6 +0.93 00:00 10403 Bxa4 Nxa4 Qxa4 c3 Nb3 Qc2 Nc6
7 < +0.85 00:00 14151 Bxa4 Nxa4 Qxa4 c3 Nb3 Qc2 Nc6 f5
7 +0.93 00:00 30905 Bxa4 Nxa4 Qxa4 c3 Nb3 Qc2 Nc6 g5
8 < +0.85 00:00 32157 Bxa4 Nxa4 Qxa4 c3 Nb3 Qc2 Nc6 g5
8 < +0.77 00:00 37302 Bxa4 Nxa4 Qxa4 c3 Nb3 Qc2 Nc6 g5
8 +0.81 00:00 83528 Rd8 Rfe1 Bxa4 Nxa4 Qxa4 Bxd4 cxd4 Ra1 Qb5
9 +0.73 00:00 132030 Rd8 Qf2 f5 exf5 gxf5 Rfe1 Bxa4 Nxa4 Qxa4 b3
10 < +0.57 00:00 271304 Rd8 Qf2 f5 exf5 gxf5 Rfe1 Bxa4 Nxa4 Qxa4 c3
10 +0.81 00:00 422871 Re8 g5 d5 Qc1 b6 Ra1 Bc6 Bd2 dxe4 dxe4
11 +0.85 00:01 605584 Re8 g5 Rf8 Nge2 f5 b3 fxe4 dxe4 Nec6 Rbd1 e5
12 < +0.77 00:01 718523 Re8 g5 b6 Nge2 d5 b3 dxe4 dxe4
12 < +0.69 00:01 950968 Re8 f5 gxf5 gxf5 exf5 Nh5 Be5+ Bf4
12 < +0.53 00:01 1171k Re8 f5 gxf5 gxf5 exf5 Nh5 Be5+ Bf4
12 +0.77 00:03 1841k Re8 g5 Rf8 Nge2 f5 b3 fxe4 dxe4 Nec6 Rbd1 Rbc8
Qd3 Rf7 Nxd4 cxd4
13 < +0.65 00:03 2189k Re8 f5 exf5 gxf5 Be5 fxg6 hxg6 Bf4 Bxa4 Nxa4 Qxa4
c3 Bxf4 Qxf4
13 < +0.53 00:04 2667k Re8 f5 exf5 gxf5 Be5 fxg6 hxg6 Bf4 Bxa4 Nxa4 Qxa4
c3 Bxf4 Qxf4
13 < +0.28 00:05 3369k Re8 f5 exf5 gxf5 Be5 fxg6 hxg6 Bf4 Bxa4 Nxa4 Qxa4
c3 Bxf4 Qxf4
13 +0.53 00:05 3568k Rd8 Qf2 f5 exf5 gxf5 Rfe1 Bf6 Nh5 Bh8 Qh4 Ng6 Qf2
Bxa4 Nxa4 Qxa4
14 < +0.24 00:07 4586k Rd8 Qf2 Bxa4 Nxa4 Qxa4 f5 exf5 exf5 gxf5 c3 Nc2
Bg5
14 +0.57 00:11 7056k f5 exf5 gxf5 Nh5 Bh8 Qd1 Be8 Rg1 Rd8 Bd2 Bxh5
gxh5 Kf7 b3 Bg7
15 < +0.40 00:13 7990k f5 exf5 gxf5 Nh5 Bh8 Qd1 Be8 Rg1 Rd8 Bd2 Bxh5
gxh5 Kf7 Ne4 Qxa4 c3
15 +0.44 00:18 10982k f5 exf5 gxf5 Nh5 Bh8 Qd1 Be8 Rg1 Bg6 Bd2 Qb6 Ne2
Ndc6 Bc3 Bxh5 Bxc6 Bxc3 Nxc3 fxg4 hxg4 Nxc6 gxh5+
Kf7
16 +0.40 00:19 11732k f5 exf5 gxf5 Nh5 Bh8 Qd1 Qb6 Re1 Rf8 Bf2 Ng6 g5
Rf7 Kg1 Qb4 Kh2 Bxa4 Bxd4 Qxd4 Nxa4 Qxa4 Rxe6
Nxf4 Rxd6 Nxh5 Qxh5 Qxc2
17 +0.40 00:40 24133k f5 exf5 gxf5 Qd1 Qb6 Rg1 Be8 Re1 Qc7 Bf2 Bc6 Bxc6
bxc6 Nh5 Ng6 b3 Qf7 g5 Nf3+ Qxf3 Bxc3 Re2 Rb4
18 < +0.32 00:52 31195k f5 exf5 gxf5 Qd1 Qb6 Rg1 Be8 Re1 Qc7 Bf2 Bc6 Bxc6
bxc6 Nh5 Ng6 Na2 fxg4 hxg4 Qf7 c3
18 +0.36 01:12 42650k f5 exf5 gxf5 Qd1 Qb6 Rg1 fxg4 hxg4 Rf8 Nh5 Bh8 b3
Ndc6 Bd2 Bd4 Re1 d5 Ne2 Bf2 Rf1
So it takes 11 seconds to find the right defensive plan in single thread mode.
After Re8 as happened in your game:
[d] 1r2r1k1/1p1bnpbp/p2pp1p1/2p5/Pq1nPPP1/2NPB1NP/1PPQ2BK/1R3R2 w - -
2 -0.61 00:00 4427 f5 exf5 gxf5 gxf5
3 -0.61 00:00 5043 f5 exf5 gxf5 gxf5
4 -0.61 00:00 6277 f5 exf5 gxf5 gxf5
5 -0.81 00:00 10302 f5 exf5 gxf5 gxf5 Bxd4 cxd4
6 -0.81 00:00 15565 f5 exf5 gxf5 gxf5 Bxd4 cxd4
7 < -0.93 00:00 21063 f5 exf5 gxf5 gxf5 Bxd4 cxd4
7 -0.81 00:00 32979 Nge2 b6 b3 Nec6 Rbe1 Rbc8 Nxd4 cxd4
8 > -0.73 00:00 50470 Nge2 Bxa4 Nxa4 Qxd2 Bxd2 Nxe2 e5
8 > -0.65 00:00 60519 Nge2 b6 b3 Nec6 Rbe1 Rbc8 g5
8 -0.73 00:00 76142 Nge2 Rbc8 b3 Nec6 Rbe1 f5 exf5 gxf5 Nxd4 cxd4
8 -0.65 00:00 95027 Qc1 f5 Ra1 Rf8 Re1 b6 exf5 gxf5 Bxd4 Bxd4
9 -0.65 00:00 106279 Qc1 f5 Ra1 Rf8 Re1 b6 exf5 gxf5 Bxd4 Bxd4
10 < -0.73 00:00 183704 Qc1 Qa5 Ra1 b6 g5 Nec6
10 > -0.57 00:00 224504 Qc1 Qa5 Ra1 b6 g5 Nec6 Nce2
10 -0.61 00:00 249940 Qc1 Qa5 Ra1 Qc7 Nge2 Nec6 Bf2 Nxe2 Nxe2 Nb4 Nc3
11 < -0.69 00:00 261112 Qc1 Qa5 g5 Nec6 Re1 Nb4
11 -0.69 00:00 304493 Qc1 Qa5 Ra1 Qc7 Bd2 b6 Rf2 f5 Kg1 c4 gxf5 gxf5
12 -0.69 00:00 397421 Qc1 Qa5 Ra1 Qc7 Bd2 b6 Rf2 f5 Kg1 c4 gxf5 gxf5
13 > -0.61 00:01 712053 Qc1 b6 Ra1 f5 Re1 fxg4 hxg4 e5 f5
13 -0.65 00:01 833237 Qc1 Qa5 Ra1 Qc7 Qd2 Rbd8 Rae1 Qb6 Qc1 Qb4 f5 exf5
gxf5 Bxa4 Nxa4 Qxa4 fxg6 hxg6
14 -0.69 00:02 1354k Qc1 Qa5 Bd2 Qc7 Rf2 f5 Ra1 b6 Be3 Rbd8 Kg1 Nec6
gxf5 exf5
14 > -0.57 00:03 2060k e5 dxe5 fxe5 Bxe5 Bf4 Bxf4 Qxf4 f5 Nge4 fxe4 Qf7+
Kh8 Qf6+ Kg8 Qf7+
15 -0.65 00:05 3200k Qc1 Qa5 Bd2 Qc7 Rf2 b6 Ra1 f5 Be3 Rf8 exf5 exf5
Nd5 Nxd5 Bxd5+ Be6 Bxe6+ Nxe6 gxf5
15 > -0.61 00:06 3683k e5 dxe5 fxe5 Bxe5 Bf4 Bxf4 Qxf4 f5 Nge4 fxe4 Qf7+
Kh8 Qf6+ Kg8 Qf7+
15 > -0.53 00:06 3892k e5 dxe5 fxe5 Bxe5 Bf4 Bxf4 Qxf4 f5 Nge4 fxe4 Qf7+
Kh8 Qf6+ Kg8 Qf7+
15 -0.53 00:07 4292k e5 d5 Qf2 Rec8 Rfc1 Nec6 Nxd5 exd5 c3 Qxa4 Ra1
Qb5 cxd4 cxd4 Bxd4 Qxd3
16 -0.57 00:10 6303k e5 d5 Qf2 Rec8 Bd2 Qb6 Be3 Nec6 Nge2 f6 f5 gxf5
gxf5 Nxe5 fxe6 Qxe6 Nxd4 cxd4 Bxd4
17 > -0.44 00:13 8140k e5 d5 Qf2 Rec8 Bd2 Qb6 Be3 Nec6 Na2 Qa5 b3
17 > -0.32 00:15 9117k e5 d5 Qf2 Nec6 Bxd5 exd5 Nxd5 Qxa4 Nb6
17 > -0.08 00:21 12661k e5 d5 Qf2 Nec6 Bxd5 exd5 Nxd5 Qxa4 Nb6
17 -0.16 00:28 16558k e5 d5 Qf2 a5 Nb5 Bxb5 axb5 Qxb5 Bxd4 cxd4 Ne2 Nc6
Nxd4 Nxd4 Qxd4 Rbc8 c3 Bf8 Qe3 Bc5 d4 Bb6 Kg3
18 -0.08 00:45 26408k e5 d5 Qf2 Rec8 Bd2 Qb6 Be3 a5 Bxd4 cxd4 Nge2 f6
f5 gxf5 Nxd4 Qc7 gxf5 exf5 Nxd5 Qxe5+ Qg3 Nxd5
Qxe5 fxe5 Bxd5+ Kh8
19 < -0.36 01:02 36301k e5 d5 Qf2 Rec8 Bd2 Qb6 Be3 a5 Bxd4 cxd4 Nge2 f6
f5 gxf5 Nxd4 fxg4 exf6 Rf8 f7+ Kh8 Ncb5 Be5+ Kh1
g3
19 +0.00 01:30 52985k f5 Nec6 Bh6 Be5 fxe6 fxe6 Rf4 Qb6 Rf2 Kh8 Rf7 Re7
Rf2 Ree8
20 > +0.12 02:08 74816k f5 gxf5 gxf5 exf5 Bf4 Qb6 Nh5 Bh8 a5 Qc7 exf5
Bxf5 Ne4 Bxe4 Bxe4 f5 Qg2+
20 > +0.24 02:34 90107k f5
20 +0.16 03:53 137125k f5 Nec6 Bh6 Be5 fxe6 Bxe6 Qf2 Kh8 Bg5 Rg8 Nd5
Qxa4 c3 Nb3 Qe3 Qa2 Bf6+ Bxf6 Nxf6 Rg7 Ne2 Ne5
Nf4 Qa4 Bf3 c4
21 > +0.36 04:56 173366k f5 exf5 exf5 gxf5 Nh5 fxg4 hxg4 Bh8 Bh6
21 > +0.57 07:03 247818k f5 Nec6 Bh6 Be5 fxg6 hxg6 Qg5 Ne7 Rf2 Ndc6 Qh4 f5
Rff1 Rbc8 Bg5 Rc7 gxf5 exf5 Kh1 c4 exf5 cxd3 f6
Bxf6 Qxb4 Nxb4 Bxf6
21 +0.48 07:23 259506k f5 Nec6 Bh6 Be5 fxg6 hxg6 Qg5 Ne7 Rf2 Ndc6 Qh4 f5
Rff1 Qd4 Bg5 b5 axb5 axb5 gxf5 gxf5 Qh6 Nd8 Be3
Bg7 Qf4 Qe5 exf5 exf5 Qxe5 Bxe5 Ra1 b4 Nd5 Nxd5
Bxd5+ Ne6
It takes 90 seconds to spot that f5 gives advantage, but still score is far from crushing.
1 thread, core i7:
2 +1.09 00:00 2846 Bxa4 Nxa4 Qxa4
3 +0.93 00:00 4589 Bxa4 Nxa4 Qxa4 b3
4 +1.05 00:00 6444 Bxa4 Nxa4 Qxa4 c3 Nb3
5 +0.93 00:00 8005 Bxa4 Nxa4 Qxa4 c3 Nb3 Qc2
6 +0.93 00:00 10403 Bxa4 Nxa4 Qxa4 c3 Nb3 Qc2 Nc6
7 < +0.85 00:00 14151 Bxa4 Nxa4 Qxa4 c3 Nb3 Qc2 Nc6 f5
7 +0.93 00:00 30905 Bxa4 Nxa4 Qxa4 c3 Nb3 Qc2 Nc6 g5
8 < +0.85 00:00 32157 Bxa4 Nxa4 Qxa4 c3 Nb3 Qc2 Nc6 g5
8 < +0.77 00:00 37302 Bxa4 Nxa4 Qxa4 c3 Nb3 Qc2 Nc6 g5
8 +0.81 00:00 83528 Rd8 Rfe1 Bxa4 Nxa4 Qxa4 Bxd4 cxd4 Ra1 Qb5
9 +0.73 00:00 132030 Rd8 Qf2 f5 exf5 gxf5 Rfe1 Bxa4 Nxa4 Qxa4 b3
10 < +0.57 00:00 271304 Rd8 Qf2 f5 exf5 gxf5 Rfe1 Bxa4 Nxa4 Qxa4 c3
10 +0.81 00:00 422871 Re8 g5 d5 Qc1 b6 Ra1 Bc6 Bd2 dxe4 dxe4
11 +0.85 00:01 605584 Re8 g5 Rf8 Nge2 f5 b3 fxe4 dxe4 Nec6 Rbd1 e5
12 < +0.77 00:01 718523 Re8 g5 b6 Nge2 d5 b3 dxe4 dxe4
12 < +0.69 00:01 950968 Re8 f5 gxf5 gxf5 exf5 Nh5 Be5+ Bf4
12 < +0.53 00:01 1171k Re8 f5 gxf5 gxf5 exf5 Nh5 Be5+ Bf4
12 +0.77 00:03 1841k Re8 g5 Rf8 Nge2 f5 b3 fxe4 dxe4 Nec6 Rbd1 Rbc8
Qd3 Rf7 Nxd4 cxd4
13 < +0.65 00:03 2189k Re8 f5 exf5 gxf5 Be5 fxg6 hxg6 Bf4 Bxa4 Nxa4 Qxa4
c3 Bxf4 Qxf4
13 < +0.53 00:04 2667k Re8 f5 exf5 gxf5 Be5 fxg6 hxg6 Bf4 Bxa4 Nxa4 Qxa4
c3 Bxf4 Qxf4
13 < +0.28 00:05 3369k Re8 f5 exf5 gxf5 Be5 fxg6 hxg6 Bf4 Bxa4 Nxa4 Qxa4
c3 Bxf4 Qxf4
13 +0.53 00:05 3568k Rd8 Qf2 f5 exf5 gxf5 Rfe1 Bf6 Nh5 Bh8 Qh4 Ng6 Qf2
Bxa4 Nxa4 Qxa4
14 < +0.24 00:07 4586k Rd8 Qf2 Bxa4 Nxa4 Qxa4 f5 exf5 exf5 gxf5 c3 Nc2
Bg5
14 +0.57 00:11 7056k f5 exf5 gxf5 Nh5 Bh8 Qd1 Be8 Rg1 Rd8 Bd2 Bxh5
gxh5 Kf7 b3 Bg7
15 < +0.40 00:13 7990k f5 exf5 gxf5 Nh5 Bh8 Qd1 Be8 Rg1 Rd8 Bd2 Bxh5
gxh5 Kf7 Ne4 Qxa4 c3
15 +0.44 00:18 10982k f5 exf5 gxf5 Nh5 Bh8 Qd1 Be8 Rg1 Bg6 Bd2 Qb6 Ne2
Ndc6 Bc3 Bxh5 Bxc6 Bxc3 Nxc3 fxg4 hxg4 Nxc6 gxh5+
Kf7
16 +0.40 00:19 11732k f5 exf5 gxf5 Nh5 Bh8 Qd1 Qb6 Re1 Rf8 Bf2 Ng6 g5
Rf7 Kg1 Qb4 Kh2 Bxa4 Bxd4 Qxd4 Nxa4 Qxa4 Rxe6
Nxf4 Rxd6 Nxh5 Qxh5 Qxc2
17 +0.40 00:40 24133k f5 exf5 gxf5 Qd1 Qb6 Rg1 Be8 Re1 Qc7 Bf2 Bc6 Bxc6
bxc6 Nh5 Ng6 b3 Qf7 g5 Nf3+ Qxf3 Bxc3 Re2 Rb4
18 < +0.32 00:52 31195k f5 exf5 gxf5 Qd1 Qb6 Rg1 Be8 Re1 Qc7 Bf2 Bc6 Bxc6
bxc6 Nh5 Ng6 Na2 fxg4 hxg4 Qf7 c3
18 +0.36 01:12 42650k f5 exf5 gxf5 Qd1 Qb6 Rg1 fxg4 hxg4 Rf8 Nh5 Bh8 b3
Ndc6 Bd2 Bd4 Re1 d5 Ne2 Bf2 Rf1
So it takes 11 seconds to find the right defensive plan in single thread mode.
After Re8 as happened in your game:
[d] 1r2r1k1/1p1bnpbp/p2pp1p1/2p5/Pq1nPPP1/2NPB1NP/1PPQ2BK/1R3R2 w - -
2 -0.61 00:00 4427 f5 exf5 gxf5 gxf5
3 -0.61 00:00 5043 f5 exf5 gxf5 gxf5
4 -0.61 00:00 6277 f5 exf5 gxf5 gxf5
5 -0.81 00:00 10302 f5 exf5 gxf5 gxf5 Bxd4 cxd4
6 -0.81 00:00 15565 f5 exf5 gxf5 gxf5 Bxd4 cxd4
7 < -0.93 00:00 21063 f5 exf5 gxf5 gxf5 Bxd4 cxd4
7 -0.81 00:00 32979 Nge2 b6 b3 Nec6 Rbe1 Rbc8 Nxd4 cxd4
8 > -0.73 00:00 50470 Nge2 Bxa4 Nxa4 Qxd2 Bxd2 Nxe2 e5
8 > -0.65 00:00 60519 Nge2 b6 b3 Nec6 Rbe1 Rbc8 g5
8 -0.73 00:00 76142 Nge2 Rbc8 b3 Nec6 Rbe1 f5 exf5 gxf5 Nxd4 cxd4
8 -0.65 00:00 95027 Qc1 f5 Ra1 Rf8 Re1 b6 exf5 gxf5 Bxd4 Bxd4
9 -0.65 00:00 106279 Qc1 f5 Ra1 Rf8 Re1 b6 exf5 gxf5 Bxd4 Bxd4
10 < -0.73 00:00 183704 Qc1 Qa5 Ra1 b6 g5 Nec6
10 > -0.57 00:00 224504 Qc1 Qa5 Ra1 b6 g5 Nec6 Nce2
10 -0.61 00:00 249940 Qc1 Qa5 Ra1 Qc7 Nge2 Nec6 Bf2 Nxe2 Nxe2 Nb4 Nc3
11 < -0.69 00:00 261112 Qc1 Qa5 g5 Nec6 Re1 Nb4
11 -0.69 00:00 304493 Qc1 Qa5 Ra1 Qc7 Bd2 b6 Rf2 f5 Kg1 c4 gxf5 gxf5
12 -0.69 00:00 397421 Qc1 Qa5 Ra1 Qc7 Bd2 b6 Rf2 f5 Kg1 c4 gxf5 gxf5
13 > -0.61 00:01 712053 Qc1 b6 Ra1 f5 Re1 fxg4 hxg4 e5 f5
13 -0.65 00:01 833237 Qc1 Qa5 Ra1 Qc7 Qd2 Rbd8 Rae1 Qb6 Qc1 Qb4 f5 exf5
gxf5 Bxa4 Nxa4 Qxa4 fxg6 hxg6
14 -0.69 00:02 1354k Qc1 Qa5 Bd2 Qc7 Rf2 f5 Ra1 b6 Be3 Rbd8 Kg1 Nec6
gxf5 exf5
14 > -0.57 00:03 2060k e5 dxe5 fxe5 Bxe5 Bf4 Bxf4 Qxf4 f5 Nge4 fxe4 Qf7+
Kh8 Qf6+ Kg8 Qf7+
15 -0.65 00:05 3200k Qc1 Qa5 Bd2 Qc7 Rf2 b6 Ra1 f5 Be3 Rf8 exf5 exf5
Nd5 Nxd5 Bxd5+ Be6 Bxe6+ Nxe6 gxf5
15 > -0.61 00:06 3683k e5 dxe5 fxe5 Bxe5 Bf4 Bxf4 Qxf4 f5 Nge4 fxe4 Qf7+
Kh8 Qf6+ Kg8 Qf7+
15 > -0.53 00:06 3892k e5 dxe5 fxe5 Bxe5 Bf4 Bxf4 Qxf4 f5 Nge4 fxe4 Qf7+
Kh8 Qf6+ Kg8 Qf7+
15 -0.53 00:07 4292k e5 d5 Qf2 Rec8 Rfc1 Nec6 Nxd5 exd5 c3 Qxa4 Ra1
Qb5 cxd4 cxd4 Bxd4 Qxd3
16 -0.57 00:10 6303k e5 d5 Qf2 Rec8 Bd2 Qb6 Be3 Nec6 Nge2 f6 f5 gxf5
gxf5 Nxe5 fxe6 Qxe6 Nxd4 cxd4 Bxd4
17 > -0.44 00:13 8140k e5 d5 Qf2 Rec8 Bd2 Qb6 Be3 Nec6 Na2 Qa5 b3
17 > -0.32 00:15 9117k e5 d5 Qf2 Nec6 Bxd5 exd5 Nxd5 Qxa4 Nb6
17 > -0.08 00:21 12661k e5 d5 Qf2 Nec6 Bxd5 exd5 Nxd5 Qxa4 Nb6
17 -0.16 00:28 16558k e5 d5 Qf2 a5 Nb5 Bxb5 axb5 Qxb5 Bxd4 cxd4 Ne2 Nc6
Nxd4 Nxd4 Qxd4 Rbc8 c3 Bf8 Qe3 Bc5 d4 Bb6 Kg3
18 -0.08 00:45 26408k e5 d5 Qf2 Rec8 Bd2 Qb6 Be3 a5 Bxd4 cxd4 Nge2 f6
f5 gxf5 Nxd4 Qc7 gxf5 exf5 Nxd5 Qxe5+ Qg3 Nxd5
Qxe5 fxe5 Bxd5+ Kh8
19 < -0.36 01:02 36301k e5 d5 Qf2 Rec8 Bd2 Qb6 Be3 a5 Bxd4 cxd4 Nge2 f6
f5 gxf5 Nxd4 fxg4 exf6 Rf8 f7+ Kh8 Ncb5 Be5+ Kh1
g3
19 +0.00 01:30 52985k f5 Nec6 Bh6 Be5 fxe6 fxe6 Rf4 Qb6 Rf2 Kh8 Rf7 Re7
Rf2 Ree8
20 > +0.12 02:08 74816k f5 gxf5 gxf5 exf5 Bf4 Qb6 Nh5 Bh8 a5 Qc7 exf5
Bxf5 Ne4 Bxe4 Bxe4 f5 Qg2+
20 > +0.24 02:34 90107k f5
20 +0.16 03:53 137125k f5 Nec6 Bh6 Be5 fxe6 Bxe6 Qf2 Kh8 Bg5 Rg8 Nd5
Qxa4 c3 Nb3 Qe3 Qa2 Bf6+ Bxf6 Nxf6 Rg7 Ne2 Ne5
Nf4 Qa4 Bf3 c4
21 > +0.36 04:56 173366k f5 exf5 exf5 gxf5 Nh5 fxg4 hxg4 Bh8 Bh6
21 > +0.57 07:03 247818k f5 Nec6 Bh6 Be5 fxg6 hxg6 Qg5 Ne7 Rf2 Ndc6 Qh4 f5
Rff1 Rbc8 Bg5 Rc7 gxf5 exf5 Kh1 c4 exf5 cxd3 f6
Bxf6 Qxb4 Nxb4 Bxf6
21 +0.48 07:23 259506k f5 Nec6 Bh6 Be5 fxg6 hxg6 Qg5 Ne7 Rf2 Ndc6 Qh4 f5
Rff1 Qd4 Bg5 b5 axb5 axb5 gxf5 gxf5 Qh6 Nd8 Be3
Bg7 Qf4 Qe5 exf5 exf5 Qxe5 Bxe5 Ra1 b4 Nd5 Nxd5
Bxd5+ Ne6
It takes 90 seconds to spot that f5 gives advantage, but still score is far from crushing.
Joona Kiiski
-
michiguel
- Posts: 6401
- Joined: Thu Mar 09, 2006 8:30 pm
- Location: Chicago, Illinois, USA
Re: Test Position
Absolutely. It is the first move any Sicilian player will consider and the lastzullil wrote:[d]1rr3k1/1p1bnpbp/p2pp1p1/2p5/Pq1nPPP1/2NPB1NP/1PPQ2BK/1R3R2 b - - 4 16
What do your favorite engines choose here? Is f5 necessary?
Miguel
-
swami
- Posts: 6664
- Joined: Thu Mar 09, 2006 4:21 am
Re: Test Position
f5 puts end or slows down white's aggressive plans on king side.
Therefore it's necessary.
Therefore it's necessary.
-
swami
- Posts: 6664
- Joined: Thu Mar 09, 2006 4:21 am
Re: Test Position
I don't think it's white's turn to move. From initial post it appears that it's actually Black's turn to move.
Your Stockfish analysis is with white playing f5.
Your Stockfish analysis is with white playing f5.
zamar wrote:Here is what current stockfish devel version thinks about the position:
1 thread, core i7:
2 +1.09 00:00 2846 Bxa4 Nxa4 Qxa4
3 +0.93 00:00 4589 Bxa4 Nxa4 Qxa4 b3
4 +1.05 00:00 6444 Bxa4 Nxa4 Qxa4 c3 Nb3
5 +0.93 00:00 8005 Bxa4 Nxa4 Qxa4 c3 Nb3 Qc2
6 +0.93 00:00 10403 Bxa4 Nxa4 Qxa4 c3 Nb3 Qc2 Nc6
7 < +0.85 00:00 14151 Bxa4 Nxa4 Qxa4 c3 Nb3 Qc2 Nc6 f5
7 +0.93 00:00 30905 Bxa4 Nxa4 Qxa4 c3 Nb3 Qc2 Nc6 g5
8 < +0.85 00:00 32157 Bxa4 Nxa4 Qxa4 c3 Nb3 Qc2 Nc6 g5
8 < +0.77 00:00 37302 Bxa4 Nxa4 Qxa4 c3 Nb3 Qc2 Nc6 g5
8 +0.81 00:00 83528 Rd8 Rfe1 Bxa4 Nxa4 Qxa4 Bxd4 cxd4 Ra1 Qb5
9 +0.73 00:00 132030 Rd8 Qf2 f5 exf5 gxf5 Rfe1 Bxa4 Nxa4 Qxa4 b3
10 < +0.57 00:00 271304 Rd8 Qf2 f5 exf5 gxf5 Rfe1 Bxa4 Nxa4 Qxa4 c3
10 +0.81 00:00 422871 Re8 g5 d5 Qc1 b6 Ra1 Bc6 Bd2 dxe4 dxe4
11 +0.85 00:01 605584 Re8 g5 Rf8 Nge2 f5 b3 fxe4 dxe4 Nec6 Rbd1 e5
12 < +0.77 00:01 718523 Re8 g5 b6 Nge2 d5 b3 dxe4 dxe4
12 < +0.69 00:01 950968 Re8 f5 gxf5 gxf5 exf5 Nh5 Be5+ Bf4
12 < +0.53 00:01 1171k Re8 f5 gxf5 gxf5 exf5 Nh5 Be5+ Bf4
12 +0.77 00:03 1841k Re8 g5 Rf8 Nge2 f5 b3 fxe4 dxe4 Nec6 Rbd1 Rbc8
Qd3 Rf7 Nxd4 cxd4
13 < +0.65 00:03 2189k Re8 f5 exf5 gxf5 Be5 fxg6 hxg6 Bf4 Bxa4 Nxa4 Qxa4
c3 Bxf4 Qxf4
13 < +0.53 00:04 2667k Re8 f5 exf5 gxf5 Be5 fxg6 hxg6 Bf4 Bxa4 Nxa4 Qxa4
c3 Bxf4 Qxf4
13 < +0.28 00:05 3369k Re8 f5 exf5 gxf5 Be5 fxg6 hxg6 Bf4 Bxa4 Nxa4 Qxa4
c3 Bxf4 Qxf4
13 +0.53 00:05 3568k Rd8 Qf2 f5 exf5 gxf5 Rfe1 Bf6 Nh5 Bh8 Qh4 Ng6 Qf2
Bxa4 Nxa4 Qxa4
14 < +0.24 00:07 4586k Rd8 Qf2 Bxa4 Nxa4 Qxa4 f5 exf5 exf5 gxf5 c3 Nc2
Bg5
14 +0.57 00:11 7056k f5 exf5 gxf5 Nh5 Bh8 Qd1 Be8 Rg1 Rd8 Bd2 Bxh5
gxh5 Kf7 b3 Bg7
15 < +0.40 00:13 7990k f5 exf5 gxf5 Nh5 Bh8 Qd1 Be8 Rg1 Rd8 Bd2 Bxh5
gxh5 Kf7 Ne4 Qxa4 c3
15 +0.44 00:18 10982k f5 exf5 gxf5 Nh5 Bh8 Qd1 Be8 Rg1 Bg6 Bd2 Qb6 Ne2
Ndc6 Bc3 Bxh5 Bxc6 Bxc3 Nxc3 fxg4 hxg4 Nxc6 gxh5+
Kf7
16 +0.40 00:19 11732k f5 exf5 gxf5 Nh5 Bh8 Qd1 Qb6 Re1 Rf8 Bf2 Ng6 g5
Rf7 Kg1 Qb4 Kh2 Bxa4 Bxd4 Qxd4 Nxa4 Qxa4 Rxe6
Nxf4 Rxd6 Nxh5 Qxh5 Qxc2
17 +0.40 00:40 24133k f5 exf5 gxf5 Qd1 Qb6 Rg1 Be8 Re1 Qc7 Bf2 Bc6 Bxc6
bxc6 Nh5 Ng6 b3 Qf7 g5 Nf3+ Qxf3 Bxc3 Re2 Rb4
18 < +0.32 00:52 31195k f5 exf5 gxf5 Qd1 Qb6 Rg1 Be8 Re1 Qc7 Bf2 Bc6 Bxc6
bxc6 Nh5 Ng6 Na2 fxg4 hxg4 Qf7 c3
18 +0.36 01:12 42650k f5 exf5 gxf5 Qd1 Qb6 Rg1 fxg4 hxg4 Rf8 Nh5 Bh8 b3
Ndc6 Bd2 Bd4 Re1 d5 Ne2 Bf2 Rf1
So it takes 11 seconds to find the right defensive plan in single thread mode.
After Re8 as happened in your game:
[d] 1r2r1k1/1p1bnpbp/p2pp1p1/2p5/Pq1nPPP1/2NPB1NP/1PPQ2BK/1R3R2 w - -
2 -0.61 00:00 4427 f5 exf5 gxf5 gxf5
3 -0.61 00:00 5043 f5 exf5 gxf5 gxf5
4 -0.61 00:00 6277 f5 exf5 gxf5 gxf5
5 -0.81 00:00 10302 f5 exf5 gxf5 gxf5 Bxd4 cxd4
6 -0.81 00:00 15565 f5 exf5 gxf5 gxf5 Bxd4 cxd4
7 < -0.93 00:00 21063 f5 exf5 gxf5 gxf5 Bxd4 cxd4
7 -0.81 00:00 32979 Nge2 b6 b3 Nec6 Rbe1 Rbc8 Nxd4 cxd4
8 > -0.73 00:00 50470 Nge2 Bxa4 Nxa4 Qxd2 Bxd2 Nxe2 e5
8 > -0.65 00:00 60519 Nge2 b6 b3 Nec6 Rbe1 Rbc8 g5
8 -0.73 00:00 76142 Nge2 Rbc8 b3 Nec6 Rbe1 f5 exf5 gxf5 Nxd4 cxd4
8 -0.65 00:00 95027 Qc1 f5 Ra1 Rf8 Re1 b6 exf5 gxf5 Bxd4 Bxd4
9 -0.65 00:00 106279 Qc1 f5 Ra1 Rf8 Re1 b6 exf5 gxf5 Bxd4 Bxd4
10 < -0.73 00:00 183704 Qc1 Qa5 Ra1 b6 g5 Nec6
10 > -0.57 00:00 224504 Qc1 Qa5 Ra1 b6 g5 Nec6 Nce2
10 -0.61 00:00 249940 Qc1 Qa5 Ra1 Qc7 Nge2 Nec6 Bf2 Nxe2 Nxe2 Nb4 Nc3
11 < -0.69 00:00 261112 Qc1 Qa5 g5 Nec6 Re1 Nb4
11 -0.69 00:00 304493 Qc1 Qa5 Ra1 Qc7 Bd2 b6 Rf2 f5 Kg1 c4 gxf5 gxf5
12 -0.69 00:00 397421 Qc1 Qa5 Ra1 Qc7 Bd2 b6 Rf2 f5 Kg1 c4 gxf5 gxf5
13 > -0.61 00:01 712053 Qc1 b6 Ra1 f5 Re1 fxg4 hxg4 e5 f5
13 -0.65 00:01 833237 Qc1 Qa5 Ra1 Qc7 Qd2 Rbd8 Rae1 Qb6 Qc1 Qb4 f5 exf5
gxf5 Bxa4 Nxa4 Qxa4 fxg6 hxg6
14 -0.69 00:02 1354k Qc1 Qa5 Bd2 Qc7 Rf2 f5 Ra1 b6 Be3 Rbd8 Kg1 Nec6
gxf5 exf5
14 > -0.57 00:03 2060k e5 dxe5 fxe5 Bxe5 Bf4 Bxf4 Qxf4 f5 Nge4 fxe4 Qf7+
Kh8 Qf6+ Kg8 Qf7+
15 -0.65 00:05 3200k Qc1 Qa5 Bd2 Qc7 Rf2 b6 Ra1 f5 Be3 Rf8 exf5 exf5
Nd5 Nxd5 Bxd5+ Be6 Bxe6+ Nxe6 gxf5
15 > -0.61 00:06 3683k e5 dxe5 fxe5 Bxe5 Bf4 Bxf4 Qxf4 f5 Nge4 fxe4 Qf7+
Kh8 Qf6+ Kg8 Qf7+
15 > -0.53 00:06 3892k e5 dxe5 fxe5 Bxe5 Bf4 Bxf4 Qxf4 f5 Nge4 fxe4 Qf7+
Kh8 Qf6+ Kg8 Qf7+
15 -0.53 00:07 4292k e5 d5 Qf2 Rec8 Rfc1 Nec6 Nxd5 exd5 c3 Qxa4 Ra1
Qb5 cxd4 cxd4 Bxd4 Qxd3
16 -0.57 00:10 6303k e5 d5 Qf2 Rec8 Bd2 Qb6 Be3 Nec6 Nge2 f6 f5 gxf5
gxf5 Nxe5 fxe6 Qxe6 Nxd4 cxd4 Bxd4
17 > -0.44 00:13 8140k e5 d5 Qf2 Rec8 Bd2 Qb6 Be3 Nec6 Na2 Qa5 b3
17 > -0.32 00:15 9117k e5 d5 Qf2 Nec6 Bxd5 exd5 Nxd5 Qxa4 Nb6
17 > -0.08 00:21 12661k e5 d5 Qf2 Nec6 Bxd5 exd5 Nxd5 Qxa4 Nb6
17 -0.16 00:28 16558k e5 d5 Qf2 a5 Nb5 Bxb5 axb5 Qxb5 Bxd4 cxd4 Ne2 Nc6
Nxd4 Nxd4 Qxd4 Rbc8 c3 Bf8 Qe3 Bc5 d4 Bb6 Kg3
18 -0.08 00:45 26408k e5 d5 Qf2 Rec8 Bd2 Qb6 Be3 a5 Bxd4 cxd4 Nge2 f6
f5 gxf5 Nxd4 Qc7 gxf5 exf5 Nxd5 Qxe5+ Qg3 Nxd5
Qxe5 fxe5 Bxd5+ Kh8
19 < -0.36 01:02 36301k e5 d5 Qf2 Rec8 Bd2 Qb6 Be3 a5 Bxd4 cxd4 Nge2 f6
f5 gxf5 Nxd4 fxg4 exf6 Rf8 f7+ Kh8 Ncb5 Be5+ Kh1
g3
19 +0.00 01:30 52985k f5 Nec6 Bh6 Be5 fxe6 fxe6 Rf4 Qb6 Rf2 Kh8 Rf7 Re7
Rf2 Ree8
20 > +0.12 02:08 74816k f5 gxf5 gxf5 exf5 Bf4 Qb6 Nh5 Bh8 a5 Qc7 exf5
Bxf5 Ne4 Bxe4 Bxe4 f5 Qg2+
20 > +0.24 02:34 90107k f5
20 +0.16 03:53 137125k f5 Nec6 Bh6 Be5 fxe6 Bxe6 Qf2 Kh8 Bg5 Rg8 Nd5
Qxa4 c3 Nb3 Qe3 Qa2 Bf6+ Bxf6 Nxf6 Rg7 Ne2 Ne5
Nf4 Qa4 Bf3 c4
21 > +0.36 04:56 173366k f5 exf5 exf5 gxf5 Nh5 fxg4 hxg4 Bh8 Bh6
21 > +0.57 07:03 247818k f5 Nec6 Bh6 Be5 fxg6 hxg6 Qg5 Ne7 Rf2 Ndc6 Qh4 f5
Rff1 Rbc8 Bg5 Rc7 gxf5 exf5 Kh1 c4 exf5 cxd3 f6
Bxf6 Qxb4 Nxb4 Bxf6
21 +0.48 07:23 259506k f5 Nec6 Bh6 Be5 fxg6 hxg6 Qg5 Ne7 Rf2 Ndc6 Qh4 f5
Rff1 Qd4 Bg5 b5 axb5 axb5 gxf5 gxf5 Qh6 Nd8 Be3
Bg7 Qf4 Qe5 exf5 exf5 Qxe5 Bxe5 Ra1 b4 Nd5 Nxd5
Bxd5+ Ne6
It takes 90 seconds to spot that f5 gives advantage, but still score is far from crushing.
-
zamar
- Posts: 613
- Joined: Sun Jan 18, 2009 7:03 am
Re: Test Position
First analysis is black to move. Second analysis is after black's Re8 which happened in Louis' fantastic game.swami wrote:I don't think it's white's turn to move. From initial post it appears that it's actually Black's turn to move.
Your Stockfish analysis is with white playing f5.
Joona Kiiski
-
zullil
- Posts: 6442
- Joined: Tue Jan 09, 2007 12:31 am
- Location: PA USA
- Full name: Louis Zulli
Re: Test Position
After an hour at 23M nps, Crafty chooses Nec6:zullil wrote:[d]1rr3k1/1p1bnpbp/p2pp1p1/2p5/Pq1nPPP1/2NPB1NP/1PPQ2BK/1R3R2 b - - 4 16
What do your favorite engines choose here? Is f5 necessary?
Code: Select all
Crafty v23.2 (8 cpus)
White(1): setboard 1rr3k1/1p1bnpbp/p2pp1p1/2p5/Pq1nPPP1/2NPB1NP/1PPQ2BK/1R3R2 b - - 4 16
Black(1): st 3600
search time set to 3600.00.
Black(1): display stats
display statistics at end of each search.
Black(1): display
+---+---+---+---+---+---+---+---+
8 | |<R>|<R>| . | | . |<K>| . |
+---+---+---+---+---+---+---+---+
7 | . |<P>| . |<B>|<N>|<P>|<B>|<P>|
+---+---+---+---+---+---+---+---+
6 |<P>| . | |<P>|<P>| . |<P>| . |
+---+---+---+---+---+---+---+---+
5 | . | |<P>| | . | | . | |
+---+---+---+---+---+---+---+---+
4 |-P-|<Q>| |<N>|-P-|-P-|-P-| . |
+---+---+---+---+---+---+---+---+
3 | . | |-N-|-P-|-B-| |-N-|-P-|
+---+---+---+---+---+---+---+---+
2 | |-P-|-P-|-Q-| | . |-B-|-K-|
+---+---+---+---+---+---+---+---+
1 | . |-R-| . | | . |-R-| . | |
+---+---+---+---+---+---+---+---+
a b c d e f g h
Black(1): go
time limit 60:00 (+0.00) (60:00)
depth time score variation (1)
starting thread 1
starting thread 2
starting thread 3
starting thread 4
starting thread 5
starting thread 6
starting thread 7
7-> 0.05 -0.47 1. ... b6 2. Nge2 Bxa4 3. Nxa4 Qxd2
4. Bxd2 Nxe2 (s=2)
8 0.07 -0.44 1. ... b6 2. f5 gxf5 3. exf5 exf5 4.
Bg5 Re8 5. Rbe1
8 0.09 -0.46 1. ... Bxa4 2. Nxa4 Qxa4 3. b3 Qb5
4. f5 gxf5 5. exf5 exf5 6. Nxf5 Nexf5
7. gxf5
8-> 0.09 -0.46 1. ... Bxa4 2. Nxa4 Qxa4 3. b3 Qb5
4. f5 gxf5 5. exf5 exf5 6. Nxf5 Nexf5
7. gxf5 (s=3)
9 0.09 -0.46 1. ... Bxa4 2. Nxa4 Qxa4 3. b3 Qb5
4. f5 gxf5 5. exf5 exf5 6. Nxf5 Nexf5
7. gxf5 (s=2)
9 0.19 -0.50 1. ... b6 2. Qf2 Rf8 3. Bd2 Nec6 4.
Rfc1 c4 5. Nge2 Bf6 6. Nxd4 Nxd4
9-> 0.21 -0.50 1. ... b6 2. Qf2 Rf8 3. Bd2 Nec6 4.
Rfc1 c4 5. Nge2 Bf6 6. Nxd4 Nxd4
10 0.29 -0.48 1. ... b6 2. h4 Bxa4 3. Nxa4 Qxa4 4.
b3 Qb5 5. f5 gxf5 6. exf5 exf5 7. Bxd4
Bxd4 8. Nxf5 Nxf5 9. Rxf5
10-> 0.33 -0.48 1. ... b6 2. h4 Bxa4 3. Nxa4 Qxa4 4.
b3 Qb5 5. f5 gxf5 6. exf5 exf5 7. Bxd4
Bxd4 8. Nxf5 Nxf5 9. Rxf5
11 0.44 -0.27 1. ... b6 2. f5 exf5 3. exf5 gxf5 4.
Bh6 Re8 5. Bxg7 Kxg7 6. Qg5+ Kf8 7.
Qh6+ Kg8 8. Qxd6 Bxa4 9. Nxa4 Qxa4
11 0.59 -0.34 1. ... Bxa4 2. Nxa4 Qxa4 3. c3 Nb3
4. Qe2 Re8 5. f5 exf5 6. exf5 Nc6 7.
fxg6 hxg6
11 1.00 -0.49 1. ... Re8 2. Nge2 Bxa4 3. Nxa4 Qxa4
4. c3 Nb3 5. Qc2 Qb5 6. g5 Nc6
11-> 1.29 -0.49 1. ... Re8 2. Nge2 Bxa4 3. Nxa4 Qxa4
4. c3 Nb3 5. Qc2 Qb5 6. g5 Nc6 (s=8)
12 1.44 -0.45 1. ... Re8 2. Nge2 b6 3. b3 Nec6 4.
Rbc1 f5 5. Rcd1 Bf6 6. g5 Nxe2 7. Nxe2
Qxd2 8. Rxd2 (s=7)
12-> 1.55 -0.45 1. ... Re8 2. Nge2 b6 3. b3 Nec6 4.
Rbc1 f5 5. Rcd1 Bf6 6. g5 Nxe2 7. Nxe2
Qxd2 8. Rxd2
13 2.28 -0.46 1. ... Re8 2. h4 Bxa4 3. Nxa4 Qxa4
4. c3 Nb3 5. Qe2 Nc6 6. h5 a5 7. g5
b5
13-> 2.61 -0.46 1. ... Re8 2. h4 Bxa4 3. Nxa4 Qxa4
4. c3 Nb3 5. Qe2 Nc6 6. h5 a5 7. g5
b5
14 4.40 -0.31 1. ... Re8 2. f5 exf5 3. exf5 gxf5
4. Nh5 Bh8 5. Bxd4 Bxd4 6. Qg5+ Ng6
7. Nf6+ Bxf6 8. Qxf6 f4 9. Qxd6
14-> 6.05 -0.31 1. ... Re8 2. f5 exf5 3. exf5 gxf5
4. Nh5 Bh8 5. Bxd4 Bxd4 6. Qg5+ Ng6
7. Nf6+ Bxf6 8. Qxf6 f4 9. Qxd6
15 7.70 -0.01 1. ... Re8 2. f5 exf5 3. exf5 gxf5
4. Nh5 Bh8 5. Bxd4 Qxd4 6. Qg5+ Kf8
7. Qh6+ Kg8 8. Qg5+
15 18.56 -0.09 1. ... Bxa4 2. Nxa4 Qxa4 3. c3 Nb3
4. Qe2 Re8 5. f5 exf5 6. exf5 gxf5
7. Nxf5 Nxf5 8. Rxf5 Be5+ 9. Kh1
15-> 20.96 -0.09 1. ... Bxa4 2. Nxa4 Qxa4 3. c3 Nb3
4. Qe2 Re8 5. f5 exf5 6. exf5 gxf5
7. Nxf5 Nxf5 8. Rxf5 Be5+ 9. Kh1 (s=14)
16 22.52 -0.18 1. ... Bxa4 2. Nxa4 Qxa4 3. c3 Nb3
4. Qc2 Re8 5. Rbe1 Kh8 6. g5 c4 7.
dxc4 Qxc4 8. Rd1 Rbd8 9. e5 (s=13)
16-> 26.17 -0.18 1. ... Bxa4 2. Nxa4 Qxa4 3. c3 Nb3
4. Qc2 Re8 5. Rbe1 Kh8 6. g5 c4 7.
dxc4 Qxc4 8. Rd1 Rbd8 9. e5 (s=2)
17 27.53 -0.19 1. ... Bxa4 2. Nxa4 Qxa4 3. c3 Nb3
4. Qc2 Re8 5. Rbe1 Kh8 6. g5 c4 7.
dxc4 Qxc4 8. Rd1 Rbd8 9. Qe2 d5 10.
Qxc4 dxc4
17-> 37.42 -0.19 1. ... Bxa4 2. Nxa4 Qxa4 3. c3 Nb3
4. Qc2 Re8 5. Rbe1 Kh8 6. g5 c4 7.
dxc4 Qxc4 8. Rd1 Rbd8 9. Qe2 d5 10.
Qxc4 dxc4 (s=8)
18 40.32 -0.19 1. ... Bxa4 2. Nxa4 Qxa4 3. c3 Nb3
4. Qc2 b5 5. f5 Nc6 6. f6 Bf8 7. Ne2
Ne5 8. g5 Rb7 9. Kh1 c4 10. d4 (s=7)
18-> 59.27 -0.19 1. ... Bxa4 2. Nxa4 Qxa4 3. c3 Nb3
4. Qc2 b5 5. f5 Nc6 6. f6 Bf8 7. Ne2
Ne5 8. g5 Rb7 9. Kh1 c4 10. d4 (s=6)
19 1:04 -0.14 1. ... Bxa4 2. Nxa4 Qxa4 3. c3 Nb3
4. Qc2 b5 5. f5 Nc6 6. f6 Bf8 7. Ne2
Ne5 8. g5 Rb7 9. Nf4 a5 10. Kh1 h6
(s=5)
19-> 1:58 -0.14 1. ... Bxa4 2. Nxa4 Qxa4 3. c3 Nb3
4. Qc2 b5 5. f5 Nc6 6. f6 Bf8 7. Ne2
Ne5 8. g5 Rb7 9. Nf4 a5 10. Kh1 h6
(s=9)
20 5:30 -0.09 1. ... Bxa4 2. f5 Nec6 3. Nxa4 Qxa4
4. f6 Bf8 5. b3 Qa3 6. Qd1 Qa5 7. Bd2
Qb6 8. c3 Nb5 9. g5 Ne5 10. Qe2 a5
11. Kh1 (s=8)
20-> 7:33 -0.09 1. ... Bxa4 2. f5 Nec6 3. Nxa4 Qxa4
4. f6 Bf8 5. b3 Qa3 6. Qd1 Qa5 7. Bd2
Qb6 8. c3 Nb5 9. g5 Ne5 10. Qe2 a5
11. Kh1
21 9:44 0.19 1. ... Bxa4 2. f5 Nec6 3. Bh6 Rc7 4.
Nxa4 Qxa4 5. f6 Bh8 6. c3 Nb3 7. Qe3
c4 8. Nh5 cxd3 9. Qxd3 Rd8 10. Ng7
Ne5 11. Qg3 b6
21 28:19 0.10 1. ... Rf8 2. f5 f6 3. fxe6 Bxe6 4.
Nge2 Nxe2 5. Qxe2 Nc6 6. Qd2 Nd4 7.
Bf4 Rbd8 8. Rf2 Rf7 9. Bh6 Bxh6 10.
Qxh6 Re8 11. Kh1 b6
21 28:54 0.01 1. ... Nec6 2. f5 Kh8 3. f6 Bf8 4.
Nge2 b5 5. axb5 axb5 6. Bh6 Ra8 7.
Nxd4 Nxd4 8. Bxf8 Rxf8 9. Qh6 <HT>
21-> 34:49 0.01 1. ... Nec6 2. f5 Kh8 3. f6 Bf8 4.
Nge2 b5 5. axb5 axb5 6. Bh6 Ra8 7.
Nxd4 Nxd4 8. Bxf8 Rxf8 9. Qh6 <HT>
22 37:42 0.08 1. ... Nec6 2. f5 Kh8 3. f6 Bf8 4.
Qc1 Qa5 5. Ra1 Nb4 6. Rf2 Qc7 7. Na2
Nxa2 8. Rxa2 d5 9. Bf4 Bd6 10. Bxd6
Qxd6 11. Qh6 Rg8 12. Ra1
22-> 47:02 0.08 1. ... Nec6 2. f5 Kh8 3. f6 Bf8 4.
Qc1 Qa5 5. Ra1 Nb4 6. Rf2 Qc7 7. Na2
Nxa2 8. Rxa2 d5 9. Bf4 Bd6 10. Bxd6
Qxd6 11. Qh6 Rg8 12. Ra1 (s=4)
23 53:05 0.13 1. ... Nec6 2. f5 Kh8 3. f6 Bf8 4.
Qc1 Qa5 5. Qd1 Ne5 6. Nge2 Qb4 7. Kh1
Bxa4 8. Nxa4 Qxa4 9. b3 Qc6 10. g5
Qb6 11. c3 Nb5 12. Qd2 a5 (s=3)
time=60:00 mat=0 n=83898943063 fh=52% nps=23.3M
extensions=1867.2M qchecks=1948.6M reduced=1811.9M pruned=846.0M
predicted=0 evals=251.1M 50move=0 EGTBprobes=0 hits=0
SMP-> splits=2354210 aborts=302629 data=67/1024 elap=60:00
terminating SMP processes.
Black(1): Nec6
time used: 60:00
-
Eelco de Groot
- Posts: 4696
- Joined: Sun Mar 12, 2006 2:40 am
- Full name: Eelco de Groot
Re: Test Position
I don't think that Rainbow Serpent is very good here, it does not really get enough depth because of re-searches using the PV-search if it does not trust the null window search and too much time in the PV-search especially for the first move because there it is using a PV-search with full depth. Mainly I think it values the exchange on a4 too high compared to f5 so maybe it still underestimates White's pawn storm if f5 is not played by Black. It does try Stockfish's pawn storm code even if both Kings have castled to the same side, the score is already computed even if it was not used so it is essentially free knowledge, but has to be toned down if there is still your own King behind the pawns, and maybe the code is too aggressive in this case even if you decrease the bonuspoints for it. It is difficult to get it right and I did not spend much time on tuning or changing the code. I guess the program would also need more knowledge of the Pawn shield in that case, I'm not a big fan of Rainbow Serpent's == Stockfish's present code for computing the Pawn shield.zullil wrote:[d]1rr3k1/1p1bnpbp/p2pp1p1/2p5/Pq1nPPP1/2NPB1NP/1PPQ2BK/1R3R2 b - - 4 16
What do your favorite engines choose here? Is f5 necessary?
This is what I use if Kings are castled to the same side, it is not very elegant code:
Code: Select all
// Pawn storms in positions with opposite castling.
// With Kings still in the middle or symmetric castling an attempt is
// made to use toned down storm code. Splitting the code further
// would maybe be advised but hopefully Pawn shield code helps enough
// to prevent weakening, in case the King has actually castled behind
// the Storm. [EdG]
if ( square_file(pos.king_square(WHITE)) >= FILE_E)
{
if (square_file(pos.king_square(BLACK)) <= FILE_D)
ei.value += make_score(ei.pi->queenside_storm_value(WHITE) - ei.pi->kingside_storm_value(BLACK), 0);
else
{
ei.value += make_score(((ei.pi->kingside_storm_value(WHITE)*(50 + phase + ei.KingDanger[BLACK]))- (ei.pi->kingside_storm_value(BLACK)*(50 + phase + ei.KingDanger[WHITE])))/1000, 0);
// Pawn structure -probably- becomes more important in case of same-side castling,
// material and center probably as well but maybe we should correct that differently, this is not yet done. [EdG]
ei.value += make_score(ei.pi->pawns_value()/10, 0);
}
}
else if (square_file(pos.king_square(WHITE)) <= FILE_D)
{
if ( square_file(pos.king_square(BLACK)) >= FILE_E)
ei.value += make_score(ei.pi->kingside_storm_value(WHITE) - ei.pi->queenside_storm_value(BLACK), 0);
else
{
ei.value += make_score(((ei.pi->queenside_storm_value(WHITE)*(50 + phase + ei.KingDanger[BLACK])) - (ei.pi->queenside_storm_value(BLACK)*(50 + phase + ei.KingDanger[WHITE])))/1000, 0);
ei.value += make_score(ei.pi->pawns_value()/12, 0);
}
}
Code: Select all
ei.KingDanger[BLACK] and ei.KingDanger[WHITE] you will not find in
Stockfish but it is not difficult to store this info separately in the eval info object. In Rainbow Serpent King Safety is added last to be able to correct it with the rest of the eval. [EdG]
[d]1rr3k1/1p1bnpbp/p2pp1p1/2p5/Pq1nPPP1/2NPB1NP/1PPQ2BK/1R3R2 b - -
Engine: Rainbow Serpent 1.6.3s(dc) Build 261 (Athlon 2009 MHz, 256 MB)
by Tord Romstad, Marco Costalba, Joona Kiiski Modifications: Dann Corbit
1.00 0:00 +1.21 1...Bxa4 (401) 1
2.00 0:00 +1.21 1...Bxa4 2.Nxa4 Qxa4 3.b3 Qe8 (3.951) 16
3.00 0:00 +1.21 1...Bxa4 2.Nxa4 Qxa4 3.c3 Ndc6 (17.498) 65
4.00 0:00 +1.01 1...Bxa4 2.Nxa4 Qxa4 3.c3 Nb3 4.Qc2 Nc6
5.Bf2 Bf6 6.Ne2 (22.736) 85
5.00 0:00 +1.17 1...Bxa4 2.Nxa4 Qxa4 3.c3 Nd5 4.Qc2 (130.345) 252
6.00 0:00 +1.05 1...Bxa4 2.Nxa4 Qxa4 3.c3 Nb3 4.Qc2 Bf6
5.f5 exf5 6.exf5 (188.846) 294
7.00 0:01 +0.88 1...Bxa4 2.Nxa4 Qxa4 3.b3 Qd7 4.c3 Nb5
5.Ne2 Nc6 6.Bf2 (627.789) 375
8.01 0:05 +0.92 1...Bxa4 2.Nxa4 Qxa4 3.b3 Qd7 4.c3 Nb5
5.Ne2 Nc6 6.g5 (2.250.267) 432
9.01 0:07 +0.68-- 1...Bxa4 2.Nxa4 Qxa4 3.b3 Qd7 4.c3 Nb5
5.Ne2 Nc6 6.g5 (3.185.065) 433
10.01 0:10 +0.76 1...Bxa4 2.Nxa4 Qxa4 3.b3 Qd7 4.c3 Nb5
5.Ne2 Nc6 6.g5 f5 7.exf5 gxf5 8.Bf2 (4.421.144) 434
11.01 1:28 +0.48 1...Bxa4 2.Nxa4 Qxa4 3.c3 Nb3 4.Qc2 d5
5.f5 d4 6.Bd2 exf5 7.exf5 Be5 8.fxg6 hxg6
9.Bg5 (39.932.864) 452
12.01 5:15 +0.05 1...Bxa4 2.Nxa4 Qxa4 3.c3 Nb3 4.Qc2 Qa2
5.f5 gxf5 6.exf5 exf5 7.Nxf5 Nxf5
8.Rxf5 Be5+ 9.Bf4 Bxf4+ 10.Rxf4 Qxb2
11.Rxb2 (142.852.041) 453
12.03 7:05 +0.20 1...f5 2.exf5 gxf5 3.Rfe1 Ng6 4.gxf5 Nh4
5.Bxd4 cxd4 6.fxe6 Bxa4 7.Nge4 dxc3
8.bxc3 Qa5 9.e7 Qxc3 10.Qe2 (190.758.604) 448
13.01 10:38 +0.27 1...f5 2.exf5 gxf5 3.Rfe1 Ng6 4.gxf5 Nh4
5.Bxd4 cxd4 6.fxe6 Bxa4 7.Nge4 dxc3
8.bxc3 Qa5 9.Nxd6 Rc6 10.Rxb7 Rxb7
11.Nxb7 Qxc3 (282.347.657) 441
14.01 28:40 +0.28 1...f5 2.exf5 gxf5 3.Rfe1 Ng6 4.Nh5 Bh8
5.Qf2 Rf8 6.Nd5 Qxa4 7.Nb6 Qb5
8.Nxd7 Qxd7 9.b4 fxg4 10.hxg4 Ne5 (766.524.198) 445
best move: f7-f5 time: 39:14.344 min n/s: 447.041 nodes: 1.052.490.213
[/code]
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
place. Therefore, if you write the code as cleverly as possible, you
are, by definition, not smart enough to debug it.
-- Brian W. Kernighan
-
Eelco de Groot
- Posts: 4696
- Joined: Sun Mar 12, 2006 2:40 am
- Full name: Eelco de Groot
Re: Test Position
Here are some deeper searches of the position, all with different Rainbow Serpents.
First one is with an older version, with a different search and eval, it is really tough to beat this, so really the challenge now is to see what are the better characteristics of the simpler, older version, and the newer slightly more complex versions.
There are also still bugs I'm afraid but those are harder for me to fix, some areas of the code like the transposition table code etc. I have not really studied and there are some subtle bugs suspected that only show up in debug mode, that the newer version of Stockfish maybe does not have, or were fixed in 1.6.3 which I have not studied yet fully or introduced in Rainbow Serpent. But this as an aside.
1rr3k1/1p1bnpbp/p2pp1p1/2p5/Pq1nPPP1/2NPB1NP/1PPQ2BK/1R3R2 b - -
Engine: Rainbow Serpent 1.6.3s(dc) Build 235 (Athlon 2009 MHz, 256 MB)
by Tord Romstad, Marco Costalba, Joona Kiiski Modifications: Dann Corbit
1.00 0:00 +1.25 1...Bxa4 (482) 1
2.00 0:00 +1.25 1...Bxa4 2.Nxa4 Qxa4 3.b3 Qe8 (3.564) 14
3.00 0:00 +1.33 1...Bxa4 2.Nxa4 Qxa4 3.c3 Nc2 4.b3 (16.987) 60
4.00 0:00 +1.21 1...Bxa4 2.Nxa4 Qxa4 3.c3 Nb3 (21.270) 71
5.00 0:00 +1.21 1...Bxa4 2.Nxa4 Qxa4 3.b3 Qa2 4.c3 Qxd2
5.Bxd2 Nb5 6.Ne2 Na7 (43.220) 125
6.00 0:00 +1.17 1...Bxa4 2.Nxa4 Qxa4 3.b3 Qa2 4.c3 Qxd2
5.Bxd2 Nb5 6.Ne2 Na7 (53.600) 142
7.00 0:00 +0.96-- 1...Bxa4 2.Nxa4 Qxa4 3.b3 Qa2 4.c3 Qxd2
5.Bxd2 Nb5 6.Ne2 Na7 7.f5 exf5
8.exf5 gxf5 (94.184) 207
8.00 0:06 +0.60 1...Bxa4 2.f5 exf5 3.gxf5 f6 4.Nxa4 Qxa4
5.b3 Qd7 6.fxg6 hxg6 7.c3 (2.694.764) 447
9.01 0:34 +0.12 1...Bxa4 2.f5 gxf5 3.exf5 Bxc2 4.f6 Bxf6
5.Bxd4 cxd4 6.Qxc2 Be5 7.Qf2 dxc3
8.Qxf7+ Kh8 9.Qxe7 c2 (15.516.353) 449
9.02 0:36 +0.52 1...f5 2.exf5 gxf5 3.g5 Ng6 4.Nh5 Bxa4
5.Nxg7 Kxg7 6.Nxa4 Qxa4 7.b4 Qxc2
8.bxc5 Qxd2 9.Bxd2 Rxc5 10.Be3 (16.400.002) 448
10.01 0:49 +0.40 1...f5 2.exf5 gxf5 3.Nh5 Bxa4 4.Nxa4 Qxa4
5.c3 Nb3 6.Qe1 Rf8 7.Qh4 Ng6 8.Qg5 Qd7 (22.009.201) 448
11.01 1:37 +0.48 1...f5 2.exf5 gxf5 3.Nh5 Bh8 4.Rfe1 Ng6
5.Qf2 Rf8 6.Kg1 Bxa4 7.Nxa4 Qxa4
8.c3 Nc2 9.b3 (44.323.009) 453
12.01 2:41 +0.44 1...f5 2.exf5 gxf5 3.Nh5 Bh8 4.Rfe1 Ng6
5.Qf2 Rf8 6.g5 Rf7 7.Kg1 b6 8.a5 bxa5
9.Ra1 Ne7 10.Rab1 e5 11.Nd5 (73.058.063) 452
13.01 6:17 +0.24-- 1...f5 2.exf5 gxf5 3.Rg1 Bxa4 4.Nxa4 Qxa4
5.c3 Nb3 6.Qe2 Ng6 7.gxf5 Nxf4
8.Bxf4 (169.429.139) 449
14.01 9:34 +0.32 1...f5 2.exf5 gxf5 3.Rg1 fxg4 4.hxg4 Bxa4
5.Nxa4 Qxa4 6.Ne4 Rd8 7.b3 Qd7 8.c3 Nb5
9.Bf2 Nd5 10.Rgc1 (256.531.809) 446
15.01 35:17 -0.07 1...f5 2.exf5 gxf5 3.Nh5 Kh8 4.Qf2 Rg8
5.Rfe1 Nec6 6.g5 Be8 7.Nxg7 Rxg7
8.Bxc6 Bxc6 9.Bxd4 Qxd4 10.Qxd4 cxd4
11.Ne2 (965.693.695) 456
16.01 84:34 0.00 1...f5 2.exf5 gxf5 3.Nh5 Bh8 4.Qf2 Rf8
5.Qg3 Ng6 6.Bxd4 Bxd4 7.gxf5 Rxf5
8.Be4 Rbf8 9.Bxf5 Rxf5 10.Qg4 c4
11.Ne4 Qxa4 12.c3 Qc2+ 13.Qg2 (2.267.159.314) 446
best move: f7-f5 time: 147:55.485 min n/s: 426.020 nodes: 3.781.140.379
This later one is reasonable, the solution is not as fast and nodes per second is a bit low but that is just because of the Neo2 motherboard downclocking the Athlon at that run.
1rr3k1/1p1bnpbp/p2pp1p1/2p5/Pq1nPPP1/2NPB1NP/1PPQ2BK/1R3R2 b - -
Engine: Rainbow Serpent 1.6.3s(dc) Build 263 (Athlon 2009 MHz, 256 MB)
by Tord Romstad, Marco Costalba, Joona Kiiski Modifications: Dann Corbit
1.00 0:00 +1.21 1...Bxa4 (401) 1
2.00 0:00 +1.21 1...Bxa4 2.Nxa4 Qxa4 3.b3 Qe8 (3.951) 15
3.00 0:00 +1.21 1...Bxa4 2.Nxa4 Qxa4 3.c3 Ndc6 (17.119) 60
4.00 0:00 +1.01 1...Bxa4 2.Nxa4 Qxa4 3.c3 Nb3 4.Qc2 Nc6
5.Bf2 Bf6 6.Ne2 (21.886) 73
5.00 0:00 +1.17 1...Bxa4 2.Nxa4 Qxa4 3.c3 Nd5 4.Qc2 (98.078) 209
6.00 0:01 +1.05 1...Bxa4 2.Nxa4 Qxa4 3.c3 Nb3 4.Qc2 Bf6
5.f5 exf5 6.gxf5 (607.275) 353
7.01 0:08 +0.80 1...Bxa4 2.f5 Bxc2 3.f6 Bxb1 4.fxg7 Bc2
5.Qf2 f5 6.exf5 gxf5 7.gxf5 Ndxf5
8.Bd5 exd5 9.Nxf5 (3.539.979) 403
8.01 0:28 +0.43 1...Bxa4 2.f5 Bxc2 3.f6 Bxf6 4.Bxd4 Qxd4
5.Qxc2 Be5 6.Qf2 Qxf2 7.Rxf2 c4
8.dxc4 Rxc4 9.Rd1 (11.735.873) 417
9.01 1:10 +0.28 1...Bxa4 2.f5 exf5 3.exf5 gxf5
4.Nxa4 Qxa4 5.b3 Qd7 6.Bxd4 cxd4
7.Nxf5 Nxf5 8.Rxf5 Qe6 9.Kg1 Qe3+
10.Qxe3 dxe3 11.Bd5 (29.927.907) 423
10.01 1:46 +0.16 1...Bxa4 2.f5 exf5 3.gxf5 Bxc2 4.f6 Bxf6
5.Bxd4 cxd4 6.Qxc2 Bh4 7.Na2 Bxg3+
8.Kxg3 Qb5 9.Qe2 Qe5+ 10.Kf2 (45.570.246) 428
10.03 2:03 +0.28 1...f5 2.exf5 gxf5 3.Nh5 Bh8 4.Rfe1 b6
5.Qd1 Rf8 6.g5 Ng6 7.Bf2 Bxa4 8.Nxa4 Qxa4
9.c3 (52.819.973) 426
11.01 3:24 +0.32 1...f5 2.exf5 gxf5 3.Nh5 Bh8 4.Rfe1 Ng6
5.Qf2 Rf8 6.g5 Bxa4 7.Kh1 Bxc3
8.bxc3 Qxc3 9.Nf6+ Rxf6 10.gxf6 Qxf6 (86.208.736) 421
12.01 4:47 +0.12-- 1...f5 2.exf5 gxf5 3.Nh5 Bh8 4.Rfe1 Ng6
5.Qf2 Rf8 6.g5 Bxa4 7.Nd5 exd5
8.Bxd5+ Rf7 9.Bxf7+ Kxf7 10.c3 (121.320.380) 421
13.01 9:14 +0.28 1...f5 2.exf5 gxf5 3.Nh5 Bh8 4.Rfe1 Rf8
5.Qd1 Ng6 6.Bd2 fxg4 7.hxg4 Nh4
8.Ne4 Qxa4 9.c3 Nxg2 10.Kxg2 Qc2
11.Qxc2 Nxc2 (234.007.336) 422
14.01 41:36 +0.24 1...f5 2.exf5 gxf5 3.Qf2 Be8 4.Rfe1 Bf7
5.g5 Rf8 6.h4 Nec6 7.Qd2 Qa5 8.Nge2 Nxe2
9.Rxe2 Bxc3 10.Qxc3 Qxa4 (1.063.479.787) 426
15.01 88:53 +0.04 1...f5 2.exf5 gxf5 3.Rg1 Kh8 4.Nh5 Ng6
5.Nxg7 Kxg7 6.a5 Bc6 7.Bxc6 bxc6
8.Qf2 Kh8 9.Rge1 Rf8 10.g5 (2.261.774.410) 424
best move: f7-f5 time: 89:09.031 min n/s: 424.052 nodes: 2.268.270.347
This ten hour run is with the latest version, it only has one small change in the search, it is using ok_to_prune not only on the move expected after a nullmove and trying to interfere with that "threat" from the other side if you do nothing (i.e you do a "null move" and Black then gets to execute a "threat move"), but on a real reply in the PV. I hope I got the right move out of the search stack It changes the PV and seemingly the PV gets even more variable than it already was, which was not really an expected effect, but in this particular positon it does not seem to slow down the solution. I can't really judge the quality of the analysis, this is just a first run.
[d]1rr3k1/1p1bnpbp/p2pp1p1/2p5/Pq1nPPP1/2NPB1NP/1PPQ2BK/1R3R2 b - -
Engine: Rainbow Serpent 1.6.3s(dc) Build 269 (Athlon 2009 MHz, 256 MB)
by Tord Romstad, Marco Costalba, Joona Kiiski Modifications: Dann Corbit
1.00 0:00 +1.21 1...Bxa4 (401) 1
2.00 0:00 +1.21 1...Bxa4 2.Nxa4 Qxa4 3.b3 Qe8 (3.951) 15
3.00 0:00 +1.21 1...Bxa4 2.Nxa4 Qxa4 3.c3 Ndc6 (17.146) 60
4.00 0:00 +1.01 1...Bxa4 2.Nxa4 Qxa4 3.c3 Nb3 4.Qc2 Nc6
5.Bf2 Bf6 6.Ne2 (21.909) 73
5.00 0:00 +1.17 1...Bxa4 2.Nxa4 Qxa4 3.c3 Nd5 4.Qc2 (99.982) 213
6.00 0:01 +1.05 1...Bxa4 2.Nxa4 Qxa4 3.c3 Nb3 4.Qc2 Bf6
5.f5 exf5 6.gxf5 (456.599) 365
7.01 0:06 +0.92 1...Bxa4 2.Nxa4 Qxa4 3.c3 Nb3 4.Qc2 d5
5.g5 b6 (2.732.371) 449
8.01 0:12 +0.56-- 1...Bxa4 2.Nxa4 Qxa4 3.c3 Nb3 4.Qc2 d5
5.f5 gxf5 6.gxf5 (5.773.152) 452
9.01 0:17 +0.28 1...Bxa4 2.f5 exf5 3.exf5 gxf5
4.Nxa4 Qxa4 5.b3 Qd7 6.Bxd4 cxd4
7.Nxf5 Nxf5 8.Rxf5 Be5+ 9.Kg1 Kh8 (7.680.268) 451
9.10 0:28 +0.29 1...f5 2.exf5 gxf5 3.Rfe1 Ng6 4.gxf5 Nh4
5.Bxd4 cxd4 6.fxe6 Bxa4 7.Bd5 dxc3
8.e7+ Kh8 9.bxc3 Bxc3 10.Rxb4 Bxd2 (12.989.183) 453
10.01 1:25 +0.24 1...f5 2.exf5 gxf5 3.Rfe1 fxg4
4.hxg4 Ng6 5.Rf1 Nh4 6.Nge4 Bc6
7.Qf2 Nxg2 8.Qxg2 Rf8 (38.382.826) 449
11.01 2:20 +0.20 1...f5 2.exf5 gxf5 3.Rfe1 b6 4.Nh5 Bh8
5.Qd1 Rf8 6.Bd2 Ng6 7.b3 Rbe8 8.Nd5 (63.538.480) 453
12.01 5:08 +0.20 1...f5 2.exf5 gxf5 3.Nh5 Bh8 4.Rfe1 Ng6
5.Qf2 Rf8 6.Nd5 Qxa4 7.Nb6 Qb5 8.Nc4 fxg4
9.hxg4 Nf5 10.gxf5 Rxf5 11.Nxd6 Rxh5+ (139.943.892) 453
13.01 11:45 +0.32 1...f5 2.exf5 gxf5 3.Rfe1 Be8 4.Qd1 Bf7
5.Bd2 Qb6 6.Be3 Qa5 7.Qd2 Nec6
8.gxf5 exf5 9.Bxc6 (321.468.533) 455
14.01 26:58 +0.08-- 1...f5 2.exf5 gxf5 3.Rfe1 Be8 4.Qd1 Bf7
5.Bd2 Qb6 6.a5 Qd8 7.Na4 fxg4 8.hxg4 Nd5
9.c3 Rc7 10.Nb6 Nxb6 11.axb6 Qh4+
12.Bh3 Rc6 13.cxd4 Bxd4 (739.695.389) 457
15.01 47:12 +0.24 1...f5 2.exf5 gxf5 3.Rfe1 Rf8 4.Nh5 Bh8
5.g5 Ng6 6.Qf2 Rf7 7.h4 Rff8 8.Kh1 Bxa4
9.Nxa4 Qxa4 10.Ra1 Qb5 (1.292.475.831) 456
16.01 85:43 +0.12 1...f5 2.exf5 gxf5 3.Nh5 Bh8 4.Qf2 Rf8
5.Rfc1 Rf7 6.g5 Ng6 7.Bd2 Qb6 8.a5 Qd8
9.b4 cxb4 10.Rxb4 (2.349.414.251) 456
17.01 383:01 +0.15 1...f5 2.exf5 gxf5 3.Rg1 Rf8 4.Nh5 Ng6
5.Nxg7 Kxg7 6.Qf2 Bc6 7.Bxc6 bxc6
8.gxf5 Nxf5 9.Ne4 Nxe3 10.Qxe3 Qxa4
11.b3 (10.746.919.146) 467
18.01 630:46 +0.19 1...f5 2.exf5 gxf5 3.Rg1 Rf8 4.Nh5 Ng6
5.Nxg7 Kxg7 6.gxf5 Nxf5 7.Be4 Bxa4
8.Nxa4 Qxa4 9.b3 Qd7 10.d4 cxd4 (17.519.102.328) 462
Eelco
First one is with an older version, with a different search and eval, it is really tough to beat this, so really the challenge now is to see what are the better characteristics of the simpler, older version, and the newer slightly more complex versions.
There are also still bugs I'm afraid but those are harder for me to fix, some areas of the code like the transposition table code etc. I have not really studied and there are some subtle bugs suspected that only show up in debug mode, that the newer version of Stockfish maybe does not have, or were fixed in 1.6.3 which I have not studied yet fully or introduced in Rainbow Serpent. But this as an aside.
1rr3k1/1p1bnpbp/p2pp1p1/2p5/Pq1nPPP1/2NPB1NP/1PPQ2BK/1R3R2 b - -
Engine: Rainbow Serpent 1.6.3s(dc) Build 235 (Athlon 2009 MHz, 256 MB)
by Tord Romstad, Marco Costalba, Joona Kiiski Modifications: Dann Corbit
1.00 0:00 +1.25 1...Bxa4 (482) 1
2.00 0:00 +1.25 1...Bxa4 2.Nxa4 Qxa4 3.b3 Qe8 (3.564) 14
3.00 0:00 +1.33 1...Bxa4 2.Nxa4 Qxa4 3.c3 Nc2 4.b3 (16.987) 60
4.00 0:00 +1.21 1...Bxa4 2.Nxa4 Qxa4 3.c3 Nb3 (21.270) 71
5.00 0:00 +1.21 1...Bxa4 2.Nxa4 Qxa4 3.b3 Qa2 4.c3 Qxd2
5.Bxd2 Nb5 6.Ne2 Na7 (43.220) 125
6.00 0:00 +1.17 1...Bxa4 2.Nxa4 Qxa4 3.b3 Qa2 4.c3 Qxd2
5.Bxd2 Nb5 6.Ne2 Na7 (53.600) 142
7.00 0:00 +0.96-- 1...Bxa4 2.Nxa4 Qxa4 3.b3 Qa2 4.c3 Qxd2
5.Bxd2 Nb5 6.Ne2 Na7 7.f5 exf5
8.exf5 gxf5 (94.184) 207
8.00 0:06 +0.60 1...Bxa4 2.f5 exf5 3.gxf5 f6 4.Nxa4 Qxa4
5.b3 Qd7 6.fxg6 hxg6 7.c3 (2.694.764) 447
9.01 0:34 +0.12 1...Bxa4 2.f5 gxf5 3.exf5 Bxc2 4.f6 Bxf6
5.Bxd4 cxd4 6.Qxc2 Be5 7.Qf2 dxc3
8.Qxf7+ Kh8 9.Qxe7 c2 (15.516.353) 449
9.02 0:36 +0.52 1...f5 2.exf5 gxf5 3.g5 Ng6 4.Nh5 Bxa4
5.Nxg7 Kxg7 6.Nxa4 Qxa4 7.b4 Qxc2
8.bxc5 Qxd2 9.Bxd2 Rxc5 10.Be3 (16.400.002) 448
10.01 0:49 +0.40 1...f5 2.exf5 gxf5 3.Nh5 Bxa4 4.Nxa4 Qxa4
5.c3 Nb3 6.Qe1 Rf8 7.Qh4 Ng6 8.Qg5 Qd7 (22.009.201) 448
11.01 1:37 +0.48 1...f5 2.exf5 gxf5 3.Nh5 Bh8 4.Rfe1 Ng6
5.Qf2 Rf8 6.Kg1 Bxa4 7.Nxa4 Qxa4
8.c3 Nc2 9.b3 (44.323.009) 453
12.01 2:41 +0.44 1...f5 2.exf5 gxf5 3.Nh5 Bh8 4.Rfe1 Ng6
5.Qf2 Rf8 6.g5 Rf7 7.Kg1 b6 8.a5 bxa5
9.Ra1 Ne7 10.Rab1 e5 11.Nd5 (73.058.063) 452
13.01 6:17 +0.24-- 1...f5 2.exf5 gxf5 3.Rg1 Bxa4 4.Nxa4 Qxa4
5.c3 Nb3 6.Qe2 Ng6 7.gxf5 Nxf4
8.Bxf4 (169.429.139) 449
14.01 9:34 +0.32 1...f5 2.exf5 gxf5 3.Rg1 fxg4 4.hxg4 Bxa4
5.Nxa4 Qxa4 6.Ne4 Rd8 7.b3 Qd7 8.c3 Nb5
9.Bf2 Nd5 10.Rgc1 (256.531.809) 446
15.01 35:17 -0.07 1...f5 2.exf5 gxf5 3.Nh5 Kh8 4.Qf2 Rg8
5.Rfe1 Nec6 6.g5 Be8 7.Nxg7 Rxg7
8.Bxc6 Bxc6 9.Bxd4 Qxd4 10.Qxd4 cxd4
11.Ne2 (965.693.695) 456
16.01 84:34 0.00 1...f5 2.exf5 gxf5 3.Nh5 Bh8 4.Qf2 Rf8
5.Qg3 Ng6 6.Bxd4 Bxd4 7.gxf5 Rxf5
8.Be4 Rbf8 9.Bxf5 Rxf5 10.Qg4 c4
11.Ne4 Qxa4 12.c3 Qc2+ 13.Qg2 (2.267.159.314) 446
best move: f7-f5 time: 147:55.485 min n/s: 426.020 nodes: 3.781.140.379
This later one is reasonable, the solution is not as fast and nodes per second is a bit low but that is just because of the Neo2 motherboard downclocking the Athlon at that run.
1rr3k1/1p1bnpbp/p2pp1p1/2p5/Pq1nPPP1/2NPB1NP/1PPQ2BK/1R3R2 b - -
Engine: Rainbow Serpent 1.6.3s(dc) Build 263 (Athlon 2009 MHz, 256 MB)
by Tord Romstad, Marco Costalba, Joona Kiiski Modifications: Dann Corbit
1.00 0:00 +1.21 1...Bxa4 (401) 1
2.00 0:00 +1.21 1...Bxa4 2.Nxa4 Qxa4 3.b3 Qe8 (3.951) 15
3.00 0:00 +1.21 1...Bxa4 2.Nxa4 Qxa4 3.c3 Ndc6 (17.119) 60
4.00 0:00 +1.01 1...Bxa4 2.Nxa4 Qxa4 3.c3 Nb3 4.Qc2 Nc6
5.Bf2 Bf6 6.Ne2 (21.886) 73
5.00 0:00 +1.17 1...Bxa4 2.Nxa4 Qxa4 3.c3 Nd5 4.Qc2 (98.078) 209
6.00 0:01 +1.05 1...Bxa4 2.Nxa4 Qxa4 3.c3 Nb3 4.Qc2 Bf6
5.f5 exf5 6.gxf5 (607.275) 353
7.01 0:08 +0.80 1...Bxa4 2.f5 Bxc2 3.f6 Bxb1 4.fxg7 Bc2
5.Qf2 f5 6.exf5 gxf5 7.gxf5 Ndxf5
8.Bd5 exd5 9.Nxf5 (3.539.979) 403
8.01 0:28 +0.43 1...Bxa4 2.f5 Bxc2 3.f6 Bxf6 4.Bxd4 Qxd4
5.Qxc2 Be5 6.Qf2 Qxf2 7.Rxf2 c4
8.dxc4 Rxc4 9.Rd1 (11.735.873) 417
9.01 1:10 +0.28 1...Bxa4 2.f5 exf5 3.exf5 gxf5
4.Nxa4 Qxa4 5.b3 Qd7 6.Bxd4 cxd4
7.Nxf5 Nxf5 8.Rxf5 Qe6 9.Kg1 Qe3+
10.Qxe3 dxe3 11.Bd5 (29.927.907) 423
10.01 1:46 +0.16 1...Bxa4 2.f5 exf5 3.gxf5 Bxc2 4.f6 Bxf6
5.Bxd4 cxd4 6.Qxc2 Bh4 7.Na2 Bxg3+
8.Kxg3 Qb5 9.Qe2 Qe5+ 10.Kf2 (45.570.246) 428
10.03 2:03 +0.28 1...f5 2.exf5 gxf5 3.Nh5 Bh8 4.Rfe1 b6
5.Qd1 Rf8 6.g5 Ng6 7.Bf2 Bxa4 8.Nxa4 Qxa4
9.c3 (52.819.973) 426
11.01 3:24 +0.32 1...f5 2.exf5 gxf5 3.Nh5 Bh8 4.Rfe1 Ng6
5.Qf2 Rf8 6.g5 Bxa4 7.Kh1 Bxc3
8.bxc3 Qxc3 9.Nf6+ Rxf6 10.gxf6 Qxf6 (86.208.736) 421
12.01 4:47 +0.12-- 1...f5 2.exf5 gxf5 3.Nh5 Bh8 4.Rfe1 Ng6
5.Qf2 Rf8 6.g5 Bxa4 7.Nd5 exd5
8.Bxd5+ Rf7 9.Bxf7+ Kxf7 10.c3 (121.320.380) 421
13.01 9:14 +0.28 1...f5 2.exf5 gxf5 3.Nh5 Bh8 4.Rfe1 Rf8
5.Qd1 Ng6 6.Bd2 fxg4 7.hxg4 Nh4
8.Ne4 Qxa4 9.c3 Nxg2 10.Kxg2 Qc2
11.Qxc2 Nxc2 (234.007.336) 422
14.01 41:36 +0.24 1...f5 2.exf5 gxf5 3.Qf2 Be8 4.Rfe1 Bf7
5.g5 Rf8 6.h4 Nec6 7.Qd2 Qa5 8.Nge2 Nxe2
9.Rxe2 Bxc3 10.Qxc3 Qxa4 (1.063.479.787) 426
15.01 88:53 +0.04 1...f5 2.exf5 gxf5 3.Rg1 Kh8 4.Nh5 Ng6
5.Nxg7 Kxg7 6.a5 Bc6 7.Bxc6 bxc6
8.Qf2 Kh8 9.Rge1 Rf8 10.g5 (2.261.774.410) 424
best move: f7-f5 time: 89:09.031 min n/s: 424.052 nodes: 2.268.270.347
This ten hour run is with the latest version, it only has one small change in the search, it is using ok_to_prune not only on the move expected after a nullmove and trying to interfere with that "threat" from the other side if you do nothing (i.e you do a "null move" and Black then gets to execute a "threat move"), but on a real reply in the PV. I hope I got the right move out of the search stack
It changes the PV and seemingly the PV gets even more variable than it already was, which was not really an expected effect, but in this particular positon it does not seem to slow down the solution. I can't really judge the quality of the analysis, this is just a first run.[d]1rr3k1/1p1bnpbp/p2pp1p1/2p5/Pq1nPPP1/2NPB1NP/1PPQ2BK/1R3R2 b - -
Engine: Rainbow Serpent 1.6.3s(dc) Build 269 (Athlon 2009 MHz, 256 MB)
by Tord Romstad, Marco Costalba, Joona Kiiski Modifications: Dann Corbit
1.00 0:00 +1.21 1...Bxa4 (401) 1
2.00 0:00 +1.21 1...Bxa4 2.Nxa4 Qxa4 3.b3 Qe8 (3.951) 15
3.00 0:00 +1.21 1...Bxa4 2.Nxa4 Qxa4 3.c3 Ndc6 (17.146) 60
4.00 0:00 +1.01 1...Bxa4 2.Nxa4 Qxa4 3.c3 Nb3 4.Qc2 Nc6
5.Bf2 Bf6 6.Ne2 (21.909) 73
5.00 0:00 +1.17 1...Bxa4 2.Nxa4 Qxa4 3.c3 Nd5 4.Qc2 (99.982) 213
6.00 0:01 +1.05 1...Bxa4 2.Nxa4 Qxa4 3.c3 Nb3 4.Qc2 Bf6
5.f5 exf5 6.gxf5 (456.599) 365
7.01 0:06 +0.92 1...Bxa4 2.Nxa4 Qxa4 3.c3 Nb3 4.Qc2 d5
5.g5 b6 (2.732.371) 449
8.01 0:12 +0.56-- 1...Bxa4 2.Nxa4 Qxa4 3.c3 Nb3 4.Qc2 d5
5.f5 gxf5 6.gxf5 (5.773.152) 452
9.01 0:17 +0.28 1...Bxa4 2.f5 exf5 3.exf5 gxf5
4.Nxa4 Qxa4 5.b3 Qd7 6.Bxd4 cxd4
7.Nxf5 Nxf5 8.Rxf5 Be5+ 9.Kg1 Kh8 (7.680.268) 451
9.10 0:28 +0.29 1...f5 2.exf5 gxf5 3.Rfe1 Ng6 4.gxf5 Nh4
5.Bxd4 cxd4 6.fxe6 Bxa4 7.Bd5 dxc3
8.e7+ Kh8 9.bxc3 Bxc3 10.Rxb4 Bxd2 (12.989.183) 453
10.01 1:25 +0.24 1...f5 2.exf5 gxf5 3.Rfe1 fxg4
4.hxg4 Ng6 5.Rf1 Nh4 6.Nge4 Bc6
7.Qf2 Nxg2 8.Qxg2 Rf8 (38.382.826) 449
11.01 2:20 +0.20 1...f5 2.exf5 gxf5 3.Rfe1 b6 4.Nh5 Bh8
5.Qd1 Rf8 6.Bd2 Ng6 7.b3 Rbe8 8.Nd5 (63.538.480) 453
12.01 5:08 +0.20 1...f5 2.exf5 gxf5 3.Nh5 Bh8 4.Rfe1 Ng6
5.Qf2 Rf8 6.Nd5 Qxa4 7.Nb6 Qb5 8.Nc4 fxg4
9.hxg4 Nf5 10.gxf5 Rxf5 11.Nxd6 Rxh5+ (139.943.892) 453
13.01 11:45 +0.32 1...f5 2.exf5 gxf5 3.Rfe1 Be8 4.Qd1 Bf7
5.Bd2 Qb6 6.Be3 Qa5 7.Qd2 Nec6
8.gxf5 exf5 9.Bxc6 (321.468.533) 455
14.01 26:58 +0.08-- 1...f5 2.exf5 gxf5 3.Rfe1 Be8 4.Qd1 Bf7
5.Bd2 Qb6 6.a5 Qd8 7.Na4 fxg4 8.hxg4 Nd5
9.c3 Rc7 10.Nb6 Nxb6 11.axb6 Qh4+
12.Bh3 Rc6 13.cxd4 Bxd4 (739.695.389) 457
15.01 47:12 +0.24 1...f5 2.exf5 gxf5 3.Rfe1 Rf8 4.Nh5 Bh8
5.g5 Ng6 6.Qf2 Rf7 7.h4 Rff8 8.Kh1 Bxa4
9.Nxa4 Qxa4 10.Ra1 Qb5 (1.292.475.831) 456
16.01 85:43 +0.12 1...f5 2.exf5 gxf5 3.Nh5 Bh8 4.Qf2 Rf8
5.Rfc1 Rf7 6.g5 Ng6 7.Bd2 Qb6 8.a5 Qd8
9.b4 cxb4 10.Rxb4 (2.349.414.251) 456
17.01 383:01 +0.15 1...f5 2.exf5 gxf5 3.Rg1 Rf8 4.Nh5 Ng6
5.Nxg7 Kxg7 6.Qf2 Bc6 7.Bxc6 bxc6
8.gxf5 Nxf5 9.Ne4 Nxe3 10.Qxe3 Qxa4
11.b3 (10.746.919.146) 467
18.01 630:46 +0.19 1...f5 2.exf5 gxf5 3.Rg1 Rf8 4.Nh5 Ng6
5.Nxg7 Kxg7 6.gxf5 Nxf5 7.Be4 Bxa4
8.Nxa4 Qxa4 9.b3 Qd7 10.d4 cxd4 (17.519.102.328) 462
Eelco
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
place. Therefore, if you write the code as cleverly as possible, you
are, by definition, not smart enough to debug it.
-- Brian W. Kernighan