Discussion of chess software programming and technical issues.
Moderators: hgm , Dann Corbit , Harvey Williamson
tttony
Posts: 264Joined: Sun Apr 24, 2011 12:33 am
Post by tttony Sat Oct 10, 2015 4:04 am
Just print till depth 10
[D] r1b1kb1r/pp1p1ppp/1qn1pn2/2p5/4P3/2N2P2/PPPPN1PP/R1BQ1BKR w kq - 6 5
Avoid Na4
My engine was playing fast games 5s+100ms and one of them lost this game playing Na4, so a mate in 4
Code: Select all
Skiull x64:
1 00:00 50 50 +0,35 d2-d3
2 00:00 578 578 +0,10 Nc3-a4 Qb6-a6
3 00:00 1.123 1.123 +0,20 Nc3-a4 Qb6-a6 Na4-c3
4 00:00 3.014 3.014 +0,15 Nc3-a4 Qb6-a6 Ne2-c3 Qa6-a5
5 00:00 11.849 740.563 +0,10 Nc3-a4 Qb6-a5 Ne2-c3 d7-d5 Bf1-d3
6 00:00 50.440 1.576.250 0,00 d2-d3 d7-d5 Ra1-b1 d5xe4 f3xe4 Bf8-d6
7 00:00 103.052 2.192.596 0,00 d2-d3 d7-d5 Ra1-b1 Bf8-d6 Bc1-g5 d5-d4 Bg5xf6
8 00:00 330.738 2.645.904 +0,05 Nc3-a4 Qb6-a5 c2-c4 d7-d5 d2-d3 Bf8-d6 Bc1-g5 Nc6-b4
9 00:00 417.553 2.676.622 +0,05 Nc3-a4 Qb6-a5 c2-c4 d7-d5 d2-d3 Bf8-e7 c4xd5 e6xd5 Bc1-e3
10 00:00 1.714.536 2.815.330 -0,05 d2-d3 Bf8-e7 Nc3-a4 Qb6-a6 Na4-c3 d7-d5 d3-d4 O-O Bc1-e3 Rf8-d8
sje
Posts: 4675Joined: Mon Mar 13, 2006 7:43 pm
Post by sje Sat Oct 10, 2015 5:14 am
Code: Select all
PV: [+0.064/1/0.002/6] 5. Nf4
PV: [LoseIn3/2/0.004/110] 5. Nf4 c4+ 6. d4 cxd3+ 7. Be3 Qxe3#
PV: [-0.224/2/0.004/178] 5. d3 Bd6
PV: [-0.283/3/0.005/777] 5. d3 c4+ 6. d4 Bd6
PV: [-0.170/3/0.005/1,237] 5. Na4 Qb5 6. Nec3
PV: [LoseIn5/4/0.007/3,564] 5. Na4 c4+ 6. Nxb6 Bc5+ 7. d4 cxd3+ 8. Nd4 Bxd4+ 9. Be3 Bxe3#
PV: [-0.231/4/0.015/12,537] 5. d3 d5 6. g3 Bd6
PV: [-0.289/5/0.052/47,155] 5. d3 c4+ 6. d4 d5 7. exd5
PV: [-0.190/6/0.220/240,838] 5. d3 d5 6. exd5 Nxd5 7. Ne4 Be7
PV: [-0.195/7/0.801/939,965] 5. d3 d5 6. Na4 Qa5 7. c4 Bd6 8. Bf4
PV: [-0.228/8/2.867/3,820,639] 5. d3 d5 6. b3 c4+ 7. d4 Bb4 8. exd5 Nxd5 9. Ne4
PV: [-0.168/9/13.512/18,094,881] 5. d3 d5 6. Na4 Qa6 7. b3 b5 8. exd5 exd5 9. Nac3
PV: [-0.255/10/46.411/68,179,323] 5. d3 d5 6. Na4 Qa6 7. exd5 exd5 8. Nac3 d4 9. Ne4 Be7
Joost Buijs
Posts: 1562Joined: Thu Jul 16, 2009 10:47 amLocation: Almere, The Netherlands
Post by Joost Buijs Sat Oct 10, 2015 7:10 am
Code: Select all
FEN: r1b1kb1r/pp1p1ppp/1qn1pn2/2p5/4P3/2N2P2/PPPPN1PP/R1BQ1BKR w kq - 6 5
Nightmare MP:
1 00:00 24 24 -0.23 Ne2-f4
2 00:00 139 139 -0.23 Ne2-f4
3 00:00 1k 1k -1.48 Nc3-a4 Qb6-a5
4 00:00 11k 2,147k -1.35 d2-d3 Bf8-d6 Bc1-f4 Bd6xf4
5 00:00 19k 1,867k -1.10 g2-g3 Bf8-d6 Kg1-g2 Nc6-e5
6 00:00 26k 2,006k -0.99 g2-g3 Bf8-d6 Kg1-g2 O-O Ne2-f4 Bd6xf4
7 00:00 34k 2,248k -0.99 g2-g3 Bf8-d6 Kg1-g2 O-O Ne2-f4 Bd6xf4 g3xf4
8 00:00 67k 2,700k -1.25 g2-g3 Bf8-d6 Kg1-g2 O-O d2-d3 c5-c4 Bc1-g5 c4xd3
9 00:00 131k 3,272k -1.16 g2-g3 Bf8-d6 Kg1-g2 h7-h5 b2-b3 h5-h4 Ne2-f4 Bd6xf4 g3xf4
10 00:00 266k 4,434k -1.07 g2-g3 Bf8-d6 Kg1-g2 O-O b2-b3 c5-c4 Nc3-a4 Qb6-a5 Bc1-b2
Nightmare does not like this position for white, thinks about Na4 once and sees it is bad.
It finds that it has to make an escape route for the king with g3.
Henk
Posts: 7210Joined: Mon May 27, 2013 10:31 am
Post by Henk Sat Oct 10, 2015 9:53 am
Code: Select all
Depth Value Time(seconds) Nodes
10 -0.71 15.01 563218 d2d3
9 -0.70 10.59 406460 d2d3
8 -0.70 3.56 131907 d2d3
7 -0.70 7.74 306200 g1f2
6 -0.70 0.91 28277 d2d3
5 -0.71 2.23 78773 g1f2
4 -0.43 0.37 9947 c3a4
3 -0.45 1.12 32599 c3a4
2 -0.41 0.06 1545 e2g3
1 -0.40 0.03 266 e2f4
Maarten Claessens
Posts: 106Joined: Mon May 12, 2014 10:08 amLocation: Near Nijmegen
Post by Maarten Claessens Sat Oct 10, 2015 10:16 am
Code: Select all
WaDuuttie> sb r1b1kb1r/pp1p1ppp/1qn1pn2/2p5/4P3/2N2P2/PPPPN1PP/R1BQ1BKR w kq - 6 5
WaDuuttie> st 60
WaDuuttie> sd 10
WaDuuttie> on
ply nodes score pv
1 30 -34 d2d3
2 289 -72 d2d3 d7d5
3 980 -60 c3a4 b6a5 e2c3
4 2849 -64 d2d3 c5c4 d3d4 d7d5 c3a4
5 4521 -88 d2d3 c5c4 d3d4 d7d5 c1g5 b6b2
6 15892 -76 d2d3 c5c4 d3d4 d7d5 h2h4 f8d6 c1g5
7 26772 -78 d2d3 c5c4 d3d4 d7d5 e4e5 f6d7 c3a4 b6b5
8 65338 -90 d2d3?
8 98046 -94 d2d3 d7d5 h2h4 f8e7 g2g4 c6e5 f1g2 c5c4 g1h2
9 132689 -78 d2d3!
9 183575 -76 d2d3 c5c4 d3d4 d7d5 e4e5 f6d7 c3a4 b6a5 c2c3 f8e7
10 284085 -76 d2d3 d7d5 c3a4 b6a6 c2c4 b7b6 c4d5 e6d5 c1g5 f8d6
g5f6
Computer's move: d2d3 (359 msecs)
Nothing is unstable (Lawrence Krauss)
matthewlai
Posts: 793Joined: Sun Aug 03, 2014 4:48 amLocation: London, UK
Post by matthewlai Sat Oct 10, 2015 11:18 am
Giraffe has no concept of depth. The first number is just an iteration count for the probability-based search. This is a ~6 seconds search.
Code: Select all
1 1 0 26 e2f4
2 1 0 52 e2f4
3 -15 0 121 d2d3
4 -10 0 269 d2d3 a7a5
5 -21 0 649 a2a3 d7d5 e4d5 e6d5
6 -25 0 1324 d2d3 a7a5 a2a3
7 -22 1 1851 d2d3 a7a5 a2a3 d7d5 e4d5 e6d5
8 -31 2 4477 d2d3 a7a5 a2a3 a8a6 e2f4
9 -37 5 8674 d2d3 a7a5 a2a3 a5a4 f3f4 d7d5 e4d5 f6d5
10 -35 12 19235 d2d3 a7a5 f3f4 d7d5 e4e5 f6d7 a2a3
11 -41 26 42014 d2d3 a7a5 a2a3 a5a4 h2h3 d7d6 a1b1
12 -42 60 96798 d2d3 a7a5 a2a3 a5a4 a1b1 d7d5 e4d5 e6d5
13 -32 129 204178 d2d3 a7a5 c3a4 b6d8 c2c4 d7d6 a4c3 h7h5 c1g5
14 -44 254 414328 d2d3 d7d5 e4d5 e6d5 d1e1 c5c4 d3d4 c8f5 e1d1 e8c8
15 -69 627 1031260 g2g4 h7h6 d2d3 d7d5 e4d5 e6d5 f1g2 d5d4
Disclosure: I work for DeepMind on the AlphaZero project, but everything I say here is personal opinion and does not reflect the views of DeepMind / Alphabet.
Daniel Anulliero
Posts: 759Joined: Fri Jan 04, 2013 4:55 pmLocation: Nice
Post by Daniel Anulliero Sat Oct 10, 2015 1:40 pm
"last best version " of Isa (1.6.0.98) :
Code: Select all
1 74 108 0 e2f4
2 22 236 0 e2f4 c5c4
3 42 1967 0 c3a4 b6b5 a4c3
4 -2 5062 0 c3a4 b6b4 a4c3 d7d5
5 0 8174 0 c3a4 b6b4 a4c3 b4b6
6 0 28373 1 c3a4 b6c7 a4c3 c7b6
7 0 116150 4 c3a4 b6b5 a4c3 b5b6
8 -4 1138789 35 g2g4 f8e7 g4g5 f6h5 d2d3 e8g8 f1g2 c5c4
9 2 1249005 71 g2g4 h7h6 g1g2 e6e5 c3d5 b6d8 e2c3 h6h5 d5f6 d8f6
10 -16 3399651 165 g2g4 h7h6 g1g2 f8e7 e2g3 e8g8 f1c4 b6c7 c3b5 c7f4
It like g4 , like Giraffe
flok
Post by flok Sat Oct 10, 2015 2:42 pm
revision 1200
Code: Select all
info string move: e2-f4 (1), alpha/score/beta: -32767/-4/32767, took 0.00/0.00, 0.03knps
info string move: c3-a4 (2), alpha/score/beta: -79/-56/71, took 0.01/0.02, 0.03knps
info string move: d2-d3 (3), alpha/score/beta: -131/-30/19, took 0.07/0.08, 0.05knps
info string move: d1-e1 (4), alpha/score/beta: -105/-57/45, took 0.48/0.56, 0.07knps
info string move: c3-a4 (5), alpha/score/beta: -132/-37/18, took 1.48/2.04, 0.05knps
info string move: c3-a4 (6), alpha/score/beta: -112/-52/38, took 9.53/11.57, 0.08knps
info string move: c3-a4 (7), alpha/score/beta: -127/-34/23, took 63.87/75.43, 0.07knps
jdart
Posts: 4361Joined: Fri Mar 10, 2006 5:23 amLocation: http://www.arasanchess.org
Post by jdart Sat Oct 10, 2015 4:01 pm
Arasan 18.0, 1 thread:
Code: Select all
1 0.00 Nd4 -3.85 17
1 0.00 Na4 -0.85 102
1 0.00 Qe1 -0.66 323
1 0.01 d3 -0.46 595
1 0.01 d3 -0.46 1061
2 0.01 d3 -0.73 1635
3 0.01 d3 -0.61 3123
4 0.01 -- -0.80 3395
4 0.01 d3 -0.83 3844
5 0.01 d3 -0.68 4526
6 0.02 Na4 -0.66 6974
6 0.02 Na4 -0.66 7140
7 0.02 Na4 -0.69 10990
8 0.03 d3 -0.53 21581
8 0.03 d3 -0.53 22118
9 0.04 -- -0.72 32882
9 0.05 d3 -0.60 36689
10 0.06 d3 -0.64 45458
mvk
Posts: 589Joined: Tue Jun 04, 2013 10:15 pm
Post by mvk Sat Oct 10, 2015 4:04 pm
[D] r1b1kb1r/pp1p1ppp/1qn1pn2/2p5/4P3/2N2P2/PPPPN1PP/R1BQ1BKR w kq -
Code: Select all
Rookie v3.9
time ply score variation
0.0 1 -0.162 5. h4
0.0 2 -0.512 5. h4 c4+ 6. Kh2
0.0 3 -0.719 5. h4 c4+ 6. Kh2 Bd6+ 7. f4
0.0 4 -0.601 5. d3 c4+ 6. d4 Bc5 7. h4 Nxd4
0.0 5 -0.688 5. Kf2 c4+ 6. Kg3 Bd6+ 7. f4 Nb4 8. e5
0.0 6 -0.687 5. Kf2 c4+ 6. Kg3 Bd6+ 7. f4 Bc5 8. d4 Nxd4
0.2 7 -0.381 5. d3 Bd6 6. h4 Nd4 7. Qd2 Be5 8. Nf4
0.3 8 -0.378 5. d3 Bd6 6. Be3 Nd4 7. Kf2 Qxb2 8. Nxd4 Qxc3
0.6 9 -0.511 5. d3 d5 6. exd5 exd5 7. h4 Bd6 8. Qd2 O-O 9. Nf4
0.7 10 -0.559 5. d3 d5 6. exd5 exd5 7. h4 Bd6 8. Bg5 O-O 9. Bxf6 gxf6
{Depth limit reached}[Account deleted]
