TCEC Season 7 - Main Thread

Discussion of computer chess matches and engine tournaments.

Moderators: bob, hgm, Harvey Williamson

Forum rules
This textbox is used to restore diagrams posted with the [d] tag before the upgrade.
User avatar
M ANSARI
Posts: 3411
Joined: Thu Mar 16, 2006 6:10 pm

Re: TCEC Season 7 - Main Thread

Post by M ANSARI » Sat Dec 06, 2014 4:53 pm

I don't know about lost, but it certainly a good position for white. SF chooses 3 moves and thinks that black's best defense is to exchange white's black squared bishop. There must be a ton of variations where that bishop really hurts and so it thinks it must go. I did a clip analysis with 3 variations on and at depth 37 this is what SF thinks

54: New game

[d]r1bq1rk1/pp2bpp1/2p1p2p/3n4/3P2P1/P2B1N2/1PP1QP1P/R1B1K2R b KQ - 0 1

Analysis by Stockfish 151014 64 SSE4.2:

1. +/= (0.62): 7...Bd6 8.Bd2 Bf4 9.0-0-0 Bxd2+ 10.Qxd2 Qf6 11.Ne5 Qf4 12.h4 Qxd2+ 13.Kxd2 Nf4 14.g5 Nxd3 15.Nxd3 b6 16.gxh6 gxh6 17.Ne5 Bb7 18.Rhg1+ Kh7 19.Nd7 Rfd8 20.Nf6+ Kh8 21.Ke3 Ba6 22.Ng4 Kh7 23.c3 Bb7 24.Nf6+ Kh8
2. +/- (1.00): 7...c5 8.Rg1 cxd4 9.g5 h5 10.g6 Qd6 11.gxf7+ Rxf7 12.Rg3 e5 13.Nxe5 Rf5 14.Ng6 Bd7 15.Bxf5 Bxf5 16.Qxh5 Qe6+ 17.Kf1 Bd6 18.f4 Re8 19.Bd2 Qe4 20.b4 Nc3 21.Rg2 Qxc2 22.Kg1 Bxg6 23.Qxg6 Qxg6 24.Rxg6 Bc7 25.Rf1 Kf7 26.Rg2 Ne2+ 27.Kh1 Re4 28.f5 Be5 29.Rf3 a6 30.a4
3. +/- (1.01): 7...e5 8.dxe5 Qa5+ 9.Bd2 Qa4 10.c4 Nb4 11.Bxb4 Bxb4+ 12.Kf1 g6 13.Qe4 Be7 14.h4 Be6 15.h5 Rad8 16.b4 Qb3 17.Be2 g5 18.Nd4 Rxd4 19.Qxd4 Rd8 20.Qe3 Bxc4 21.Qxb3 Bxb3 22.Rh3 Be6 23.Ke1 Rd5 24.Re3 Rd4 25.Rd1 Rf4 26.f3 c5 27.bxc5

(Spearfish, 06.12.2014)

IanO
Posts: 478
Joined: Wed Mar 08, 2006 8:45 pm
Location: Portland, OR
Contact:

Re: TCEC Season 7 - Main Thread

Post by IanO » Tue Dec 09, 2014 2:57 pm

At the halfway mark, Komodo is leading the field, thanks to four early wins against Gull.

Code: Select all

 N Engine           Rtng  Pts  Gm    SB Ko     St     Ho     Gu    

 1 Komodo 1331      3208 11.0  18 89.75 ······ ==0=== ==1=== 1111==
 2 Stockfish 301114 3222  9.0  18 82.25 ==1=== ······ ==0=== =10=01
 3 Houdini 4        3195  8.5  18 77.75 ==0=== ==1=== ······ 0==0=1
 4 Gull 3           3109  7.5  18 67.75 0000== =01=10 1==1=0 ······

JJJ
Posts: 1287
Joined: Sat Apr 19, 2014 11:47 am

Re: TCEC Season 7 - Main Thread

Post by JJJ » Wed Dec 10, 2014 4:41 pm

What is the probability of a final between houdini and gull ? :p

User avatar
Laskos
Posts: 9509
Joined: Wed Jul 26, 2006 8:21 pm
Full name: Kai Laskos

Re: TCEC Season 7 - Main Thread

Post by Laskos » Wed Dec 10, 2014 4:45 pm

JJJ wrote:What is the probability of a final between houdini and gull ? :p
Probably around 0.2 times 0.1, or 2%. I didn t do simulations this time, just a gut feeling about the order of magnitude.

JJJ
Posts: 1287
Joined: Sat Apr 19, 2014 11:47 am

Re: TCEC Season 7 - Main Thread

Post by JJJ » Wed Dec 10, 2014 8:32 pm

Thx you for your opinion.

whereagles
Posts: 561
Joined: Thu Nov 13, 2014 11:03 am

Re: TCEC Season 7 - Main Thread

Post by whereagles » Thu Dec 11, 2014 12:48 am

anyone got a clue why SF is having problems this season, despite +40 elo vs last season? :?:

Jouni
Posts: 2029
Joined: Wed Mar 08, 2006 7:15 pm

Re: TCEC Season 7 - Main Thread

Post by Jouni » Thu Dec 11, 2014 8:01 am

Framework testing was 60s+0.05 th 1, but TCEC is 120m+30s th 16. Something like 1000x more nodes/move :!:.
Jouni

JJJ
Posts: 1287
Joined: Sat Apr 19, 2014 11:47 am

Re: TCEC Season 7 - Main Thread

Post by JJJ » Thu Dec 11, 2014 8:12 am

I hope for the final, the true best version of it will play.

User avatar
Laskos
Posts: 9509
Joined: Wed Jul 26, 2006 8:21 pm
Full name: Kai Laskos

Re: TCEC Season 7 - Main Thread

Post by Laskos » Thu Dec 11, 2014 11:44 am

whereagles wrote:anyone got a clue why SF is having problems this season, despite +40 elo vs last season? :?:
Things happen, in 10-20-30 games. Even measuring +40 Elo points with 95% confidence needs more than 100 games.

Martin Thoresen
Posts: 1833
Joined: Wed Jun 21, 2006 10:07 pm

Re: TCEC Season 7 - Main Thread

Post by Martin Thoresen » Tue Dec 16, 2014 3:49 pm

Congratulations to Komodo and Stockfish for qualifying for the Superfinal.
I think many people will say Stage 4 was very exciting - it was very close and not clear until the end who would qualify for the Superfinal.

The Superfinal is now live and can be followed on the live page:
http://tcec.chessdom.com/live.php

The Superfinal is a 64 game match where they will play black & white of the same opening. The openings are again assembled by Nelson Hernandez & Adam Hair and filtered by IM Erik Kislik.
You can find some notes from Erik in the "Opening Book" link at the top of the TCEC website.

The rating list has also been updated after Stage 4, you can find it here in the archive page:
http://tcec.chessdom.com/archive.php

Updated engines for the Superfinal:
Komodo 1333
Stockfish 141214

Enjoy!

Post Reply