TCEC openings

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

Moderators: hgm, Rebel, chrisw

jdart
Posts: 4366
Joined: Fri Mar 10, 2006 5:23 am
Location: http://www.arasanchess.org

TCEC openings

Post by jdart »

I am surprised by some of the opening choices used in the TCEC match, and also am not surprised but disappointed in how badly the engines play the opening once out of book.

Once example is below. 9. h3 is the last book move. At this point Chessbase Megabase shows a winning percentage for White of 75%, so this is not a balanced position. I think Gull's first non-book move 9. .. Bf5 is questionable even though this is the most commonly played and seems to be favored recently (Hiarcs book recommends .. h6, preventing Bg5, which looks better). White continued to build an advantage through the whole game.

I don't use these games in building my opening book because of issues like this.

--Jon

[pgn]
[Event "TCEC Season 9 - Stage 3"]
[Site "http://tcec.chessdom.com"]
[Date "2016.08.22"]
[Round "36"]
[White "Fire 5"]
[Black "Gull 3"]
[Result "1-0"]
[ECO "D20"]
[WhiteElo "3099"]
[BlackElo "3123"]
[PlyCount "134"]
[EventDate "2016.07.23"]

1. d4 d5 2. c4 dxc4 3. e3 e5 4. Bxc4 exd4 5. exd4 Nf6 6. Nf3 Bb4+ 7. Nc3 O-O 8.
O-O Nc6 9. h3 Bf5 10. Bg5 Qd6 11. Nb5 Qd7 12. a3 a6 13. Bxf6 axb5 14. Bxb5 gxf6
15. axb4 Rxa1 16. Qxa1 Be4 17. Qa3 Qf5 18. Bxc6 Bxc6 19. Rc1 Re8 20. Re1 Re6
21. Rxe6 fxe6 22. Ne1 Qe4 23. Qc3 Kf7 24. b5 Bxb5 25. Qxc7+ Ke8 26. Qa5 Bc6 27.
Qd2 Qd5 28. Qe3 Bb5 29. Qf4 Ke7 30. Nc2 Bc6 31. Ne3 Qe4 32. Qxe4 Bxe4 33. f3
Bd3 34. g4 b6 35. Kf2 Kd6 36. f4 Bg6 37. Kg3 Bf7 38. h4 Ke7 39. g5 f5 40. d5
Kd6 41. dxe6 Kxe6 42. Kf3 Bh5+ 43. Kf2 Be8 44. Ke2 Bb5+ 45. Kd2 Bd7 46. Kd3 Be8
47. Kd4 Bh5 48. Nd5 Be2 49. Nxb6 Kf7 50. Nd5 Bd1 51. Ke5 Be2 52. Nf6 Kg7 53.
Nxh7 Kxh7 54. Kxf5 Ba6 55. g6+ Kh6 56. g7 Kxg7 57. Kg5 Bb5 58. b3 Bc6 59. b4
Bb5 60. f5 Bc4 61. h5 Kh8 62. Kf6 Bd3 63. Ke6 Bc2 64. b5 Kg8 65. b6 Bb3+ 66.
Ke5 Kg7 67. b7 Bf7 1-0

[/pgn]
Krzysztof Grzelak
Posts: 1525
Joined: Tue Jul 15, 2014 12:47 pm

Re: TCEC openings

Post by Krzysztof Grzelak »

I have always believed that the book debuts should have a maximum of 4 movements, or do not use any books. There will always be speculation as to the debut book no matter who did it.
Sven
Posts: 4052
Joined: Thu May 15, 2008 9:57 pm
Location: Berlin, Germany
Full name: Sven Schüle

Re: TCEC openings

Post by Sven »

jdart wrote:At this point Chessbase Megabase shows a winning percentage for White of 75%, so this is not a balanced position.
How many games between strong players (say, ELO >= 2400) are included in that 75% percentage?
Dirt
Posts: 2851
Joined: Wed Mar 08, 2006 10:01 pm
Location: Irvine, CA, USA

Re: TCEC openings

Post by Dirt »

Give these monster balanced openings and they will draw almost every game. The tournament organizers are going for openings they hope will give an advantage, but not a sure win. I think they are doing a very good job.
Deasil is the right way to go.
Adam Hair
Posts: 3226
Joined: Wed May 06, 2009 10:31 pm
Location: Fuquay-Varina, North Carolina

Re: TCEC openings

Post by Adam Hair »

Dirt wrote:Give these monster balanced openings and they will draw almost every game. The tournament organizers are going for openings they hope will give an advantage, but not a sure win. I think they are doing a very good job.
Nelson will be pleased to see that there is some approval for his selected openings.
User avatar
cdani
Posts: 2204
Joined: Sat Jan 18, 2014 10:24 am
Location: Andorra

Re: TCEC openings

Post by cdani »

Dirt wrote:Give these monster balanced openings and they will draw almost every game. The tournament organizers are going for openings they hope will give an advantage, but not a sure win. I think they are doing a very good job.
+1
corres
Posts: 3657
Joined: Wed Nov 18, 2015 11:41 am
Location: hungary

Re: TCEC openings

Post by corres »

Starting from the same opening position the chances are different for a mainly tactical monster like Stockfish and for a mainly positional monster like Komodo. A tactical engine finds a good continuation easier in an unbalanced position than in a balanced position and vice verse a positional engine does. So in my opinion a really good arrange is 50% balanced and 50% unbalanced opening position played with reversed color.
User avatar
hgm
Posts: 27790
Joined: Fri Mar 10, 2006 10:06 am
Location: Amsterdam
Full name: H G Muller

Re: TCEC openings

Post by hgm »

jdart wrote:Once example is below. 9. h3 is the last book move. At this point Chessbase Megabase shows a winning percentage for White of 75%, so this is not a balanced position.
TCEC intentionally uses unbalanced openings, to avoid bloodless draws. 75% is just the sort of imbalance you need: roughly a 50-50 mix of wins and draws. The slightest difference in quality of play will immediately express itself as half a point.
Hai
Posts: 598
Joined: Sun Aug 04, 2013 1:19 pm

Re: TCEC openings

Post by Hai »

hgm wrote:
jdart wrote:Once example is below. 9. h3 is the last book move. At this point Chessbase Megabase shows a winning percentage for White of 75%, so this is not a balanced position.
TCEC intentionally uses unbalanced openings, to avoid bloodless draws. 75% is just the sort of imbalance you need: roughly a 50-50 mix of wins and draws. The slightest difference in quality of play will immediately express itself as half a point.
= Take the 20 possible opening moves:
1. Crazy games
2. Fair play
3. Opening phase too
4. Draw rate 50%
5. IDeA says all moves are 0.00.
User avatar
hgm
Posts: 27790
Joined: Fri Mar 10, 2006 10:06 am
Location: Amsterdam
Full name: H G Muller

Re: TCEC openings

Post by hgm »

Well, 0.00 is exactly what you don't want. It would mean that not all the wins are for the same side.