Engine Analysis

AdminX · Post by **AdminX** » Wed Apr 02, 2008 10:30 pm

I played a game of chess (15,3 Rated) online at playchess this past weekend. My opponent moved so fast I thought he might have been a cheater. Anyway, I prefer to give my opponents the benefit of any doubts I may have had and have learned to keep the blame and the focus on myself when it comes to my poor play.

After the game I let the chess engine Deep Shredder 11 analyze my game. I was surprised to find that up to move 19, the whole line had been book, since my knowledge ended after white's 11th move. I bring this up because I normally let my engine use a reference database (Megabase 2008). The problem is the program stops all analysis at the move right before the Novelty. Long story, short I see now that the whole line I went down was bad. However the engine did not point that out. Which would have otherwise lead me to believe my novelty was bad (In this case it was, but by that time I was already in time trouble). Now I wonder is it best to kill the reference database when doing engine analysis?

[Event "Rated game, 15m + 3s"]
[Site "Café"]
[Date "2008.03.30"]
[Round "?"]
[White "Rishmawi"]
[Black "NYTed"]
[Result "1-0"]
[ECO "B98"]
[WhiteElo "1937"]
[BlackElo "1951"]
[Annotator "Deep Shredder 11 UCI (30s)"]
[PlyCount "45"]
[EventDate "2008.03.30"]
[TimeControl "900+3"]

{B98: Sicilian Najdorf: 6 Bg5 e6 7 f4 Be7, sidelines} 1. e4 {[%emt 0:00:03]} c5
{[%emt 0:00:02]} 2. Nf3 {[%emt 0:00:02]} d6 {[%emt 0:00:01]} 3. d4 {
[%emt 0:00:04]} cxd4 {[%emt 0:00:01]} 4. Nxd4 {[%emt 0:00:19]} Nf6 {
[%emt 0:00:01]} 5. Nc3 {[%emt 0:00:03]} a6 {[%emt 0:00:04]} 6. Bg5 {
[%emt 0:00:26]} e6 {[%emt 0:00:04]} 7. f4 {[%emt 0:00:16]} Be7 {[%emt 0:00:03]}
8. Qf3 {[%emt 0:00:07]} Qc7 {[%emt 0:00:07]} 9. O-O-O {[%emt 0:00:06]} Nc6 {
[%emt 0:00:03]} 10. Nxc6 {[%emt 0:00:06]} Qxc6 {[%emt 0:00:23]} 11. Qg3 {
[%emt 0:00:09]} h6 {[%emt 0:01:05]} 12. Bxf6 {[%emt 0:00:24]} Bxf6 {
[%emt 0:00:04]} 13. e5 {[%emt 0:00:11]} dxe5 {[%emt 0:00:06]} 14. fxe5 {
[%emt 0:00:05]} Bg5+ {[%emt 0:00:05]} 15. Kb1 {[%emt 0:00:13]} O-O {
[%emt 0:00:10]} 16. Bd3 {[%emt 0:00:15]} Rb8 {[%emt 0:01:19]} 17. Ne4 {
[%emt 0:00:20]} Be7 {[%emt 0:01:14]} 18. Nf6+ {[%emt 0:00:06]} Kh8 {
[%emt 0:00:02]} 19. Rhf1 {[%emt 0:00:25]} Qc7 $146 {[%emt 0:00:33]} (19... Qb6
20. Ne8 Rg8 21. Be4 Rxe8 22. Rxf7 Bg5 23. Qf3 Kg8 24. Bg6 Qe3 25. Rxg7+ {
1-0 Peregi,Z (2079)-Pais,N (2153)/Hungary 2007/EXT 2008}) (19... b6 {
is still a small chance} 20. Nh5 g5 $18) 20. Ne8 $3 {Mate threat [%emt 0:00:55]
} Rxe8 {[%emt 0:00:21]} (20... Rxe8 21. Rxf7 {Combination}) (20... -- 21. Qxg7#
{Mate threat}) 21. Rxf7 {[%emt 0:00:11]} g5 {[%emt 0:01:19]} 22. Qf3 {
[%emt 0:00:16]} Qxe5 {[%emt 0:01:18]} (22... Qxc2+ {does not save the day} 23.
Kxc2 Kg8 24. Qh5 Kh8 25. Qxh6+ Kg8 26. Qh7#) 23. Rh7+ {
NYTed resigns (Lag: Av=1.63s, max=2.4s) [%emt 0:00:24]} (23. Rh7+ Kg8 24. Qf7#
) 1-0

Dr.Wael Deeb · Post by **Dr.Wael Deeb** » Wed Apr 02, 2008 10:53 pm

My answer is definitely yes Ted as I suffered from this particular approach more than once....

Dann Corbit · Post by **Dann Corbit** » Wed Apr 02, 2008 10:56 pm

To analyze an opening, turn off the book.

Also, for openings the only engine I trust is Rybka and then only at fairly long time control.

If the opening is closed, it will take 30 minutes per move.
If the opening is open, then 10 minutes per move is good enough to know it is reasonable.

Proof:
Try your favorite engines with this position from the Evans Gambit at 5 minutes:
[d]r1bqk1nr/pppp1ppp/2n5/b7/2BpP3/2P2N2/P4PPP/RNBQ1RK1 b kq - bm Nge7;

I predict lousy choices from all of them (that don't have the right move in book).

Mike S. · Post by **Mike S.** » Thu Apr 03, 2008 9:43 am

Dann Corbit wrote: Try your favorite engines with this position from the Evans Gambit at 5 minutes:

r1bqk1nr/pppp1ppp/2n5/b7/2BpP3/2P2N2/P4PPP/RNBQ1RK1 b kq - bm Nge7;

I predict lousy choices from all of them (that don't have the right move in book).

Do you mean bm=Nge7 only? In the ChessOK statistics, some alternatives have good values.

Code: Select all

Move Games White won Black won Draws White ELO Black ELO Year 
d7d6 d6 157 51%  33.1%  15.9%  2780 2637 2006 
g8e7 Nge7 147 46.9%  36.1%  17%  2780 2755 2007 
d4c3 dxc3 126 66.7%  23.8%  9.5%  2780 2550 2007 
g8f6 Nf6 63 42.9%  34.9%  22.2%  2680 2632 2006 
d4d3 d3 58 58.6%  27.6%  13.8%  2519 2545 2007 
a5b6 Bb6 44 45.5%  38.6%  15.9%  2780 2600 2005

http://chessok.com/?page_id=98

AdminX · Post by **AdminX** » Thu Apr 03, 2008 12:04 pm

Hi Dan,

I now understand your point about analysis time. As is normal, results vary from engine to engine.

Thanks
Ted Summers

Results after 5 minutes analysis time:

Analysis by Rybka 2.3.2a mp :

1...Nf6 2.e5 Ne4 3.Qe1
=/+ (-0.38) Depth: 5 00:00:00
= (0.20) Depth: 7 00:00:00 23kN
1...dxc3 2.Qb3 Nf6 3.e5 d5 4.exf6 dxc4 5.fxg7 Rg8
= (-0.22) Depth: 7 00:00:00 50kN
= (0.08) Depth: 10 00:00:02 328kN
1...Nf6 2.e5 Ne4 3.Qe1 Nc5 4.Bg5 Ne7 5.Nxd4 h6 6.Bh4
= (0.07) Depth: 10 00:00:02 363kN
= (0.00) Depth: 13 00:00:05 804kN
1...dxc3 2.Qb3 Qf6 3.Bg5 Qg6 4.Nxc3 Nf6 5.e5 d5 6.Nxd5 Nxd5 7.Bxd5 0-0 8.Rad1
= (-0.17) Depth: 13 00:00:08 1372kN
= (-0.15) Depth: 19 00:03:40 38322kN

Analysis by HIARCS 12 MP:

1...Nge7 2.Nxd4 Ne5 3.Bb3 0-0 4.Qh5 d6 5.f4 N5c6 6.Be3 Bd7 7.Rd1 Qe8 8.f5
=/+ (-0.32) Depth: 8/25 00:00:00 233kN
=/+ (-0.47) Depth: 11/26 00:00:02 1041kN
1...Nf6 2.e5 Ne4 3.Qb3 0-0
=/+ (-0.48) Depth: 11/26 00:00:03 1441kN
= (-0.23) Depth: 12/35 00:00:05 2800kN
1...Nge7 2.cxd4 d5 3.exd5 Nxd5 4.Ba3 Be6 5.Bb5 Nde7 6.Bxe7 Kxe7 7.Bxc6 bxc6 8.Qa4 Qd5 9.Ne5 Bd7
= (-0.24) Depth: 12/35 00:00:06 3170kN
=/+ (-0.58) Depth: 13/35 00:00:18 9033kN
1...dxc3 2.Qb3 Qf6 3.Bg5 Qg6 4.Nxc3 Nf6 5.e5 d5 6.Bxd5 Nxd5 7.Nxd5 0-0 8.Rac1 Qe4 9.Qb5 a6 10.Ne7+ Nxe7 11.Qxa5
=/+ (-0.59) Depth: 13/35 00:00:25 12117kN
=/+ (-0.35) Depth: 16/41 00:04:04 116mN

Analysis by Deep Shredder 11 UCI:

1...Bxc3 2.Nxc3 dxc3
= (-0.24) Depth: 1/3 00:00:00
1...dxc3 2.Bg5
-/+ (-0.80) Depth: 1/3 00:00:00
= (-0.02) Depth: 2/3 00:00:00
1...Bxc3
= (-0.24) Depth: 2/5 00:00:00
1...Nf6 2.e5 d5
=/+ (-0.52) Depth: 2/12 00:00:00
=/+ (-0.51) Depth: 3/13 00:00:00
1...dxc3 2.Qb3 Qe7
-/+ (-0.71) Depth: 3/13 00:00:00
=/+ (-0.29) Depth: 5/14 00:00:00
1...Nf6 2.e5 Ne4 3.Qb3 0-0 4.cxd4 h6 5.Bd5 Ng5 6.Qc4 Nxf3+ 7.Bxf3
=/+ (-0.43) Depth: 5/17 00:00:00
=/+ (-0.45) Depth: 9/15 00:00:00 296kN
1...dxc3 2.Qb3 Qf6 3.Bg5 Qg6 4.Nxc3 Nf6 5.Bxf6 Qxf6 6.Nd5 Qh6 7.Qa3 Qd6 8.Qxd6 cxd6 9.Rfd1 0-0 10.Ne3 Re8 11.Nf5 Bb4 12.Rac1 h6
-/+ (-0.74) Depth: 9/24 00:00:00 541kN
-/+ (-0.78) Depth: 18/46 00:02:58 172mN

Analysis by Naum 3:

1...b5
-/+ (-0.77) Depth: 1/15 00:00:00
1...dxc3 2.Qb3 Nf6
=/+ (-0.55) Depth: 2/13 00:00:00
= (-0.09) Depth: 3/19 00:00:00
1...b5 2.Bxb5 Rb8 3.Bxc6
=/+ (-0.67) Depth: 3/19 00:00:00
= (0.19) Depth: 4/18 00:00:00
1...dxc3 2.Qb3 Nf6 3.Bxf7+
= (-0.11) Depth: 4/18 00:00:00
1...Nf6 2.e5 Ne4 3.Qb3 0-0 4.cxd4 Bb6 5.Nc3 Nxc3 6.Qxc3 d5 7.exd6 Ba5 8.Qd3 Qxd6 9.Ng5 g6 10.Ba3
=/+ (-0.54) Depth: 4/18 00:00:00
= (-0.08) Depth: 18/36 00:04:15 249mN

Analysis by Zappa Mexico II:

1...Nf6 2.cxd4 Nxe4 3.d5 Ne7 4.Qd4 Nd6 5.Qxg7
=/+ (-0.34) Depth: 8/33 00:00:02 224kN
= (-0.11) Depth: 10/33 00:00:04 1493kN
1...dxc3 2.Qb3 Qe7 3.Nxc3 Nf6 4.Ba3 d6 5.Nd5 Qd8 6.Bb2 Nxe4 7.Bxg7
= (-0.12) Depth: 10/39 00:00:05 2110kN
= (-0.21) Depth: 12/39 00:00:12 6296kN
1...Nf6 2.cxd4 Nxe4 3.d5 Ne7 4.Qd4 Nd6 5.Qxg7 Rg8 6.Qd4 Nef5 7.Qe5+ Qe7 8.Bd3 Qxe5 9.Nxe5
= (-0.22) Depth: 12/39 00:00:16 9246kN
= (-0.19) Depth: 13/45 00:00:28 17036kN
1...dxc3 2.Qb3 Kf8 3.Nxc3 d6 4.Bxf7 Qf6 5.Bh5 Bxc3 6.Bg5 Qe6 7.Qxc3 Nf6 8.Bxf6 Qxf6
= (-0.24) Depth: 13/45 00:00:41 25830kN
= (-0.15) Depth: 14/45 00:01:11 45447kN
1...Nf6 2.cxd4 Nxe4 3.d5 Ne7 4.Qd4 Nd6 5.Qxg7 Rg8 6.Qd4 Nef5 7.Qe5+ Qe7 8.Be2 b5
= (-0.19) Depth: 14/45 00:01:28 57257kN
=/+ (-0.35) Depth: 16/51 00:04:54 193mN

Eelco de Groot · Post by **Eelco de Groot** » Thu Apr 03, 2008 4:39 pm

5 minutes or 50 minutes still seems at least a magnitude short for opening positions that have not been analyzed or played in tournament practice, if you just have a single CPU. But with modern multi CPUs this kind of analysis is, it seems to me now much more feasible to do in methodic fashion than say 5 years ago.

The first choice of the computer often seems wrong though even after long analysis. I think that must be shortcomings of the positional eval, but if that is the case it should be possible to correct it as positional characteristics do not vary very dramatically across the openings, one because the distance from the opening position is fairly small, the number of moves played is exactly known and a lot of important positional features are known for just about every opening. I think it just is not developed much because putting knowledge in opening books is usually more efficient. It just does not work in case there is no opening book move available anymore.

Here is some analysis again for Dann's given Evans Gambit line, from Shredder, first Shredder Default in three best moves:

Engine: Shredder 11 Default (256 MB)
by Stefan Meyer-Kahlen
loading shredderbase C:\Documents and Settings\E. de Groot\Mijn documenten\ShredderChess\Shredderbases\all345.tbe
shredderbases with 5 pieces found (157 MB)
-----
18 12:55 +0.62 1...dxc3 2.Qb3 Qf6 3.Bg5 Qg6 4.Nxc3 Bxc3 5.Qxc3 Nf6 6.e5 Ne4 7.Qe3 O-O 8.Bd3 f5 (400.044.741) 516
18 16:50 +0.21 1...Bb6 2.cxd4 d6 3.Bb5 a6 4.Ba4 Kf8 5.Bb2 Bg4 6.Bxc6 bxc6 7.Nbd2 h6 8.Qc2 Ne7 9.Qc3 f6 10.Nh4 Kg8 11.Qg3 (526.586.004) 520
18 18:03 +0.13 1...Bxc3 2.Nxc3 dxc3 3.Qb3 Qf6 4.e5 Qf5 5.Qxc3 Nge7 6.Bd3 Nd5 7.Qb3 Qe6 8.Ba3 Nf4 9.Bc4 Qg6 10.g3 h6 (565.793.460) 521
-----
18 12:55 +0.62 1...dxc3 2.Qb3 Qf6 3.Bg5 Qg6 4.Nxc3 Bxc3 5.Qxc3 Nf6 6.e5 Ne4 7.Qe3 O-O 8.Bd3 f5 (400.044.741) 516
18 16:50 +0.21 1...Bb6 2.cxd4 d6 3.Bb5 a6 4.Ba4 Kf8 5.Bb2 Bg4 6.Bxc6 bxc6 7.Nbd2 h6 8.Qc2 Ne7 9.Qc3 f6 10.Nh4 Kg8 11.Qg3 (526.586.004) 520
18 41:39 +0.16 1...Nge7 2.cxd4 Bb6 3.Ng5 d5 4.exd5 Na5 5.Bb5+ c6 6.Ba3 O-O 7.dxc6 bxc6 8.Bd3 h6 9.Bh7+ Kh8 10.Re1 (1.252.578.867) 501 SB:2
-----
19 47:36 +0.57 1...dxc3 2.Qb3 Qf6 3.Bg5 Qg6 4.Nxc3 Bxc3 5.Qxc3 Nf6 6.e5 Ne4 7.Qe3 O-O 8.Bd3 f5 (1.435.778.456) 502 SB:2
18 16:50 +0.21 1...Bb6 2.cxd4 d6 3.Bb5 a6 4.Ba4 Kf8 5.Bb2 Bg4 6.Bxc6 bxc6 7.Nbd2 h6 8.Qc2 Ne7 9.Qc3 f6 10.Nh4 Kg8 11.Qg3 (526.586.004) 520 SB:2
18 41:39 +0.16 1...Nge7 2.cxd4 Bb6 3.Ng5 d5 4.exd5 Na5 5.Bb5+ c6 6.Ba3 O-O 7.dxc6 bxc6 8.Bd3 h6 9.Bh7+ Kh8 10.Re1 (1.252.578.867) 501 SB:2
-----
19 47:36 +0.57 1...dxc3 2.Qb3 Qf6 3.Bg5 Qg6 4.Nxc3 Bxc3 5.Qxc3 Nf6 6.e5 Ne4 7.Qe3 O-O 8.Bd3 f5 (1.435.778.456) 502 SB:2
19 70:28 +0.33 1...Bb6 2.cxd4 d6 3.d5 Na5 4.Bd3 Nf6 5.Nc3 O-O 6.h3 h6 7.Bb2 c6 8.Na4 Bc7 9.Nc3 Bd7 (2.165.716.013) 512 SB:2
18 41:39 +0.16 1...Nge7 2.cxd4 Bb6 3.Ng5 d5 4.exd5 Na5 5.Bb5+ c6 6.Ba3 O-O 7.dxc6 bxc6 8.Bd3 h6 9.Bh7+ Kh8 10.Re1 (1.252.578.867) 501 SB:2
-----
19 47:36 +0.57 1...dxc3 2.Qb3 Qf6 3.Bg5 Qg6 4.Nxc3 Bxc3 5.Qxc3 Nf6 6.e5 Ne4 7.Qe3 O-O 8.Bd3 f5 (1.435.778.456) 502 SB:2
19 70:28 +0.33 1...Bb6 2.cxd4 d6 3.d5 Na5 4.Bd3 Nf6 5.Nc3 O-O 6.h3 h6 7.Bb2 c6 8.Na4 Bc7 9.Nc3 Bd7 (2.165.716.013) 512 SB:2
19 75:21 +0.32 1...Nge7 2.cxd4 Bb6 3.Ng5 d5 4.exd5 Na5 5.Bb5+ c6 6.Ba3 O-O 7.dxc6 bxc6 8.Bd3 g6 9.Nd2 Bxd4 10.Rc1 Be3 11.fxe3 (2.321.001.660) 513 SB:2
-----
20 96:43 +0.53 1...dxc3 2.Qb3 Qf6 3.Bg5 Qg6 4.Nxc3 Nf6 5.e5 d5 (3.039.017.182) 523 SB:2
19 70:28 +0.33 1...Bb6 2.cxd4 d6 3.d5 Na5 4.Bd3 Nf6 5.Nc3 O-O 6.h3 h6 7.Bb2 c6 8.Na4 Bc7 9.Nc3 Bd7 (2.165.716.013) 512 SB:2
19 75:21 +0.32 1...Nge7 2.cxd4 Bb6 3.Ng5 d5 4.exd5 Na5 5.Bb5+ c6 6.Ba3 O-O 7.dxc6 bxc6 8.Bd3 g6 9.Nd2 Bxd4 10.Rc1 Be3 11.fxe3 (2.321.001.660) 513 SB:2
best move: d4xc3 time: 107:04.328 min n/s: 527.783 CPU 96.5% n/s(1CPU): 546.925 nodes: 3.390.647.316 SB: 2

Judging from lack of tournament practice 1... dxc3 is not good, but is still best move for the computer for at least 20 plies and more, 1... Nge7 only enters the top three after about 40 minutes.

With the slightly different 11+ settings the situation is not much different, the depth achieved is a little better though and not so much time is needed to get Nge7 in the top three. Already after four plies we can see Nge7 in third spot, the eval for 1..dxc3 seems too high though. But 100% sure that it is not playable I am not with this result and I have not analyzed the line:

r1bqk1nr/pppp1ppp/2n5/b7/2BpP3/2P2N2/P4PPP/RNBQ1RK1 b kq -

Engine: Shredder 11+ Beta4 (256 MB)
by Stefan Meyer-Kahlen

4 0:00 +0.92 1...dxc3 2.Qc2 Nf6 3.Ba3 d6 (1.492) 220
4 0:00 +0.44 1...Nf6 2.e5 Ne4 3.cxd4 d5 (1.493) 220
4 0:00 +0.13 1...Nge7 2.Nxd4 O-O 3.Bg5 d6 (4.811) 300

r1bqk1nr/pppp1ppp/2n5/b7/2BpP3/2P2N2/P4PPP/RNBQ1RK1 b kq -

Engine: Shredder 11+ Beta4 (256 MB)
by Stefan Meyer-Kahlen

15 1:57 +0.72 1...dxc3 2.Qb3 Qf6 3.Bg5 Qg6 4.Nxc3 Nf6 5.Bxf6 Qxf6 6.Nd5 Qh6 7.Qa3 Qd6 8.Qxd6 cxd6 9.Rfd1 a6 10.Nf4 b5 11.Be2 Bc3 12.Rac1 Bb2 13.Rc2 (59.047.188) 500
15 2:10 +0.43 1...Nf6 2.e5 d5 (66.267.862) 506
15 2:23 +0.29 1...Nge7 2.cxd4 d5 3.exd5 Nxd5 4.Qb3 Nce7 5.Ba3 c6 6.Bxe7 Kxe7 7.Bxd5 Qxd5 8.Qa3+ Ke8 9.Nc3 Bxc3 10.Qxc3 Kf8 11.Rfe1 Bg4 12.Re5 (72.872.676) 506

____________________________________________________________

16 2:52 +0.68 1...dxc3 2.Qb3 Qf6 3.Bg5 Qg6 4.Nxc3 Nf6 5.Bxf6 Qxf6 6.Nd5 Qh6 7.Qa3 Qd6 8.Qxd6 cxd6 9.Rfd1 O-O 10.Rac1 Re8 11.Ne3 h6 12.Nf5 Bb4 13.Bd5 (87.907.395) 510
16 3:56 +0.35 1...Nf6 2.Ba3 d6 3.e5 Ne4 4.exd6 cxd6 5.Re1 O-O 6.Rxe4 d5 7.Rh4 dxc4 8.Bxf8 Kxf8 9.Rxh7 Kg8 10.Rh4 g5 11.Rh6 Kg7 12.Rxc6 (122.067.282) 516
16 4:05 +0.27 1...Nge7 2.cxd4 d5 3.exd5 Nxd5 4.Qb3 Nce7 5.Ba3 c6 6.Bxe7 Kxe7 7.Bxd5 Qxd5 8.Qa3+ Ke8 9.Nc3 Bxc3 10.Qxc3 Kf8 11.Rfe1 Bg4 12.Re5 Qd6 (127.094.718) 517

____________________________________________________________

17 5:21 +0.72 1...dxc3 2.Qb3 Qf6 3.Bg5 Qg6 4.Nxc3 Nf6 5.Bxf6 Qxf6 6.Nd5 Qh6 7.Qa3 Qd6 8.Qxd6 cxd6 9.Rfd1 O-O 10.Rac1 Re8 11.Ne3 h6 12.Nf5 Bb4 13.Bd5 (167.730.946) 521
17 5:53 +0.35 1...Nf6 2.Ba3 d6 3.e5 Ne4 4.exd6 cxd6 5.Re1 O-O 6.Rxe4 d5 7.Rh4 dxc4 8.Bxf8 Kxf8 9.Rxh7 Kg8 10.Rh4 g5 11.Rh6 Kg7 12.Rxc6 bxc6 13.Nxd4 (184.959.318) 523
17 6:11 +0.11 1...Nge7 2.cxd4 d5 3.exd5 Nxd5 4.Qb3 Nce7 5.Ba3 c6 6.Bxe7 Kxe7 7.Bxd5 Qxd5 8.Qa3+ Ke8 9.Nc3 Bxc3 10.Qxc3 Kf8 11.Rfe1 Bg4 12.Re5 Qd7 13.Rd1 (194.888.403) 524

____________________________________________________________

18 12:41 +0.73 1...dxc3 2.Qb3 Qf6 3.Bg5 Qg6 4.Nxc3 Nf6 5.Bxf6 Qxf6 6.Nd5 Qh6 7.Qa3 Qd6 8.Qxd6 cxd6 9.Rfd1 O-O 10.Ne3 Re8 11.Nf5 Bb4 12.Ng5 Ne5 13.Bb3 d5 (402.590.421) 528
18 15:23 +0.38 1...Nf6 2.Ba3 d6 3.e5 Ne4 4.exd6 cxd6 5.Re1 O-O 6.Rxe4 d5 7.Rh4 dxc4 8.Bxf8 Kxf8 9.Rxh7 Kg8 10.Rh4 g5 11.Rh6 Kg7 12.Rxc6 bxc6 13.Nxd4 (491.960.875) 532
18 20:07 +0.19 1...Nge7 2.cxd4 d5 3.exd5 Nxd5 4.Qb3 Nce7 5.Ba3 c6 6.Bxe7 Kxe7 7.Bxd5 Qxd5 8.Qa3+ Ke8 9.Nc3 Bxc3 10.Qxc3 Kf8 11.Rfe1 Bg4 12.Re5 Qd6 (635.603.491) 526

____________________________________________________________

19 30:04 +0.51 1...dxc3 2.Qb3 Qf6 3.Bg5 Qg6 4.Nxc3 Nf6 5.e5 d5 (951.552.203) 527
19 35:01 +0.32 1...Nf6 2.Ba3 d6 3.e5 Ne4 4.exd6 cxd6 5.Re1 O-O 6.Rxe4 d5 7.Rh4 dxc4 8.Bxf8 Kxf8 9.cxd4 Bf5 10.Nbd2 Bd3 11.Rh5 (1.114.760.059) 530
19 36:11 +0.19 1...Nge7 2.cxd4 d5 3.exd5 Nxd5 4.Qb3 Nce7 5.Ba3 c6 6.Bxe7 Kxe7 7.Bxd5 Qxd5 8.Qa3+ Ke8 9.Nc3 Bxc3 10.Qxc3 Kf8 11.Rfe1 h6 12.Re5 (1.153.087.409) 531

____________________________________________________________

20 48:24 +0.53 1...dxc3 2.Qb3 Qf6 3.Bg5 Qg6 4.Nxc3 Nf6 5.e5 d5 (1.550.684.318) 533
20 54:11 +0.32 1...Nf6 2.Ba3 d6 3.e5 Ne4 4.exd6 cxd6 5.Re1 O-O 6.Rxe4 d5 7.Rh4 dxc4 8.Bxf8 Kxf8 9.cxd4 Bf5 10.Nbd2 Bd3 11.Rc1 b5 12.Kh1 Kg8 (1.743.775.104) 536
20 79:08 +0.24 1...Nge7 2.cxd4 Bb6 3.Ng5 d5 4.exd5 Na5 5.Bb5+ c6 6.Ba3 O-O 7.dxc6 bxc6 8.Bd3 g6 9.Nd2 Re8 10.Re1 (2.564.343.779) 540

____________________________________________________________

21 121:24 +0.43 1...dxc3 2.Qb3 Qf6 3.Bg5 Qg6 4.Nxc3 Nf6
5.e5 d5 (3.966.879.113) 544 SB:16

21 256:37 +0.28 1...Nge7 2.cxd4 Bb6 3.Ng5 d5 4.exd5 Na5
5.Bb5+ c6 6.Ba3 O-O 7.dxc6 bxc6
8.Bd3 g6 9.Nd2 Bxd4 10.Rc1 Be3 (8.604.946.762) 558 SB:48

21 242:40 +0.26 1...Nf6 2.e5 Ne4 3.Ba3 d5 (8.122.679.745) 557 SB:48

____________________________________________________________

21 121:24 +0.43 1...dxc3 2.Qb3 Qf6 3.Bg5 Qg6 4.Nxc3 Nf6
5.e5 d5 (3.966.879.113) 544 SB:16

21 256:37 +0.28 1...Nge7 2.cxd4 Bb6 3.Ng5 d5 4.exd5 Na5
5.Bb5+ c6 6.Ba3 O-O 7.dxc6 bxc6
8.Bd3 g6 9.Nd2 Bxd4 10.Rc1 Be3 (8.604.946.762) 558 SB:48

21 421:31 +0.28 1...Bb6 2.cxd4 Nge7 3.Ng5 d5 4.exd5 Na5
5.Bb5+ c6 6.Ba3 O-O 7.dxc6 bxc6
8.Bd3 g6 9.Nd2 Bxd4 10.Rc1 Be3 (13.752.390.980) 543 SB:63

Dann Corbit · Post by **Dann Corbit** » Thu Apr 03, 2008 9:40 pm

Mike S. wrote:
Dann Corbit wrote: Try your favorite engines with this position from the Evans Gambit at 5 minutes:

r1bqk1nr/pppp1ppp/2n5/b7/2BpP3/2P2N2/P4PPP/RNBQ1RK1 b kq - bm Nge7;

I predict lousy choices from all of them (that don't have the right move in book).
Do you mean bm=Nge7 only? In the ChessOK statistics, some alternatives have good values.
Code: Select all
Move Games White won Black won Draws White ELO Black ELO Year 
d7d6 d6 157 51%  33.1%  15.9%  2780 2637 2006 
g8e7 Nge7 147 46.9%  36.1%  17%  2780 2755 2007 
d4c3 dxc3 126 66.7%  23.8%  9.5%  2780 2550 2007 
g8f6 Nf6 63 42.9%  34.9%  22.2%  2680 2632 2006 
d4d3 d3 58 58.6%  27.6%  13.8%  2519 2545 2007 
a5b6 Bb6 44 45.5%  38.6%  15.9%  2780 2600 2005 
http://chessok.com/?page_id=98

Filter on Elo very high and see what results you get.
The strongest players choose Nge7, and dxc3 is a losing blunder that lots of computer like until several hours of analysis where they finally see that it loses.
d7d6 is a viable alternative, but probably not as good as Nge7.

Mike S. · Post by **Mike S.** » Thu Apr 03, 2008 11:35 pm

Dann Corbit wrote: dxc3 is a losing blunder that lots of computer like until several hours of analysis where they finally see that it loses.

It is surprising to find how many famous players chose 7...dxc3, among them Steinitz, Tarrasch, Lasker... (I am not talking about simuls; they also played it against strong opponents in regular games).

[Event "New York"]
[Site "New York"]
[Date "1963.??.??"]
[Round "?"]
[White "Fischer, Robert James"]
[Black "Fine, Reuben"]
[Result "1-0"]
[ECO "C52"]
[PlyCount "33"]

1. e4 e5 2. Nf3 Nc6 3. Bc4 Bc5 4. b4 Bxb4 5. c3 Ba5 6. d4 exd4 7. O-O dxc3 8.
Qb3 Qe7 9. Nxc3 Nf6 10. Nd5 Nxd5 11. exd5 Ne5 12. Nxe5 Qxe5 13. Bb2 Qg5 14. h4
Qxh4 15. Bxg7 Rg8 16. Rfe1+ Kd8 17. Qg3 1-0

I don't find dxc3 examples from comparable players, after 1963

Engine Analysis

Engine Analysis

Re: Engine Analysis

Re: Engine Analysis

Re: Engine Analysis

Re: Engine Analysis

Re: Engine Analysis

Re: Engine Analysis

Re: Engine Analysis