Opening ideas for TCEC 9

[Leto]

There have been discussions about how to make TCEC 9's superfinal more interesting than TCEC8's, and an opening exit eval of about .80 has been suggested by Kai Laskos as ideal. In this thread I'll post some openings I've found that come close to that, and others can post their openings as well. Perhaps we can work on a list of openings that the organizers of TCEC may consider for TCEC9.

Evals from Komodo 9.3:

1) -.79 d29 1. e4 c6 2. d4 d5 3. Nc3 dxe4 4. f3 exf3 5. Nxf3 Nf6 6. Bc4 Bf5 7.Ne5 e6 8. O-O


2) .75 d31 1. f4 Nc6 2. Nf3 d6 3. e4 g6 4. d4 e6 5. c3
Nf6 6. Bd3 Bd7

3) -.75 d31
1. e4 e5 2. f4 Bc5 3. Qg4 Nf6

4) .89 d34 1. d4 d6 2. c4 g6 3. Nf3 Bg7 4. g3 c6 5.
Nc3 e5 6. Bg2 Bg4

[Leto]

Evals from K9.3 c=0:

5) .87 d32 1. e4 e6 2. Nf3 c5 3. Nc3 d5 4. d4 dxe4 5.
Nxe4 Qc7

6) .85 d31 1. e4 e5 2. d4 Bd6

[Leto]

Evals from K9.3 c=0:

7. .84 d31 1. d4 Nf6 2. Nf3 c5 3. c4 d5 4. g3 dxc4 5.
Qa4+ Nc6 6. dxc5 Qd5

[Norm Pollock]

In tcec 8 there was a distinction between "balanced" and "unbalanced" opening positions. "Balanced" openings occurred in 66% of the games.

Continuing that approach requires two different sets of openings. But first, I would like to see unambiguous definitions of "balanced" and "unbalanced".

My understanding is:

"balanced" is when the sets of pieces are exactly the same or will be the same after an obvious recapture on the next move. No distinction is made for bishops if they are on different colors.

"unbalanced" is not the above.

If I'm missing something, please let me know.

[Peter Berger]

There have been discussions about how to make TCEC 9's superfinal more interesting than TCEC8's, and an opening exit eval of about .80 has been suggested by Kai Laskos as ideal.

I don't think this is sufficient given that we want sth like an equal number of wins and draws. Or to put it a little more clearly: 0.80 is definitely not enough.

The "unbalanced" TCEC openings were an obvious failure, but it is not 100% trivial to do better. I have worked some hours on this problem by now and want to point out the two obvious obstacles I faced:
a.) Advantages might tend to peter out with deeper search depths.
b.) Clear advantages might be too close to being a forced win to make sense.

I found only one line I really like so far and that meets all my demands.

1. e4 Nf6 2. e5 Nd5 3. d4 d6 4. c4 Nb6 5. f4 dxe5 6. fxe5 c5 7. d5 e6 8. Nc3 exd5

Peter

[jdart]

I don't think you should go just by eval. If you do that you may just include inferior opening lines that won't lead to more interesting games, just more one-sided games.

I think if you want to make it more interesting you should go for complex positions, regardless of eval.

Here is one for example (Najdorf, B97):

1. e4 c5 2. Nf3 d6 3. d4 cxd4 4. Nxd4 Nf6 5. Nc3 a6 6. Bg5 e6 7. f4 Qb6 8. Qd2 Qxb2 9. Rb1 Qa3 10. Bxf6 gxf6 11. e5

11. e5 is probably inferior to 11. f5 but leads to very complex play.

--Jon

[michiguel]

There have been discussions about how to make TCEC 9's superfinal more interesting than TCEC8's, and an opening exit eval of about .80 has been suggested by Kai Laskos as ideal. In this thread I'll post some openings I've found that come close to that, and others can post their openings as well. Perhaps we can work on a list of openings that the organizers of TCEC may consider for TCEC9.

Evals from Komodo 9.3:

1) -.79 d29 1. e4 c6 2. d4 d5 3. Nc3 dxe4 4. f3 exf3 5. Nxf3 Nf6 6. Bc4 Bf5 7.Ne5 e6 8. O-O


2) .75 d31 1. f4 Nc6 2. Nf3 d6 3. e4 g6 4. d4 e6 5. c3
Nf6 6. Bd3 Bd7

3) -.75 d31
1. e4 e5 2. f4 Bc5 3. Qg4 Nf6

4) .89 d34 1. d4 d6 2. c4 g6 3. Nf3 Bg7 4. g3 c6 5.
Nc3 e5 6. Bg2 Bg4

Openings were fine.

You need to take into account that chess players may be observing this (who may want to learn from the theory generated), not only computer chess enthusiasts who wants to see blood, even if the games are distorted with unorthodox starts. We are double nerdy. Let's the single nerds enjoy chess.

The real alternatives are
1) play without opening books and allow learning.
2) play with truly short opening books (2 or 3 moves).

Unfortunately, 1 is not going to happen.

Miguel

[Ajedrecista]

Hello Leto:

I think that the definition of ±(0.8 ± |epsilon|) is ill-defined because it is not the same -0.8 by Komodo 9.3 than -0.8 for SF dev., for example, and the same applies for the rest of engines. I see the same flaw with depth: which depth? The more, the merrier, of course.

Anyway, here is my modest contribution of openings of circa ±0.8 eval using SF 6 as the reference (evals are from the side to move). I hope no typos:

------------------------

a) 1. Nf3 Nf6 2. Ng1 Rg8 3. Nf3 Rh8 4. Ng1 Nf6

The loss of castling kingside for the black side.

Code: Select all

FEN: rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQKBNR w KQq - 8 5

[...]

 26/34-	06:38	 247.986.857	622.641	+0,77	d2d4 Ng8f6
 26/34	07:08	 266.827.823	622.405	+0,73	d2d4 d7d5 e2e3 Bc8f5 c2c4 c7c6 Ng1f3 e7e6 Nb1c3 Nb8d7 Bf1d3 Bf5xd3 Qd1xd3 Ng8f6 OO h7h5 b2b3 Bf8b4 Bc1b2 g7g6 Rf1c1 a7a6 c4xd5 Bb4xc3 Bb2xc3 Nf6xd5
 27/36+	08:22	 312.431.039	621.481	+0,79	d2d4
 27/36-	08:43	 325.518.628	621.458	+0,73	d2d4 d7d5
 27/36+	08:48	 328.272.965	621.341	+0,79	d2d4
 27/36	09:00	 335.697.025	620.905	+0,78	d2d4 d7d5 Ng1f3 Nb8c6 c2c4 Bc8g4 Nb1c3 e7e6 c4xd5 e6xd5 e2e3 Ng8f6 h2h3 Bg4xf3 Qd1xf3 Bf8b4 a2a3 Bb4xc3+ b2xc3 Qd8e7 Bf1d3 a7a6 OO OOO Bc1b2 Kc8b8 c3c4 h7h6 Rf1c1
 28/36-	10:05	 375.997.202	620.536	+0,72	d2d4 d7d5
 28/36+	10:29	 390.751.759	620.625	+0,78	d2d4
 28/36-	11:04	 412.206.700	620.661	+0,72	d2d4 d7d5
 28/36	14:56	 553.967.007	618.061	+0,67	d2d4 d7d5 Ng1f3 Bc8f5 c2c4 e7e6 e2e3 Nb8c6 a2a3 Ng8f6 c4xd5 Qd8xd5 b2b4 a7a6 Nb1c3 Qd5d7 Bf1e2 Nf6d5 Bc1d2 OOO OO Kc8b8 Nf3h4 g7g6 Nh4xf5 g6xf5 b4b5 a6xb5

------------------------

b) 1. e4 e5 2. Nf3 f5 3. exf5

The Latvian Countergambit.

Code: Select all

FEN: rnbqkbnr/pppp2pp/8/4pP2/8/5N2/PPPP1PPP/RNBQKB1R b KQkq - 0 3

[...]

 26/39	05:23	 210.235.412	649.811	-0,77	Ng8f6 Nb1c3 Nb8c6 d2d4 e5xd4 Nf3xd4 Bf8b4 Nd4xc6 b7xc6 Qd1e2+ Qd8e7 Qe2xe7+ Ke8xe7 Bf1d3 Ke7f7 OO Rh8e8 Bc1f4 d7d6 Rf1e1 Re8xe1+ Ra1xe1 Ra8b8 a2a3 Bb4xc3 b2xc3 Bc8d7 Bf4d2 Nf6g4 Bd2g5 Ng4e5
 27/39-	06:46	 263.622.663	647.769	-0,83	Ng8f6 d2d4
 27/39+	07:05	 275.645.527	647.789	-0,77	Ng8f6
 27/39-	07:15	 282.351.568	647.941	-0,83	Ng8f6 d2d4
 27/39-	08:29	 330.790.376	648.686	-0,97	Ng8f6 d2d4
 27/39+	09:07	 354.069.883	646.368	-0,86	Ng8f6
 27/39	09:28	 367.286.035	646.308	-0,72	Ng8f6 d2d4 e5xd4 Nf3xd4 Nb8c6 Bc1e3 Nc6xd4 Qd1xd4 c7c5 Qd4a4 Qd8b6 b2b3 Qb6c7 Nb1d2 Bf8e7 Bf1c4 Qc7c6 Qa4xc6 d7xc6 Bc4e6 Bc8xe6 f5xe6 Ra8d8 OOO OO Nd2c4 Nf6g4 Rd1xd8
 28/39-	09:59	 387.421.420	646.408	-0,78	Ng8f6 d2d4
 28/39+	10:46	 418.554.988	647.415	-0,72	Ng8f6
 28/39-	10:59	 426.835.645	647.224	-0,78	Ng8f6 d2d4
 28/39	11:14	 435.999.682	646.792	-0,72	Ng8f6 d2d4 e5xd4 Nf3xd4 Nb8c6 Bc1e3 d7d5 Bf1b5 Bc8d7 Nd4e6 Bf8b4+ Be3d2 Bd7xe6 f5xe6 Qd8e7 OO OO Bb5xc6 b7xc6 Bd2xb4 Qe7xb4 Nb1d2 Ra8e8 Rf1e1 Nf6e4 Nd2xe4 d5xe4 Qd1g4 Qb4b6 Qg4xe4 Rf8xf2

------------------------

c) 1. e4 e5 2. Nf3 Nc6 3. Bb5 Qe7

Ruy López, Vinogradov variation.

Code: Select all

FEN: r1b1kbnr/ppppqppp/2n5/1B2p3/4P3/5N2/PPPP1PPP/RNBQK2R w KQkq - 4 4 

[...]

 26/34	03:06	 115.062.047	618.091	+0,78	OO a7a6 Bb5a4 Ng8f6 Nb1c3 d7d6 d2d4 b7b5 Ba4b3 Nc6a5 Bc1d2 Na5xb3 a2xb3 h7h6 Nc3xb5 Nf6xe4 Rf1e1 Bc8b7 Nb5c3 f7f5 h2h3 Qe7e6 Ra1a5 Ne4xd2 Nf3xd2 c7c5 d4xe5 d6xe5
 27/34-	03:24	 126.804.734	619.311	+0,72	OO a7a6
 27/34+	03:32	 131.908.827	619.377	+0,78	OO
 27/37-	04:38	 172.818.895	619.453	+0,72	OO Ng8f6
 27/37+	05:46	 215.125.430	620.514	+0,80	OO
 27/37-	06:11	 230.625.345	620.506	+0,69	OO Ng8f6
 27/40	07:14	 270.221.187	621.240	+0,74	OO Ng8f6 Bb5a4 d7d6 Nb1c3 Bc8g4 d2d4 e5xd4 Qd1xd4 Bg4xf3 g2xf3 Qe7d7 Ba4xc6 b7xc6 Rf1d1 Bf8e7 e4e5 d6xe5 Qd4xe5 Qd7e6 Bc1f4 Qe6xe5 Bf4xe5 OO Be5xc7 Rf8e8 Bc7d6 h7h6 Bd6xe7 Re8xe7
 28/40-	09:26	 351.962.245	621.222	+0,67	OO Ng8f6
 28/40+	09:48	 365.537.069	621.280	+0,74	OO
 28/40+	10:17	 384.121.481	621.569	+0,83	OO
 28/40-	10:39	 397.799.082	621.787	+0,75	OO Ng8f6
 28/40	11:06	 414.428.592	621.738	+0,68	OO Ng8f6 Bb5a4 d7d6 Nb1c3 Bc8d7 Rf1e1 a7a6 Nc3d5 Qe7d8 d2d4 h7h6 c2c3 Bf8e7 Nd5xe7 Qd8xe7 d4d5 Nc6a5 h2h3 OO Ba4c2 Na5c4 Qd1e2 Nc4b6 b2b3 Rf8e8 Bc1b2 Ra8d8 c3c4

------------------------

d) 1. e4 e5 2. Nf3 Nf6 3. Nxe5 d6 4. Nxf7 Kxf7 5. d4 Nxe4 6. Qh5+ g6 7. Qd5+ Kg7 8. Qxe4

Petrov Defense, Cochrane Gambit.

Code: Select all

FEN: rnbq1b1r/ppp3kp/3p2p1/8/3PQ3/8/PPP2PPP/RNB1KB1R b KQ - 0 8 

[...]

 26/34	02:13	  91.799.546	688.116	-0,80	Nb8c6 d4d5 Nc6e5 Bf1e2 Bf8e7 OO Bc8f5 Qe4a4 a7a6 Bc1e3 Be7f6 Nb1d2 Rh8e8 Ra1e1 Qd8d7 Qa4xd7+ Ne5xd7 g2g4 Bf5e4 Nd2xe4 Re8xe4 c2c3 Nd7e5 g4g5 Bf6e7 Kg1g2 Ra8e8 h2h3 Ne5d7 c3c4
 27/37	03:19	 136.409.279	683.913	-0,81	Nb8c6 d4d5 Nc6e5 Qe4f4 Bc8f5 Bf1e2 h7h6 OO Kg7h7 Nb1a3 c7c6 Na3c4 Ne5xc4 Qf4xc4 Bf8g7 Bc1e3 c6c5 c2c3 Rh8e8 Qc4b3 Ra8b8 Ra1e1 a7a6 a2a4 Qd8f6 a4a5 Bf5e4 h2h3
 28/39	05:05	 213.705.844	699.775	-0,81	Nb8c6 d4d5 Qd8e7 Qe4xe7+ Nc6xe7 c2c4 Bc8f5 Nb1c3 Kg7f7 Nc3b5 Ra8c8 Nb5xa7 Rc8e8 Bc1e3 Bf8g7 Na7b5 Bg7xb2 Ra1d1 Bf5d7 Nb5xc7 Re8c8 Nc7e6 Bb2c3+ Be3d2 Bc3xd2+ Rd1xd2 Bd7xe6 d5xe6+ Kf7xe6 Rd2b2 Rc8c7 Bf1d3 d6d5 c4xd5+ Ne7xd5 OO

------------------------

The hash (1 GB) was cleared between analysis. I have not considered limits of moves. Most of my proposals are too short, sure. Deeper searches are needed.

Regards from Spain.

Ajedrecista.

[Graham Banks]

There have been discussions about how to make TCEC 9's superfinal more interesting than TCEC8's, and an opening exit eval of about .80 has been suggested by Kai Laskos as ideal

Not a good idea, as you'll see too many one-sided games.

People should just accept that in computer chess, there is going to be a high draw rate between two closely matched opponents.

[jdart]

Couple more gambits, maybe interesting:

King's Gambit, Flude Variation:

1. e4 e5 2. f4 exf4 3. Nf3 g5 4. h4 g4 5. Ne5 Nf6 6. Bc4 d5 7. exd5 Bd6 8. d4
Nh5 9. Nc3 O-O 10. Ne4 f5 11. Ng5 Nd7 12. Qd3

Blackmar-Diemer Gambit:

1. d4 d5 2. e4 dxe4 3. Nc3 Nf6 4. f3 Bf5 5. g4 Bg6 6. g5 Nd5 7. Nxe4 e6