Doch evaluation contest

[DomLeste]

Why is it that the sample.per plays weaker/evaluates differently then the default settings if they use the same settings?

Finding stronger setting will be like pulling blood from a stone...More testing needed

Don..can you describe how the engine handles the 1st and 2nd column numbers. These 2 numbers seem to effect one other if you change one of them.

example..

pval 850 1150


"Each adjustable weight in Doch is defined by the name of the
evaluation term and 2 integer values describing the opening and
endgame weight to be assigned. Doch interpolates these values between
24 different game stages."

[Don]

Why is it that the sample.per plays weaker/evaluates differently then the default settings if they use the same settings?

Because there is a bug in this. It was fixed with version 1.3.3.

I ran over 20,000 games as a sanity test to prove that this works correctly in version 1.3.3 and got results with 1/2 of a percent. Well within statistical error. One version used sample.per and the other used the defaults.

If you think that is still happening, you have to play thousands of games to prove it unless the effect is really pronounced.

Finding stronger setting will be like pulling blood from a stone...More testing needed

Don..can you describe how the engine handles the 1st and 2nd column numbers. These 2 numbers seem to effect one other if you change one of them.

example..

pval 850 1150

As I say the values are interpolated. In the opening the value 850 is used but in the ending 1150 is used. Once pieces come off the board a value somewhere between them will be used.

Doch has 25 game stages. The game stage is computed by summing up the values of all the non-pawn pieces using this scale:

0 - Knight
1 - Bishop
2 - Rook
6 - Queen

If the stage goes about 24 it is clipped at 24.

"Each adjustable weight in Doch is defined by the name of the
evaluation term and 2 integer values describing the opening and
endgame weight to be assigned. Doch interpolates these values between
24 different game stages."

[David Dahlem]

I haven't played 20,000 games with Doch 1.3.3 yet, but i do seem to notice a difference in strength between the <empty> version and the sample.per version. For sure there is a very noticable differenct in analysis mode.

Doch 1.3.3 sample.per:
fen = rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQKBNR w - -
1 00:00 20 0 +1.06 Nb1c3
2 00:00 102 6.528 +0.09 Nb1c3 Nb8c6
3 00:00 208 13.312 +1.06 Nb1c3 Nb8c6 Ng1f3
4 00:00 662 42.368 +0.09 Nb1c3 Nb8c6 Ng1f3 Ng8f6
5 00:00 821 52.544 +0.40 Nb1c3 Nb8c6 Ng1f3 Ng8f6 d2d3
5 00:00 1.100 70.400 +0.40 Nb1c3 Nb8c6 Ng1f3 Ng8f6 d2d3
6 00:00 2.115 135.360 +0.29 Nb1c3 Nb8c6 Ng1f3 Ng8f6 d2d3 d7d6 a2a3
6 00:00 2.821 180.544 +0.29 Nb1c3 Nb8c6 Ng1f3 Ng8f6 d2d3 d7d6 a2a3
7 00:00 4.308 275.712 +0.09 Nb1c3 Nb8c6 Ng1f3 Ng8f6 d2d3 d7d6 a2a3 a7a6
7 00:00 9.174 293.568 +0.09 Nb1c3 Nb8c6 Ng1f3 Ng8f6 d2d3 d7d6 a2a3 a7a6
8 00:00 14.902 238.432 +0.31 Nb1c3 Nb8c6 Ng1f3 Ng8f6 d2d4 d7d5 a2a3 a7a6 e2e3
8 00:00 18.463 295.408 +0.31 Nb1c3 Nb8c6 Ng1f3 Ng8f6 d2d4 d7d5 a2a3 a7a6 e2e3
9 00:00 32.135 293.805 +0.27 Nb1c3 Nb8c6 Ng1f3 Ng8f6 d2d4 d7d5 e2e3 a7a6 a2a3 Nf6e4 Qd1d3 Bc8f5 Nc3xe4 d5xe4
9 00:00 37.711 301.688 +0.27 Nb1c3 Nb8c6 Ng1f3 Ng8f6 d2d4 d7d5 e2e3 a7a6 a2a3 Nf6e4 Qd1d3 Bc8f5 Nc3xe4 d5xe4
10 00:00 101.618 342.292 +0.02 Nb1c3 d7d6 Ng1f3 Ng8f6 d2d4 h7h6 a2a3 Nb8d7 Qd1d3 c7c6 h2h3
10 00:00 165.824 353.757 +0.09 Ng1f3 Ng8f6 h2h3 d7d6 d2d3 Nb8d7 Nb1d2 h7h6 g2g4 g7g5 b2b4 b7b5
10 00:00 196.462 349.265 +0.09 Ng1f3 Ng8f6 h2h3 d7d6 d2d3 Nb8d7 Nb1d2 h7h6 g2g4 g7g5 b2b4 b7b5
11 00:00 240.038 349.146 +0.02 Ng1f3 Ng8f6 h2h3 d7d6 d2d3 Nb8d7 Nb1d2 h7h6 g2g4 g7g5 c2c4 Bf8g7 Bf1g2 00 00
11 00:00 287.677 347.383 +0.02 Ng1f3 Ng8f6 h2h3 d7d6 d2d3 Nb8d7 Nb1d2 h7h6 g2g4 g7g5 c2c4 Bf8g7 Bf1g2 00 00
12 00:01 382.356 349.582 +0.09 Ng1f3 Ng8f6 h2h3 d7d6 d2d3 Nb8d7 Nb1d2 h7h6 g2g4 g7g5 c2c3 Bf8g7 Bf1g2 c7c6 00 00 b2b4 b7b5
12 00:01 513.940 361.452 +0.18 e2e3 d7d6 Bf1d3 Ng8f6 Nb1c3 Nb8d7 Ng1e2 c7c6 00 Nd7e5 f2f3 Ne5xd3 c2xd3 h7h6 a2a3 g7g5 Ne2g3
12 00:01 548.640 358.295 +0.18 e2e3 d7d6 Bf1d3 Ng8f6 Nb1c3 Nb8d7 Ng1e2 c7c6 00 Nd7e5 f2f3 Ne5xd3 c2xd3 h7h6 a2a3 g7g5 Ne2g3
13 00:02 856.848 370.528 +0.25 e2e3 d7d6 Nb1c3 Ng8f6 Ng1f3 Nb8d7 Bf1e2 h7h6 00 e7e6 h2h3 Bf8e7 d2d4 00 Qd1d3 c7c6 a2a3
13 00:02 924.820 369.928 +0.25 e2e3 d7d6 Nb1c3 Ng8f6 Ng1f3 Nb8d7 Bf1e2 h7h6 00 e7e6 h2h3 Bf8e7 d2d4 00 Qd1d3 c7c6 a2a3
14 00:04 1.824.289 381.550 +0.13 e2e3 d7d6 Nb1c3 Ng8f6 Bf1d3 Nb8d7 Ng1e2 c7c6 00 Nd7e5 h2h3 h7h6 Bd3e4 d6d5 f2f4 Nf6xe4 f4xe5 Ne4xc3 d2xc3 g7g5 Ne2g3
14 00:05 2.034.389 379.594 +0.13 e2e3 d7d6 Nb1c3 Ng8f6 Bf1d3 Nb8d7 Ng1e2 c7c6 00 Nd7e5 h2h3 h7h6 Bd3e4 d6d5 f2f4 Nf6xe4 f4xe5 Ne4xc3 d2xc3 g7g5 Ne2g3
15 00:10 4.019.373 384.514 +0.18 e2e3 d7d6 Nb1c3 Ng8f6 Ng1f3 e7e6 h2h3 Nb8c6 d2d4 Bf8e7 Bf1d3 00 00 h7h6 a2a3 a7a6 b2b4 d6d5 Bc1b2 Qd8d6 Ra1c1 b7b5 Nc3e2
15 00:20 7.436.346 372.690 +0.18 e2e3 d7d6 Nb1c3 Ng8f6 Ng1f3 e7e6 h2h3 Nb8c6 d2d4 Bf8e7 Bf1d3 00 00 h7h6 a2a3 a7a6 b2b4 d6d5 Bc1b2 Qd8d6 Ra1c1 b7b5 Nc3e2
16 00:25 9.381.755 375.035 +0.07 e2e3 Ng8f6 Nb1c3 Nb8c6 Ng1f3 e7e6 h2h3 Bf8e7 d2d4 00 Bf1d3 Nc6b4 00 h7h6 Bd3e2 d7d6 a2a3 Nb4d5 Nc3xd5 e6xd5 Qd1d3 c7c6 Rf1d1


Doch 1.3.3 <empty>:
fen = rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQKBNR w - -
1 00:00 20 0 +0.51 e2e3
2 00:00 122 0 +0.07 e2e3 d7d5
3 00:00 524 0 +0.49 d2d3 d7d6 Ng1f3
4 00:00 1.320 0 +0.09 d2d3 d7d6 Ng1f3 Ng8f6
5 00:00 2.073 132.672 +0.31 d2d3 Nb8c6 Ng1f3 e7e6 Nb1c3
5 00:00 2.745 175.680 +0.33 e2e3 d7d5 Nb1c3 Ng8f6 Ng1f3
5 00:00 3.584 229.376 +0.35 Nb1c3 d7d5 d2d4 Ng8f6 Ng1f3
5 00:00 4.462 285.568 +0.37 Ng1f3 Nb8c6 d2d4 e7e6 Nb1c3
5 00:00 4.632 296.448 +0.37 Ng1f3 Nb8c6 d2d4 e7e6 Nb1c3
6 00:00 7.400 236.800 +0.31 Ng1f3 Ng8f6 d2d3 d7d5 Nb1c3 Qd8d6
6 00:00 8.734 279.488 +0.31 Ng1f3 Ng8f6 d2d3 d7d5 Nb1c3 Qd8d6
7 00:00 17.906 286.496 +0.09 Ng1f3 Ng8f6 d2d3 Nb8c6 Nb1c3 d7d6 Bc1d2 Bc8e6
7 00:00 33.574 358.122 +0.09 Ng1f3 Ng8f6 d2d3 Nb8c6 Nb1c3 d7d6 Bc1d2 Bc8e6
8 00:00 54.006 345.638 +0.11 Ng1f3 Ng8f6 Nb1c3 d7d6 d2d4 Bc8f5 Nf3g5 h7h6 e2e4 h6xg5 e4xf5
8 00:00 64.748 345.322 +0.11 Ng1f3 Ng8f6 Nb1c3 d7d6 d2d4 Bc8f5 Nf3g5 h7h6 e2e4 h6xg5 e4xf5
9 00:00 121.853 339.069 +0.18 Ng1f3 Ng8f6 Nb1c3 d7d5 d2d3 d5d4 Nc3e4 Nb8c6 Bc1f4 Nf6d5 Qd1d2 Bc8f5
9 00:00 188.160 354.183 +0.18 Ng1f3 Ng8f6 Nb1c3 d7d5 d2d3 d5d4 Nc3e4 Nb8c6 Bc1f4 Nf6d5 Qd1d2 Bc8f5
10 00:01 355.921 345.135 +0.22 Ng1f3 Ng8f6 Nb1c3 Nb8c6 d2d4 d7d6 d4d5 Nc6e5 Bc1f4 Ne5g6 g2g3 Bc8f5 Bf4e3
10 00:01 376.121 343.882 +0.22 Ng1f3 Ng8f6 Nb1c3 Nb8c6 d2d4 d7d6 d4d5 Nc6e5 Bc1f4 Ne5g6 g2g3 Bc8f5 Bf4e3
11 00:01 592.932 341.870 +0.20 Ng1f3 Ng8f6 Nb1c3 Nb8c6 d2d4 e7e6 e2e3 d7d5 Bf1d3 Bf8d6 00 00 h2h3 h7h6 Bc1d2
11 00:01 666.542 344.021 +0.20 Ng1f3 Ng8f6 Nb1c3 Nb8c6 d2d4 e7e6 e2e3 d7d5 Bf1d3 Bf8d6 00 00 h2h3 h7h6 Bc1d2
12 00:03 1.284.816 345.496 +0.13 Ng1f3 Ng8f6 Nb1c3 Nb8c6 d2d4 d7d5 a2a3 e7e6 g2g3 Bf8d6 Bf1g2 00 00 Bc8d7 Bc1g5 h7h6
12 00:04 1.475.197 347.105 +0.13 Ng1f3 Ng8f6 Nb1c3 Nb8c6 d2d4 d7d5 a2a3 e7e6 g2g3 Bf8d6 Bf1g2 00 00 Bc8d7 Bc1g5 h7h6
13 00:05 2.027.194 345.974 +0.14 Ng1f3 Ng8f6 Nb1c3 Nb8c6 d2d4 d7d5 a2a3 a7a6 g2g3 g7g6 Bf1g2 Bf8g7 00 00 Bc1f4 Bc8f5 Nf3g5 Qd8d7
13 00:10 3.666.665 343.582 +0.26 e2e4 Ng8f6 e4e5 Nf6d5 Nb1c3 Nd5xc3 d2xc3 d7d6 Ng1f3 Nb8d7 Bc1f4 Nd7c5 Bf1b5+ c7c6 Bb5d3 Nc5e6 Bf4e3 g7g6
13 00:11 3.760.158 343.785 +0.26 e2e4 Ng8f6 e4e5 Nf6d5 Nb1c3 Nd5xc3 d2xc3 d7d6 Ng1f3 Nb8d7 Bc1f4 Nd7c5 Bf1b5+ c7c6 Bb5d3 Nc5e6 Bf4e3 g7g6
14 00:23 8.202.259 347.185 +0.49 e2e4 Ng8f6 e4e5 Nf6d5 Nb1c3 Nd5xc3 d2xc3 d7d6 Ng1f3 Nb8d7 e5xd6 e7xd6 Bf1b5 Bf8e7 00 00 Rf1e1 c7c6 Bb5d3
14 00:23 8.267.189 347.406 +0.49 e2e4 Ng8f6 e4e5 Nf6d5 Nb1c3 Nd5xc3 d2xc3 d7d6 Ng1f3 Nb8d7 e5xd6 e7xd6 Bf1b5 Bf8e7 00 00 Rf1e1 c7c6 Bb5d3
15 00:35 12.294.112 345.704 +0.48 e2e4 Ng8f6 e4e5 Nf6d5 Ng1f3 d7d6 d2d4 d6xe5 d4xe5 e7e6 Bf1e2 Bf8e7 00 00 a2a3 Nb8c6 c2c4 Nd5b6 Nb1c3 Qd8xd1 Rf1xd1 Bc8d7
15 00:36 12.468.338 346.042 +0.48 e2e4 Ng8f6 e4e5 Nf6d5 Ng1f3 d7d6 d2d4 d6xe5 d4xe5 e7e6 Bf1e2 Bf8e7 00 00 a2a3 Nb8c6 c2c4 Nd5b6 Nb1c3 Qd8xd1 Rf1xd1 Bc8d7

[Don]

When you load sample.per do you get a message in the information box saying, "43 parameters changed out of 43 total" ?

Don

I haven't played 20,000 games with Doch 1.3.3 yet, but i do seem to notice a difference in strength between the <empty> version and the sample.per version. For sure there is a very noticable differenct in analysis mode.

Doch 1.3.3 sample.per:
fen = rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQKBNR w - -
1 00:00 20 0 +1.06 Nb1c3
2 00:00 102 6.528 +0.09 Nb1c3 Nb8c6
3 00:00 208 13.312 +1.06 Nb1c3 Nb8c6 Ng1f3
4 00:00 662 42.368 +0.09 Nb1c3 Nb8c6 Ng1f3 Ng8f6
5 00:00 821 52.544 +0.40 Nb1c3 Nb8c6 Ng1f3 Ng8f6 d2d3
5 00:00 1.100 70.400 +0.40 Nb1c3 Nb8c6 Ng1f3 Ng8f6 d2d3
6 00:00 2.115 135.360 +0.29 Nb1c3 Nb8c6 Ng1f3 Ng8f6 d2d3 d7d6 a2a3
6 00:00 2.821 180.544 +0.29 Nb1c3 Nb8c6 Ng1f3 Ng8f6 d2d3 d7d6 a2a3
7 00:00 4.308 275.712 +0.09 Nb1c3 Nb8c6 Ng1f3 Ng8f6 d2d3 d7d6 a2a3 a7a6
7 00:00 9.174 293.568 +0.09 Nb1c3 Nb8c6 Ng1f3 Ng8f6 d2d3 d7d6 a2a3 a7a6
8 00:00 14.902 238.432 +0.31 Nb1c3 Nb8c6 Ng1f3 Ng8f6 d2d4 d7d5 a2a3 a7a6 e2e3
8 00:00 18.463 295.408 +0.31 Nb1c3 Nb8c6 Ng1f3 Ng8f6 d2d4 d7d5 a2a3 a7a6 e2e3
9 00:00 32.135 293.805 +0.27 Nb1c3 Nb8c6 Ng1f3 Ng8f6 d2d4 d7d5 e2e3 a7a6 a2a3 Nf6e4 Qd1d3 Bc8f5 Nc3xe4 d5xe4
9 00:00 37.711 301.688 +0.27 Nb1c3 Nb8c6 Ng1f3 Ng8f6 d2d4 d7d5 e2e3 a7a6 a2a3 Nf6e4 Qd1d3 Bc8f5 Nc3xe4 d5xe4
10 00:00 101.618 342.292 +0.02 Nb1c3 d7d6 Ng1f3 Ng8f6 d2d4 h7h6 a2a3 Nb8d7 Qd1d3 c7c6 h2h3
10 00:00 165.824 353.757 +0.09 Ng1f3 Ng8f6 h2h3 d7d6 d2d3 Nb8d7 Nb1d2 h7h6 g2g4 g7g5 b2b4 b7b5
10 00:00 196.462 349.265 +0.09 Ng1f3 Ng8f6 h2h3 d7d6 d2d3 Nb8d7 Nb1d2 h7h6 g2g4 g7g5 b2b4 b7b5
11 00:00 240.038 349.146 +0.02 Ng1f3 Ng8f6 h2h3 d7d6 d2d3 Nb8d7 Nb1d2 h7h6 g2g4 g7g5 c2c4 Bf8g7 Bf1g2 00 00
11 00:00 287.677 347.383 +0.02 Ng1f3 Ng8f6 h2h3 d7d6 d2d3 Nb8d7 Nb1d2 h7h6 g2g4 g7g5 c2c4 Bf8g7 Bf1g2 00 00
12 00:01 382.356 349.582 +0.09 Ng1f3 Ng8f6 h2h3 d7d6 d2d3 Nb8d7 Nb1d2 h7h6 g2g4 g7g5 c2c3 Bf8g7 Bf1g2 c7c6 00 00 b2b4 b7b5
12 00:01 513.940 361.452 +0.18 e2e3 d7d6 Bf1d3 Ng8f6 Nb1c3 Nb8d7 Ng1e2 c7c6 00 Nd7e5 f2f3 Ne5xd3 c2xd3 h7h6 a2a3 g7g5 Ne2g3
12 00:01 548.640 358.295 +0.18 e2e3 d7d6 Bf1d3 Ng8f6 Nb1c3 Nb8d7 Ng1e2 c7c6 00 Nd7e5 f2f3 Ne5xd3 c2xd3 h7h6 a2a3 g7g5 Ne2g3
13 00:02 856.848 370.528 +0.25 e2e3 d7d6 Nb1c3 Ng8f6 Ng1f3 Nb8d7 Bf1e2 h7h6 00 e7e6 h2h3 Bf8e7 d2d4 00 Qd1d3 c7c6 a2a3
13 00:02 924.820 369.928 +0.25 e2e3 d7d6 Nb1c3 Ng8f6 Ng1f3 Nb8d7 Bf1e2 h7h6 00 e7e6 h2h3 Bf8e7 d2d4 00 Qd1d3 c7c6 a2a3
14 00:04 1.824.289 381.550 +0.13 e2e3 d7d6 Nb1c3 Ng8f6 Bf1d3 Nb8d7 Ng1e2 c7c6 00 Nd7e5 h2h3 h7h6 Bd3e4 d6d5 f2f4 Nf6xe4 f4xe5 Ne4xc3 d2xc3 g7g5 Ne2g3
14 00:05 2.034.389 379.594 +0.13 e2e3 d7d6 Nb1c3 Ng8f6 Bf1d3 Nb8d7 Ng1e2 c7c6 00 Nd7e5 h2h3 h7h6 Bd3e4 d6d5 f2f4 Nf6xe4 f4xe5 Ne4xc3 d2xc3 g7g5 Ne2g3
15 00:10 4.019.373 384.514 +0.18 e2e3 d7d6 Nb1c3 Ng8f6 Ng1f3 e7e6 h2h3 Nb8c6 d2d4 Bf8e7 Bf1d3 00 00 h7h6 a2a3 a7a6 b2b4 d6d5 Bc1b2 Qd8d6 Ra1c1 b7b5 Nc3e2
15 00:20 7.436.346 372.690 +0.18 e2e3 d7d6 Nb1c3 Ng8f6 Ng1f3 e7e6 h2h3 Nb8c6 d2d4 Bf8e7 Bf1d3 00 00 h7h6 a2a3 a7a6 b2b4 d6d5 Bc1b2 Qd8d6 Ra1c1 b7b5 Nc3e2
16 00:25 9.381.755 375.035 +0.07 e2e3 Ng8f6 Nb1c3 Nb8c6 Ng1f3 e7e6 h2h3 Bf8e7 d2d4 00 Bf1d3 Nc6b4 00 h7h6 Bd3e2 d7d6 a2a3 Nb4d5 Nc3xd5 e6xd5 Qd1d3 c7c6 Rf1d1


Doch 1.3.3 <empty>:
fen = rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQKBNR w - -
1 00:00 20 0 +0.51 e2e3
2 00:00 122 0 +0.07 e2e3 d7d5
3 00:00 524 0 +0.49 d2d3 d7d6 Ng1f3
4 00:00 1.320 0 +0.09 d2d3 d7d6 Ng1f3 Ng8f6
5 00:00 2.073 132.672 +0.31 d2d3 Nb8c6 Ng1f3 e7e6 Nb1c3
5 00:00 2.745 175.680 +0.33 e2e3 d7d5 Nb1c3 Ng8f6 Ng1f3
5 00:00 3.584 229.376 +0.35 Nb1c3 d7d5 d2d4 Ng8f6 Ng1f3
5 00:00 4.462 285.568 +0.37 Ng1f3 Nb8c6 d2d4 e7e6 Nb1c3
5 00:00 4.632 296.448 +0.37 Ng1f3 Nb8c6 d2d4 e7e6 Nb1c3
6 00:00 7.400 236.800 +0.31 Ng1f3 Ng8f6 d2d3 d7d5 Nb1c3 Qd8d6
6 00:00 8.734 279.488 +0.31 Ng1f3 Ng8f6 d2d3 d7d5 Nb1c3 Qd8d6
7 00:00 17.906 286.496 +0.09 Ng1f3 Ng8f6 d2d3 Nb8c6 Nb1c3 d7d6 Bc1d2 Bc8e6
7 00:00 33.574 358.122 +0.09 Ng1f3 Ng8f6 d2d3 Nb8c6 Nb1c3 d7d6 Bc1d2 Bc8e6
8 00:00 54.006 345.638 +0.11 Ng1f3 Ng8f6 Nb1c3 d7d6 d2d4 Bc8f5 Nf3g5 h7h6 e2e4 h6xg5 e4xf5
8 00:00 64.748 345.322 +0.11 Ng1f3 Ng8f6 Nb1c3 d7d6 d2d4 Bc8f5 Nf3g5 h7h6 e2e4 h6xg5 e4xf5
9 00:00 121.853 339.069 +0.18 Ng1f3 Ng8f6 Nb1c3 d7d5 d2d3 d5d4 Nc3e4 Nb8c6 Bc1f4 Nf6d5 Qd1d2 Bc8f5
9 00:00 188.160 354.183 +0.18 Ng1f3 Ng8f6 Nb1c3 d7d5 d2d3 d5d4 Nc3e4 Nb8c6 Bc1f4 Nf6d5 Qd1d2 Bc8f5
10 00:01 355.921 345.135 +0.22 Ng1f3 Ng8f6 Nb1c3 Nb8c6 d2d4 d7d6 d4d5 Nc6e5 Bc1f4 Ne5g6 g2g3 Bc8f5 Bf4e3
10 00:01 376.121 343.882 +0.22 Ng1f3 Ng8f6 Nb1c3 Nb8c6 d2d4 d7d6 d4d5 Nc6e5 Bc1f4 Ne5g6 g2g3 Bc8f5 Bf4e3
11 00:01 592.932 341.870 +0.20 Ng1f3 Ng8f6 Nb1c3 Nb8c6 d2d4 e7e6 e2e3 d7d5 Bf1d3 Bf8d6 00 00 h2h3 h7h6 Bc1d2
11 00:01 666.542 344.021 +0.20 Ng1f3 Ng8f6 Nb1c3 Nb8c6 d2d4 e7e6 e2e3 d7d5 Bf1d3 Bf8d6 00 00 h2h3 h7h6 Bc1d2
12 00:03 1.284.816 345.496 +0.13 Ng1f3 Ng8f6 Nb1c3 Nb8c6 d2d4 d7d5 a2a3 e7e6 g2g3 Bf8d6 Bf1g2 00 00 Bc8d7 Bc1g5 h7h6
12 00:04 1.475.197 347.105 +0.13 Ng1f3 Ng8f6 Nb1c3 Nb8c6 d2d4 d7d5 a2a3 e7e6 g2g3 Bf8d6 Bf1g2 00 00 Bc8d7 Bc1g5 h7h6
13 00:05 2.027.194 345.974 +0.14 Ng1f3 Ng8f6 Nb1c3 Nb8c6 d2d4 d7d5 a2a3 a7a6 g2g3 g7g6 Bf1g2 Bf8g7 00 00 Bc1f4 Bc8f5 Nf3g5 Qd8d7
13 00:10 3.666.665 343.582 +0.26 e2e4 Ng8f6 e4e5 Nf6d5 Nb1c3 Nd5xc3 d2xc3 d7d6 Ng1f3 Nb8d7 Bc1f4 Nd7c5 Bf1b5+ c7c6 Bb5d3 Nc5e6 Bf4e3 g7g6
13 00:11 3.760.158 343.785 +0.26 e2e4 Ng8f6 e4e5 Nf6d5 Nb1c3 Nd5xc3 d2xc3 d7d6 Ng1f3 Nb8d7 Bc1f4 Nd7c5 Bf1b5+ c7c6 Bb5d3 Nc5e6 Bf4e3 g7g6
14 00:23 8.202.259 347.185 +0.49 e2e4 Ng8f6 e4e5 Nf6d5 Nb1c3 Nd5xc3 d2xc3 d7d6 Ng1f3 Nb8d7 e5xd6 e7xd6 Bf1b5 Bf8e7 00 00 Rf1e1 c7c6 Bb5d3
14 00:23 8.267.189 347.406 +0.49 e2e4 Ng8f6 e4e5 Nf6d5 Nb1c3 Nd5xc3 d2xc3 d7d6 Ng1f3 Nb8d7 e5xd6 e7xd6 Bf1b5 Bf8e7 00 00 Rf1e1 c7c6 Bb5d3
15 00:35 12.294.112 345.704 +0.48 e2e4 Ng8f6 e4e5 Nf6d5 Ng1f3 d7d6 d2d4 d6xe5 d4xe5 e7e6 Bf1e2 Bf8e7 00 00 a2a3 Nb8c6 c2c4 Nd5b6 Nb1c3 Qd8xd1 Rf1xd1 Bc8d7
15 00:36 12.468.338 346.042 +0.48 e2e4 Ng8f6 e4e5 Nf6d5 Ng1f3 d7d6 d2d4 d6xe5 d4xe5 e7e6 Bf1e2 Bf8e7 00 00 a2a3 Nb8c6 c2c4 Nd5b6 Nb1c3 Qd8xd1 Rf1xd1 Bc8d7

[David Dahlem]

Yes, i do see the message "43 parameters changed out of 43 total" when using sample.per.

[DomLeste]

Yes, i do see the message "43 parameters changed out of 43 total" when using sample.per.

I've never seen that message in Fritz GUI when loading sample.per. What GUI are you using david?

If only the engine can tell us the values by a DOS Prompt commands or something I dont mind manually changing it that way.

[Don]

David and everybody else too:

Your debug dump was very useful to me. I checked this out careful and discovered that this 1.3.3 still has the bug that I fixed. I went back to the buggy version and was able to duplicate your search results exactly.

The version that Jim Ablett compiled was somehow regressed and it was probably my fault. I had sent him too many versions and I had not properly labelled them so I was begging for trouble. I think he may have patched in some of the changes but not others.

The bug is no longer whether the program can load these files - it gets all of that correct now. But there was a typo bug that caused the first 3 or 4 values to be applied to the wrong positional weight. So it would be difficult indeed to make an improved version with this kind of bug.

My fear is that I'm testing everybody's patience with this including Jim Ablett's who works so hard to provide this very useful service for us.

But to make personalities work correctly we need another bug fix version and I have asked Jim Ablett to make yet another version with the fix. It will be Doch-1.3.4 to make things clear.

Don Dailey

I haven't played 20,000 games with Doch 1.3.3 yet, but i do seem to notice a difference in strength between the <empty> version and the sample.per version. For sure there is a very noticable differenct in analysis mode.

Doch 1.3.3 sample.per:
fen = rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQKBNR w - -
1 00:00 20 0 +1.06 Nb1c3
2 00:00 102 6.528 +0.09 Nb1c3 Nb8c6
3 00:00 208 13.312 +1.06 Nb1c3 Nb8c6 Ng1f3
4 00:00 662 42.368 +0.09 Nb1c3 Nb8c6 Ng1f3 Ng8f6
5 00:00 821 52.544 +0.40 Nb1c3 Nb8c6 Ng1f3 Ng8f6 d2d3
5 00:00 1.100 70.400 +0.40 Nb1c3 Nb8c6 Ng1f3 Ng8f6 d2d3
6 00:00 2.115 135.360 +0.29 Nb1c3 Nb8c6 Ng1f3 Ng8f6 d2d3 d7d6 a2a3
6 00:00 2.821 180.544 +0.29 Nb1c3 Nb8c6 Ng1f3 Ng8f6 d2d3 d7d6 a2a3
7 00:00 4.308 275.712 +0.09 Nb1c3 Nb8c6 Ng1f3 Ng8f6 d2d3 d7d6 a2a3 a7a6
7 00:00 9.174 293.568 +0.09 Nb1c3 Nb8c6 Ng1f3 Ng8f6 d2d3 d7d6 a2a3 a7a6
8 00:00 14.902 238.432 +0.31 Nb1c3 Nb8c6 Ng1f3 Ng8f6 d2d4 d7d5 a2a3 a7a6 e2e3
8 00:00 18.463 295.408 +0.31 Nb1c3 Nb8c6 Ng1f3 Ng8f6 d2d4 d7d5 a2a3 a7a6 e2e3
9 00:00 32.135 293.805 +0.27 Nb1c3 Nb8c6 Ng1f3 Ng8f6 d2d4 d7d5 e2e3 a7a6 a2a3 Nf6e4 Qd1d3 Bc8f5 Nc3xe4 d5xe4
9 00:00 37.711 301.688 +0.27 Nb1c3 Nb8c6 Ng1f3 Ng8f6 d2d4 d7d5 e2e3 a7a6 a2a3 Nf6e4 Qd1d3 Bc8f5 Nc3xe4 d5xe4
10 00:00 101.618 342.292 +0.02 Nb1c3 d7d6 Ng1f3 Ng8f6 d2d4 h7h6 a2a3 Nb8d7 Qd1d3 c7c6 h2h3
10 00:00 165.824 353.757 +0.09 Ng1f3 Ng8f6 h2h3 d7d6 d2d3 Nb8d7 Nb1d2 h7h6 g2g4 g7g5 b2b4 b7b5
10 00:00 196.462 349.265 +0.09 Ng1f3 Ng8f6 h2h3 d7d6 d2d3 Nb8d7 Nb1d2 h7h6 g2g4 g7g5 b2b4 b7b5
11 00:00 240.038 349.146 +0.02 Ng1f3 Ng8f6 h2h3 d7d6 d2d3 Nb8d7 Nb1d2 h7h6 g2g4 g7g5 c2c4 Bf8g7 Bf1g2 00 00
11 00:00 287.677 347.383 +0.02 Ng1f3 Ng8f6 h2h3 d7d6 d2d3 Nb8d7 Nb1d2 h7h6 g2g4 g7g5 c2c4 Bf8g7 Bf1g2 00 00
12 00:01 382.356 349.582 +0.09 Ng1f3 Ng8f6 h2h3 d7d6 d2d3 Nb8d7 Nb1d2 h7h6 g2g4 g7g5 c2c3 Bf8g7 Bf1g2 c7c6 00 00 b2b4 b7b5
12 00:01 513.940 361.452 +0.18 e2e3 d7d6 Bf1d3 Ng8f6 Nb1c3 Nb8d7 Ng1e2 c7c6 00 Nd7e5 f2f3 Ne5xd3 c2xd3 h7h6 a2a3 g7g5 Ne2g3
12 00:01 548.640 358.295 +0.18 e2e3 d7d6 Bf1d3 Ng8f6 Nb1c3 Nb8d7 Ng1e2 c7c6 00 Nd7e5 f2f3 Ne5xd3 c2xd3 h7h6 a2a3 g7g5 Ne2g3
13 00:02 856.848 370.528 +0.25 e2e3 d7d6 Nb1c3 Ng8f6 Ng1f3 Nb8d7 Bf1e2 h7h6 00 e7e6 h2h3 Bf8e7 d2d4 00 Qd1d3 c7c6 a2a3
13 00:02 924.820 369.928 +0.25 e2e3 d7d6 Nb1c3 Ng8f6 Ng1f3 Nb8d7 Bf1e2 h7h6 00 e7e6 h2h3 Bf8e7 d2d4 00 Qd1d3 c7c6 a2a3
14 00:04 1.824.289 381.550 +0.13 e2e3 d7d6 Nb1c3 Ng8f6 Bf1d3 Nb8d7 Ng1e2 c7c6 00 Nd7e5 h2h3 h7h6 Bd3e4 d6d5 f2f4 Nf6xe4 f4xe5 Ne4xc3 d2xc3 g7g5 Ne2g3
14 00:05 2.034.389 379.594 +0.13 e2e3 d7d6 Nb1c3 Ng8f6 Bf1d3 Nb8d7 Ng1e2 c7c6 00 Nd7e5 h2h3 h7h6 Bd3e4 d6d5 f2f4 Nf6xe4 f4xe5 Ne4xc3 d2xc3 g7g5 Ne2g3
15 00:10 4.019.373 384.514 +0.18 e2e3 d7d6 Nb1c3 Ng8f6 Ng1f3 e7e6 h2h3 Nb8c6 d2d4 Bf8e7 Bf1d3 00 00 h7h6 a2a3 a7a6 b2b4 d6d5 Bc1b2 Qd8d6 Ra1c1 b7b5 Nc3e2
15 00:20 7.436.346 372.690 +0.18 e2e3 d7d6 Nb1c3 Ng8f6 Ng1f3 e7e6 h2h3 Nb8c6 d2d4 Bf8e7 Bf1d3 00 00 h7h6 a2a3 a7a6 b2b4 d6d5 Bc1b2 Qd8d6 Ra1c1 b7b5 Nc3e2
16 00:25 9.381.755 375.035 +0.07 e2e3 Ng8f6 Nb1c3 Nb8c6 Ng1f3 e7e6 h2h3 Bf8e7 d2d4 00 Bf1d3 Nc6b4 00 h7h6 Bd3e2 d7d6 a2a3 Nb4d5 Nc3xd5 e6xd5 Qd1d3 c7c6 Rf1d1


Doch 1.3.3 <empty>:
fen = rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQKBNR w - -
1 00:00 20 0 +0.51 e2e3
2 00:00 122 0 +0.07 e2e3 d7d5
3 00:00 524 0 +0.49 d2d3 d7d6 Ng1f3
4 00:00 1.320 0 +0.09 d2d3 d7d6 Ng1f3 Ng8f6
5 00:00 2.073 132.672 +0.31 d2d3 Nb8c6 Ng1f3 e7e6 Nb1c3
5 00:00 2.745 175.680 +0.33 e2e3 d7d5 Nb1c3 Ng8f6 Ng1f3
5 00:00 3.584 229.376 +0.35 Nb1c3 d7d5 d2d4 Ng8f6 Ng1f3
5 00:00 4.462 285.568 +0.37 Ng1f3 Nb8c6 d2d4 e7e6 Nb1c3
5 00:00 4.632 296.448 +0.37 Ng1f3 Nb8c6 d2d4 e7e6 Nb1c3
6 00:00 7.400 236.800 +0.31 Ng1f3 Ng8f6 d2d3 d7d5 Nb1c3 Qd8d6
6 00:00 8.734 279.488 +0.31 Ng1f3 Ng8f6 d2d3 d7d5 Nb1c3 Qd8d6
7 00:00 17.906 286.496 +0.09 Ng1f3 Ng8f6 d2d3 Nb8c6 Nb1c3 d7d6 Bc1d2 Bc8e6
7 00:00 33.574 358.122 +0.09 Ng1f3 Ng8f6 d2d3 Nb8c6 Nb1c3 d7d6 Bc1d2 Bc8e6
8 00:00 54.006 345.638 +0.11 Ng1f3 Ng8f6 Nb1c3 d7d6 d2d4 Bc8f5 Nf3g5 h7h6 e2e4 h6xg5 e4xf5
8 00:00 64.748 345.322 +0.11 Ng1f3 Ng8f6 Nb1c3 d7d6 d2d4 Bc8f5 Nf3g5 h7h6 e2e4 h6xg5 e4xf5
9 00:00 121.853 339.069 +0.18 Ng1f3 Ng8f6 Nb1c3 d7d5 d2d3 d5d4 Nc3e4 Nb8c6 Bc1f4 Nf6d5 Qd1d2 Bc8f5
9 00:00 188.160 354.183 +0.18 Ng1f3 Ng8f6 Nb1c3 d7d5 d2d3 d5d4 Nc3e4 Nb8c6 Bc1f4 Nf6d5 Qd1d2 Bc8f5
10 00:01 355.921 345.135 +0.22 Ng1f3 Ng8f6 Nb1c3 Nb8c6 d2d4 d7d6 d4d5 Nc6e5 Bc1f4 Ne5g6 g2g3 Bc8f5 Bf4e3
10 00:01 376.121 343.882 +0.22 Ng1f3 Ng8f6 Nb1c3 Nb8c6 d2d4 d7d6 d4d5 Nc6e5 Bc1f4 Ne5g6 g2g3 Bc8f5 Bf4e3
11 00:01 592.932 341.870 +0.20 Ng1f3 Ng8f6 Nb1c3 Nb8c6 d2d4 e7e6 e2e3 d7d5 Bf1d3 Bf8d6 00 00 h2h3 h7h6 Bc1d2
11 00:01 666.542 344.021 +0.20 Ng1f3 Ng8f6 Nb1c3 Nb8c6 d2d4 e7e6 e2e3 d7d5 Bf1d3 Bf8d6 00 00 h2h3 h7h6 Bc1d2
12 00:03 1.284.816 345.496 +0.13 Ng1f3 Ng8f6 Nb1c3 Nb8c6 d2d4 d7d5 a2a3 e7e6 g2g3 Bf8d6 Bf1g2 00 00 Bc8d7 Bc1g5 h7h6
12 00:04 1.475.197 347.105 +0.13 Ng1f3 Ng8f6 Nb1c3 Nb8c6 d2d4 d7d5 a2a3 e7e6 g2g3 Bf8d6 Bf1g2 00 00 Bc8d7 Bc1g5 h7h6
13 00:05 2.027.194 345.974 +0.14 Ng1f3 Ng8f6 Nb1c3 Nb8c6 d2d4 d7d5 a2a3 a7a6 g2g3 g7g6 Bf1g2 Bf8g7 00 00 Bc1f4 Bc8f5 Nf3g5 Qd8d7
13 00:10 3.666.665 343.582 +0.26 e2e4 Ng8f6 e4e5 Nf6d5 Nb1c3 Nd5xc3 d2xc3 d7d6 Ng1f3 Nb8d7 Bc1f4 Nd7c5 Bf1b5+ c7c6 Bb5d3 Nc5e6 Bf4e3 g7g6
13 00:11 3.760.158 343.785 +0.26 e2e4 Ng8f6 e4e5 Nf6d5 Nb1c3 Nd5xc3 d2xc3 d7d6 Ng1f3 Nb8d7 Bc1f4 Nd7c5 Bf1b5+ c7c6 Bb5d3 Nc5e6 Bf4e3 g7g6
14 00:23 8.202.259 347.185 +0.49 e2e4 Ng8f6 e4e5 Nf6d5 Nb1c3 Nd5xc3 d2xc3 d7d6 Ng1f3 Nb8d7 e5xd6 e7xd6 Bf1b5 Bf8e7 00 00 Rf1e1 c7c6 Bb5d3
14 00:23 8.267.189 347.406 +0.49 e2e4 Ng8f6 e4e5 Nf6d5 Nb1c3 Nd5xc3 d2xc3 d7d6 Ng1f3 Nb8d7 e5xd6 e7xd6 Bf1b5 Bf8e7 00 00 Rf1e1 c7c6 Bb5d3
15 00:35 12.294.112 345.704 +0.48 e2e4 Ng8f6 e4e5 Nf6d5 Ng1f3 d7d6 d2d4 d6xe5 d4xe5 e7e6 Bf1e2 Bf8e7 00 00 a2a3 Nb8c6 c2c4 Nd5b6 Nb1c3 Qd8xd1 Rf1xd1 Bc8d7
15 00:36 12.468.338 346.042 +0.48 e2e4 Ng8f6 e4e5 Nf6d5 Ng1f3 d7d6 d2d4 d6xe5 d4xe5 e7e6 Bf1e2 Bf8e7 00 00 a2a3 Nb8c6 c2c4 Nd5b6 Nb1c3 Qd8xd1 Rf1xd1 Bc8d7

[DomLeste]

Lucky were here to keep you honest Don!

Once this is fixed we can start this contest count me in. Its only patience of the people creating stronger personality thats been tested I'm sure the personality testers will have some hairs left after all these false starts.

[David Dahlem]

Thanks Don

I ran the <empty> and sample.per versions on this test suite, and although you've already confirmed the bug, here are my results -

Analysis from C:\Arena\Epd\STS 1 Undermine.epd
Analyzing engine: Doch 1.3.3 <empty>
77 of 100 matching moves
1/8/2010 2:28:57 PM, Total time: Rated time: 04:36 = 276 Seconds

Analysis from C:\Arena\Epd\STS 1 Undermine.epd
Analyzing engine: Doch 1.3.3 sample.per
67 of 100 matching moves
1/8/2010 2:49:01 PM, Total time: Rated time: 06:25 = 385 Seconds

Looking forward to a Doch version that handles personality files correctly.

[Ron Langeveld]

Thanks on the info Don, and don't worry about such slips. Happens to the best. It's called progress Looks like we're back in business. Looking forward to the new compile.