ColChess 6.0 notes

Archive of the old Parsimony forum. Some messages couldn't be restored. Limitations: Search for authors does not work, Parsimony specific formats do not work, threaded view does not work properly. Posting is disabled.

ColChess 6.0 notes

Postby Dann Corbit » 06 Mar 2000, 23:31

Geschrieben von:/Posted by: Dann Corbit at 06 March 2000 23:31:08:
Looks like another big improvement. I ran it on a slow machine, and got 221/300 on WAC at 5 seconds:
Test #300 : White to play
Current Position
----------------
b . . b . r . k
. . . R . p p p pP=Pawn
. . . . q P . . rR=Rook
. . . . p . P Q nN=Knight lower=Black
. . . . P . . . bB=Bishop CAPS=White
. . . . . B . . qQ=Queen
. . . . N . K . kK=King
. . . . . . . .
Thinking 2 345 30 1727 g5g6 h7h6
3 160 130 5202 g5g6 h7h6 f6g7
4 160 531 38272 g5g6 h7h6 f6g7 h8g7
5 160 2023 87452 g5g6 h7h6 f6g7 h8g7 d7f7
6 148 5017 342561 g5g6 h7h6 f6g7 h8g7 d7f7 f8f7
---Answer: g6 **-Correct-**
Correct So Far: 221[61]/300
Result: 221[61]/300 (73%)[20%]
I will run it on a fast machine later tonight, when the machine isn't wasting so much time doing 'actual work'
;-)
I would like to comment that your centipawn evaluation numbers for checkmates are not correct.
For instance, here are some samples:
2rr3k/pp3pp1/1nnqbN1p/3pN3/2pP4/2P3Q1/PPB4P/R4RK1 w - - acd 0; acn 1547; acs 0; bm Qg6; ce 49999824; pv; c0 "**-Correct-**";
r1bq2rk/pp3pbp/2p1p1pQ/7P/3P4/2PB1N2/PP3PPR/2KR4 w - - acd 0; acn 87; acs 0; bm Qxh7+; ce 50000064; pv; c0 "**-Correct-**";
5k2/6pp/p1qN4/1p1p4/3P4/2PKP2Q/PP3r2/3R4 b - - acd 0; acn 253; acs 0; bm Qc4+; ce 49999607; pv; c0 "**-Correct-**";
3q1rk1/p4pp1/2pb3p/3p4/6Pr/1PNQ4/P1PB1PP1/4RRK1 b - - acd 1; acn 5575; acs 0; bm Bh2+; ce 49999651; pv h4xg4 a2a4; c0 "**-Correct-**";
r2rb1k1/pp1q1p1p/2n1p1p1/2bp4/5P2/PP1BPR1Q/1BPN2PP/R5K1 w - - acd 4; acn 321; acs 0; bm Qxh7+; ce 50000053; pv h3xh7+ g8f8 b2f6 c5xe3+ f3xe3; c0 "**-Correct-**";
This one is pretty bizarre:
7k/p7/1R5K/6r1/6p1/6P1/8/8 w - - acd 11; acn 77588; acs 0; bm Rb7; ce 10806; pv b6b7 g5b5 b7xb5 h8g8 b5f5 a7a5 f5f4 a5a4 f4xa4 g8f7 a4xg4; c0 "**-Correct-**";
I can't figure out what that centipawn evaluation might mean.
Here is what the PGN standard says:
16.2.5.6: Opcode "ce": centipawn evaluation
The opcode "ce" indicates the evaluation of the indicated position in centipawn
units. It takes a single operand, an optionally signed integer that gives an
evaluation of the position from the viewpoint of the active player; i.e., the
player with the move. Positive values indicate a position favorable to the
moving player while negative values indicate a position favorable to the
passive player; i.e., the player without the move. A centipawn evaluation
value close to zero indicates a neutral positional evaluation.
Values are restricted to integers that are equal to or greater than -32767 and
are less than or equal to 32766.
A value greater than 32000 indicates the availability of a forced mate to the
active player. The number of plies until mate is given by subtracting the
evaluation from the value 32767. Thus, a winning mate in N fullmoves is a mate
in ((2 * N) - 1) halfmoves (or ply) and has a corresponding centipawn
evaluation of (32767 - ((2 * N) - 1)). For example, a mate on the move (mate
in one) has a centipawn evaluation of 32766 while a mate in five has a
centipawn evaluation of 32758.
A value less than -32000 indicates the availability of a forced mate to the
passive player. The number of plies until mate is given by subtracting the
evaluation from the value -32767 and then negating the result. Thus, a losing
mate in N fullmoves is a mate in (2 * N) halfmoves (or ply) and has a
corresponding centipawn evaluation of (-32767 + (2 * N)). For example, a mate
after the move (losing mate in one) has a centipawn evaluation of -32765 while
a losing mate in five has a centipawn evaluation of -32757.
A value of -32767 indicates an illegal position. A stalemate position has a
centipawn evaluation of zero as does a position drawn due to insufficient
mating material. Any other position known to be a certain forced draw also has
a centipawn evaluation of zero.


My FTP site
Dann Corbit
 

Re: ColChess 6.0 notes

Postby Colin Frayn » 07 Mar 2000, 01:47

Geschrieben von:/Posted by: Colin Frayn at 07 March 2000 01:47:30:
Als Antwort auf:/As an answer to: ColChess 6.0 notes geschrieben von:/posted by: Dann Corbit at 06 March 2000 23:31:08:
Looks like another big improvement. I ran it on a slow machine, and got 221/300 on WAC at 5 seconds:
I would like to comment that your centipawn evaluation numbers for checkmates are not correct.
This one is pretty bizarre:
7k/p7/1R5K/6r1/6p1/6P1/8/8 w - - acd 11; acn 77588; acs 0; bm Rb7; ce 10806; pv b6b7 g5b5 b7xb5 h8g8 b5f5 a7a5 f5f4 a5a4 f4xa4 g8f7 a4xg4; c0 "**-Correct-**";
I can't figure out what that centipawn evaluation might mean.
Splendid news :) I'm getting there.....
Yeah, I don't use the suggested values. In fact I didn't even know that there _were_ any :)
It means 'definite win'. I just add on a large number (10,000) to the score when the position is a definite win, i.e. one in which the side to play, with correct play, can always get checkmate. That's a bit arbitrary I'm afraid but it helps pruning quite a lot if you can give one branch a very large score.
I'll have a look at sorting out the values for when it prints out to EPD. I'll try to get the 'mate in n' analysis working too. At the moment it just says 'mate' if it sees a CM and it's usually obvious what the depth is (because ColChess is brute force, and only rarely do the quiescence width extensions actually help it to spot a mate in more than the expected depth).
Hope that helps,
Cheers,
Col

ColChess Homepage
Colin Frayn
 

Re: ColChess 6.0 notes

Postby Dann Corbit » 07 Mar 2000, 05:05

Geschrieben von:/Posted by: Dann Corbit at 07 March 2000 05:05:46:
Als Antwort auf:/As an answer to: Re: ColChess 6.0 notes geschrieben von:/posted by: Colin Frayn at 07 March 2000 01:47:30:
Here is the last output from WAC at 5 seconds per move on a PIII 500 MHz machine:
Test #300 : White to play
Current Position
----------------
b . . b . r . k
. . . R . p p p pP=Pawn
. . . . q P . . rR=Rook
. . . . p . P Q nN=Knight lower=Black
. . . . P . . . bB=Bishop CAPS=White
. . . . . B . . qQ=Queen
. . . . N . K . kK=King
. . . . . . . .
Thinking 2 345 10 1730 g5g6 h7h6
3 160 70 5205 g5g6 h7h6 f6g7
4 160 320 38285 g5g6 h7h6 f6g7 h8g7
5 160 1212 87462 g5g6 h7h6 f6g7 h8g7 d7f7
6 148 4677 584509 g5g6 h7h6 f6g7 h8g7 d7f7 f8f7
7 148 5027 605054
---Answer: g6 **-Correct-**
Correct So Far: 238[62]/300
Result: 238[62]/300 (79%)[20%]


My FTP site
Dann Corbit
 


Return to Archive (Old Parsimony Forum)

Who is online

Users browsing this forum: No registered users and 31 guests