Gromit has finished playing

[Pete Galati]

Geschrieben von:/Posted by: Pete Galati am 23 Dezember 1999 02:50:25:
Gromit has just finished, all 60 games are replayed, Gromit finished at over 50%, TSCP had trouble erroring out when palying against Gromit in matches so I'll have to get around to running debug files after the match is over and pass that on to Tom. I believe that TSCP is probably stronger when not being used through Winboard, I don't have time to test that right now.
Old GNU plays the next round of matches.
Pete

Results from file knightx.pgn:
No. Name Win Draw Loss Unf. Score Games %
----------------------------------------------------------------------
1 Comet-wb +44 =6 -10 *0 47.0 60 78.3%
2 TCB +42 =8 -10 *0 46.0 60 76.7%
3 Phalanx +41 =5 -14 *0 43.5 60 72.5%
4 Crafty +37 =11 -12 *0 42.5 60 70.8%
5 Knightx +27 =9 -24 *0 31.5 60 52.5%
6 Gromit20 +27 =8 -25 *0 31.0 60 51.7%
7 GNUchess +9 =4 -23 *0 11.0 36 30.6%
8 GNUches5 +5 =3 -28 *0 6.5 36 18.1%
9 Dragon311 +5 =2 -29 *0 6.0 36 16.7%
10 Fortress +3 =3 -30 *0 4.5 36 12.5%
11 TSCP +0 =1 -35 *0 0.5 36 1.4%
Total Games: 270
White Wins: 137 (50.7%)
Black Wins: 103 (38.1%)
Draws: 30 (11.1%)
Unfinished: 0 (0.0%)


view ascii text file of games

[Pete Galati]

Geschrieben von:/Posted by: Pete Galati am 23 Dezember 1999 06:07:21:
Als Antwort auf:/As an answer to: Gromit has finished playing geschrieben von:/posted by: Pete Galati am 23 Dezember 1999 02:50:25:
Probably as interesting as anything in this tournament is the recent games between the new GNUchess5 (the one that came with Winboard, and the old GNUchess. The old GNU is ahead in the tournament at this point, but the new GNUchess5 won the 6 game match between them.
Pete

Results from file gnugnu.pgn:
No. Name Win Draw Loss Unf. Score Games %
----------------------------------------------------------------------
1 GNUches5 +4 =0 -2 *0 4.0 6 66.7%
2 GNUchess +2 =0 -4 *0 2.0 6 33.3%
Total Games: 6
White Wins: 5 (83.3%)
Black Wins: 1 (16.7%)
Draws: 0 (0.0%)
Unfinished: 0 (0.0%)


view GNU vs GNU game file

[Pete Galati]

Geschrieben von:/Posted by: Pete Galati am 25 Dezember 1999 01:27:44:
Als Antwort auf:/As an answer to: Re: GNU vs GNU geschrieben von:/posted by: Pete Galati am 23 Dezember 1999 06:07:21:
My tournament is finished. A couple of quick notes, both Knightx and TSCP don't seem to admit being checkmated (is that only in matches?) to Winboard. When they're checkmated they do stop calculating moves and eventually Winboard says that they have lost on time because Winboard doesn't know that there was a checkmate either.
Also, some of Crafty bad showing can probably be accounted for because Crafty was useing _my_ opening book and not it's own. So I can retest the Crafty games using Crafty's own book as a separate tournament to see what kind of difference that would have made, but I'm _not_ going to do that this year.
Also, I beleive TSCP is crippled by operating as a Winboard engine since it seems to restricting the move to an approximately 2 second search time in Winboard, I think that TSCP will probably beat 50% of these programs on a regular basis if operated on it's own, not interfaced through Winboard, but it did get 6 wins and 1 draw anyhow
Also, the only tablebases that any of these programs used were the ones that came with Knightx, and TCB didn't use them at all.
There was a 305 move game between Dragon and TSCP that appears to have been a draw several several moves before 100, but TSCP is given credit for winning on time.
Pete

Results from file knightx.pgn:
No. Name Win Draw Loss Unf. Score Games %
----------------------------------------------------------------------
1 Comet-wb +44 =6 -10 *0 47.0 60 78.3%
2 TCB +42 =8 -10 *0 46.0 60 76.7%
3 Phalanx +41 =5 -14 *0 43.5 60 72.5%
4 Crafty +37 =11 -12 *0 42.5 60 70.8%
5 Knightx +27 =9 -24 *0 31.5 60 52.5%
6 Gromit20 +27 =8 -25 *0 31.0 60 51.7%
7 GNUchess +24 =5 -31 *0 26.5 60 44.2%
8 GNUches5 +21 =4 -35 *0 23.0 60 38.3%
9 Dragon311 +16 =4 -40 *0 18.0 60 30.0%
10 Fortress +13 =3 -44 *0 14.5 60 24.2%
11 TSCP +6 =1 -53 *0 6.5 60 10.8%
Total Games: 330
White Wins: 171 (51.8%)
Black Wins: 127 (38.5%)
Draws: 32 (9.7%)
Unfinished: 0 (0.0%)


View the games pgn file

[Frank Quisinsky]

Geschrieben von:/Posted by: Frank Quisinsky am 26 Dezember 1999 22:06:17:
Als Antwort auf:/As an answer to: Re: my tournament is finished geschrieben von:/posted by: Pete Galati am 25 Dezember 1999 01:27:44:

My tournament is finished. A couple of quick notes, both Knightx and TSCP don't seem to admit being checkmated (is that only in matches?) to Winboard. When they're checkmated they do stop calculating moves and eventually Winboard says that they have lost on time because Winboard doesn't know that there was a checkmate either.
Also, some of Crafty bad showing can probably be accounted for because Crafty was useing _my_ opening book and not it's own. So I can retest the Crafty games using Crafty's own book as a separate tournament to see what kind of difference that would have made, but I'm _not_ going to do that this year.
Also, I beleive TSCP is crippled by operating as a Winboard engine since it seems to restricting the move to an approximately 2 second search time in Winboard, I think that TSCP will probably beat 50% of these programs on a regular basis if operated on it's own, not interfaced through Winboard, but it did get 6 wins and 1 draw anyhow
Also, the only tablebases that any of these programs used were the ones that came with Knightx, and TCB didn't use them at all.
There was a 305 move game between Dragon and TSCP that appears to have been a draw several several moves before 100, but TSCP is given credit for winning on time.
Pete

Results from file knightx.pgn:
No. Name Win Draw Loss Unf. Score Games %
----------------------------------------------------------------------
1 Comet-wb +44 =6 -10 *0 47.0 60 78.3%
2 TCB +42 =8 -10 *0 46.0 60 76.7%
3 Phalanx +41 =5 -14 *0 43.5 60 72.5%
4 Crafty +37 =11 -12 *0 42.5 60 70.8%
5 Knightx +27 =9 -24 *0 31.5 60 52.5%
6 Gromit20 +27 =8 -25 *0 31.0 60 51.7%
7 GNUchess +24 =5 -31 *0 26.5 60 44.2%
8 GNUches5 +21 =4 -35 *0 23.0 60 38.3%
9 Dragon311 +16 =4 -40 *0 18.0 60 30.0%
10 Fortress +13 =3 -44 *0 14.5 60 24.2%
11 TSCP +6 =1 -53 *0 6.5 60 10.8%
Total Games: 330
White Wins: 171 (51.8%)
Black Wins: 127 (38.5%)
Draws: 32 (9.7%)
Unfinished: 0 (0.0%)

I will put the tournament later on my page. I hope that is OK !
Thanks Pete !
Kind regards
Frank

[Pete Galati]

Geschrieben von:/Posted by: Pete Galati am 27 Dezember 1999 00:10:03:
Als Antwort auf:/As an answer to: Re: my tournament is finished geschrieben von:/posted by: Frank Quisinsky am 26 Dezember 1999 22:06:17:

My tournament is finished. A couple of quick notes, both Knightx and TSCP don't seem to admit being checkmated (is that only in matches?) to Winboard. When they're checkmated they do stop calculating moves and eventually Winboard says that they have lost on time because Winboard doesn't know that there was a checkmate either.
Also, some of Crafty bad showing can probably be accounted for because Crafty was useing _my_ opening book and not it's own. So I can retest the Crafty games using Crafty's own book as a separate tournament to see what kind of difference that would have made, but I'm _not_ going to do that this year.
Also, I beleive TSCP is crippled by operating as a Winboard engine since it seems to restricting the move to an approximately 2 second search time in Winboard, I think that TSCP will probably beat 50% of these programs on a regular basis if operated on it's own, not interfaced through Winboard, but it did get 6 wins and 1 draw anyhow
Also, the only tablebases that any of these programs used were the ones that came with Knightx, and TCB didn't use them at all.
There was a 305 move game between Dragon and TSCP that appears to have been a draw several several moves before 100, but TSCP is given credit for winning on time.
Pete

Results from file knightx.pgn:
No. Name Win Draw Loss Unf. Score Games %
----------------------------------------------------------------------
1 Comet-wb +44 =6 -10 *0 47.0 60 78.3%
2 TCB +42 =8 -10 *0 46.0 60 76.7%
3 Phalanx +41 =5 -14 *0 43.5 60 72.5%
4 Crafty +37 =11 -12 *0 42.5 60 70.8%
5 Knightx +27 =9 -24 *0 31.5 60 52.5%
6 Gromit20 +27 =8 -25 *0 31.0 60 51.7%
7 GNUchess +24 =5 -31 *0 26.5 60 44.2%
8 GNUches5 +21 =4 -35 *0 23.0 60 38.3%
9 Dragon311 +16 =4 -40 *0 18.0 60 30.0%
10 Fortress +13 =3 -44 *0 14.5 60 24.2%
11 TSCP +6 =1 -53 *0 6.5 60 10.8%
Total Games: 330
White Wins: 171 (51.8%)
Black Wins: 127 (38.5%)
Draws: 32 (9.7%)
Unfinished: 0 (0.0%)

I will put the tournament later on my page. I hope that is OK !
Thanks Pete !
Kind regards
Frank

Yes, that will be ok, in the next day or 2 I'll send you the cfg files from all the engines and the winboard.ini file too, along with my computer information.
Pete

[Frank Quisinsky]

Geschrieben von:/Posted by: Frank Quisinsky am 28 Dezember 1999 21:19:45:
Als Antwort auf:/As an answer to: Re: my tournament is finished geschrieben von:/posted by: Pete Galati am 27 Dezember 1999 00:10:03:

My tournament is finished. A couple of quick notes, both Knightx and TSCP don't seem to admit being checkmated (is that only in matches?) to Winboard. When they're checkmated they do stop calculating moves and eventually Winboard says that they have lost on time because Winboard doesn't know that there was a checkmate either.
Also, some of Crafty bad showing can probably be accounted for because Crafty was useing _my_ opening book and not it's own. So I can retest the Crafty games using Crafty's own book as a separate tournament to see what kind of difference that would have made, but I'm _not_ going to do that this year.
Also, I beleive TSCP is crippled by operating as a Winboard engine since it seems to restricting the move to an approximately 2 second search time in Winboard, I think that TSCP will probably beat 50% of these programs on a regular basis if operated on it's own, not interfaced through Winboard, but it did get 6 wins and 1 draw anyhow
Also, the only tablebases that any of these programs used were the ones that came with Knightx, and TCB didn't use them at all.
There was a 305 move game between Dragon and TSCP that appears to have been a draw several several moves before 100, but TSCP is given credit for winning on time.
Pete

Results from file knightx.pgn:
No. Name Win Draw Loss Unf. Score Games %
----------------------------------------------------------------------
1 Comet-wb +44 =6 -10 *0 47.0 60 78.3%
2 TCB +42 =8 -10 *0 46.0 60 76.7%
3 Phalanx +41 =5 -14 *0 43.5 60 72.5%
4 Crafty +37 =11 -12 *0 42.5 60 70.8%
5 Knightx +27 =9 -24 *0 31.5 60 52.5%
6 Gromit20 +27 =8 -25 *0 31.0 60 51.7%
7 GNUchess +24 =5 -31 *0 26.5 60 44.2%
8 GNUches5 +21 =4 -35 *0 23.0 60 38.3%
9 Dragon311 +16 =4 -40 *0 18.0 60 30.0%
10 Fortress +13 =3 -44 *0 14.5 60 24.2%
11 TSCP +6 =1 -53 *0 6.5 60 10.8%
Total Games: 330
White Wins: 171 (51.8%)
Black Wins: 127 (38.5%)
Draws: 32 (9.7%)
Unfinished: 0 (0.0%)

I will put the tournament later on my page. I hope that is OK !
Thanks Pete !
Kind regards
Frank

Yes, that will be ok, in the next day or 2 I'll send you the cfg files from all the engines and the winboard.ini file too, along with my computer information.
Pete

Pete,
OK, I have all data and the tournament stand in the tournament rubrik !
Thanks another ...
And the next ?
Havegheer 3
WT-6
Boissel 12
)
Regards
Frank

[Pete Galati]

Geschrieben von:/Posted by: Pete Galati am 29 Dezember 1999 06:05:24:
Als Antwort auf:/As an answer to: Re: On my webpage ! geschrieben von:/posted by: Frank Quisinsky am 28 Dezember 1999 21:19:45:

My tournament is finished. A couple of quick notes, both Knightx and TSCP don't seem to admit being checkmated (is that only in matches?) to Winboard. When they're checkmated they do stop calculating moves and eventually Winboard says that they have lost on time because Winboard doesn't know that there was a checkmate either.
Also, some of Crafty bad showing can probably be accounted for because Crafty was useing _my_ opening book and not it's own. So I can retest the Crafty games using Crafty's own book as a separate tournament to see what kind of difference that would have made, but I'm _not_ going to do that this year.
Also, I beleive TSCP is crippled by operating as a Winboard engine since it seems to restricting the move to an approximately 2 second search time in Winboard, I think that TSCP will probably beat 50% of these programs on a regular basis if operated on it's own, not interfaced through Winboard, but it did get 6 wins and 1 draw anyhow
Also, the only tablebases that any of these programs used were the ones that came with Knightx, and TCB didn't use them at all.
There was a 305 move game between Dragon and TSCP that appears to have been a draw several several moves before 100, but TSCP is given credit for winning on time.
Pete

Results from file knightx.pgn:
No. Name Win Draw Loss Unf. Score Games %
----------------------------------------------------------------------
1 Comet-wb +44 =6 -10 *0 47.0 60 78.3%
2 TCB +42 =8 -10 *0 46.0 60 76.7%
3 Phalanx +41 =5 -14 *0 43.5 60 72.5%
4 Crafty +37 =11 -12 *0 42.5 60 70.8%
5 Knightx +27 =9 -24 *0 31.5 60 52.5%
6 Gromit20 +27 =8 -25 *0 31.0 60 51.7%
7 GNUchess +24 =5 -31 *0 26.5 60 44.2%
8 GNUches5 +21 =4 -35 *0 23.0 60 38.3%
9 Dragon311 +16 =4 -40 *0 18.0 60 30.0%
10 Fortress +13 =3 -44 *0 14.5 60 24.2%
11 TSCP +6 =1 -53 *0 6.5 60 10.8%
Total Games: 330
White Wins: 171 (51.8%)
Black Wins: 127 (38.5%)
Draws: 32 (9.7%)
Unfinished: 0 (0.0%)

I will put the tournament later on my page. I hope that is OK !
Thanks Pete !
Kind regards
Frank

Yes, that will be ok, in the next day or 2 I'll send you the cfg files from all the engines and the winboard.ini file too, along with my computer information.
Pete

Pete,
OK, I have all data and the tournament stand in the tournament rubrik !
Thanks another ...
And the next ?
Havegheer 3
WT-6
Boissel 12
)
Regards
Frank

Thanks Frank, I'm wondering how different these tournament will be in the finishing order, we're probably all using different time controls, that seems to make big differences.
Pete