Battle of the Crowns -- calibration results so far

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.

Battle of the Crowns -- calibration results so far

Postby Dann Corbit » 15 Jun 2000, 01:56

Geschrieben von: / Posted by: Dann Corbit at 15 June 2000 02:56:46:
Program Elo + - Games Score Av.Op. Draws
------------- ---- --- --- ----- -------- ------- -------
Crafty : 2499 17 24 977 68.4 % 2365 24.6 %
SOS : 2430 32 36 348 60.2 % 2358 20.4 %
LGoliath : 2416 23 19 827 53.3 % 2393 26.5 %
Comet : 2414 19 18 1105 55.9 % 2372 24.6 %
AnMon : 2395 26 22 670 52.4 % 2378 25.1 %
Phalanx : 2385 20 24 796 49.4 % 2390 19.6 %
TCBishop : 2359 23 26 627 46.7 % 2382 21.5 %
Gromit3 : 2351 60 51 131 51.1 % 2343 19.8 %
Gromit2 : 2344 25 27 545 44.1 % 2385 24.4 %
Francesca : 2325 28 27 486 40.8 % 2389 25.7 %
Amy : 2316 35 62 215 70.0 % 2169 11.6 %
ZChess : 2312 22 25 705 46.5 % 2337 23.4 %
Bringer : 2292 31 32 401 45.4 % 2324 20.0 %
Arasan : 2286 45 37 239 50.4 % 2283 19.7 %
GnuChess4 : 2268 75 75 79 45.6 % 2299 12.7 %
LambChop : 2265 41 34 291 38.3 % 2347 19.6 %
Yace : 2257 44 46 195 57.7 % 2203 20.0 %
InmiChess3 : 2247 67 71 89 57.3 % 2196 15.7 %
Ant : 2244 34 36 339 57.1 % 2194 17.4 %
ExChess3 : 2196 28 38 417 62.8 % 2104 13.4 %
Knightx : 2189 97 83 54 50.9 % 2182 16.7 %
InmiChess2 : 2187 96 76 59 34.7 % 2297 18.6 %
Dragon : 2165 89 46 122 26.2 % 2345 11.5 %
Fortress : 2160 71 80 77 59.1 % 2097 14.3 %
GnuChess5 : 2132 51 66 126 63.9 % 2033 16.7 %
LDBlanche : 2116 73 44 140 28.6 % 2275 17.1 %
Amyan : 2086 64 58 122 44.3 % 2126 8.2 %
Cilian : 2036 52 44 182 38.7 % 2116 18.1 %
SSEChess : 1997 82 49 115 29.1 % 2152 14.8 %
Averno : 1987 77 54 120 38.3 % 2070 3.3 %
Sjeng : 1972 54 44 177 38.4 % 2054 15.8 %
RivalChess : 1917 127 113 32 39.1 % 1995 15.6 %
Freyr : 1885 123 166 28 60.7 % 1810 0.0 %
DChess : 1883 77 59 105 39.5 % 1957 6.7 %
Monik : 1858 74 81 68 45.6 % 1888 20.6 %
Faile : 1830 150 73 50 24.0 % 2030 8.0 %
Chessterfield: 1805 84 20 456 14.7 % 2110 3.9 %
Snail : 1805 222 97 28 16.1 % 2092 10.7 %
TSCP : 1764 154 75 51 26.5 % 1942 2.0 %
Zephyr : 1724 274 150 16 21.9 % 1945 6.2 %
Crux : 1719 0 0 4 0.0 % 2319 0.0 %
Noonian : 1683 209 66 49 12.2 % 2025 8.2 %
Ozwald : 1641 129 129 28 50.0 % 1641 14.3 %
Storm : 1613 331 104 24 12.5 % 1951 0.0 %
Golem01 : 1303 254 78 34 5.9 % 1785 11.8 %
Need games at G/60 for:
RivalChess, Freyr, Faile, Snail, TSCP, Zephyr, Crux, Noonian, Ozwald, Storm, Golem01.


My FTP site
Dann Corbit
 

Re: Battle of the Crowns -- calibration results so far

Postby Pete Galati » 15 Jun 2000, 05:07

Geschrieben von: / Posted by: Pete Galati at 15 June 2000 06:07:17:
Als Antwort auf: / As an answer to: Battle of the Crowns -- calibration results so far geschrieben von: / posted by: Dann Corbit at 15 June 2000 02:56:46:
Sorry if this has been gone over, but how many people are actually running these games in this tournament?
Look at SOS way up there. How many games are sceduled to be played?

Program Elo + - Games Score Av.Op. Draws
------------- ---- --- --- ----- -------- ------- -------
Crafty : 2499 17 24 977 68.4 % 2365 24.6 %
SOS : 2430 32 36 348 60.2 % 2358 20.4 %
LGoliath : 2416 23 19 827 53.3 % 2393 26.5 %
Comet : 2414 19 18 1105 55.9 % 2372 24.6 %
AnMon : 2395 26 22 670 52.4 % 2378 25.1 %
Phalanx : 2385 20 24 796 49.4 % 2390 19.6 %
TCBishop : 2359 23 26 627 46.7 % 2382 21.5 %
Gromit3 : 2351 60 51 131 51.1 % 2343 19.8 %
Gromit2 : 2344 25 27 545 44.1 % 2385 24.4 %
Francesca : 2325 28 27 486 40.8 % 2389 25.7 %
Amy : 2316 35 62 215 70.0 % 2169 11.6 %

{snip}
Pete Galati
 

Re: Battle of the Crowns -- calibration results so far

Postby Jouni » 15 Jun 2000, 06:14

Geschrieben von: / Posted by: Jouni at 15 June 2000 07:14:15:
Als Antwort auf: / As an answer to: Battle of the Crowns -- calibration results so far geschrieben von: / posted by: Dann Corbit at 15 June 2000 02:56:46:
Program Elo + - Games Score Av.Op. Draws
------------- ---- --- --- ----- -------- ------- -------
Crafty : 2499 17 24 977 68.4 % 2365 24.6 %
SOS : 2430 32 36 348 60.2 % 2358 20.4 %
LGoliath : 2416 23 19 827 53.3 % 2393 26.5 %
Comet : 2414 19 18 1105 55.9 % 2372 24.6 %
AnMon : 2395 26 22 670 52.4 % 2378 25.1 %
Phalanx : 2385 20 24 796 49.4 % 2390 19.6 %
TCBishop : 2359 23 26 627 46.7 % 2382 21.5 %
Gromit3 : 2351 60 51 131 51.1 % 2343 19.8 %
Gromit2 : 2344 25 27 545 44.1 % 2385 24.4 %
Francesca : 2325 28 27 486 40.8 % 2389 25.7 %
Amy : 2316 35 62 215 70.0 % 2169 11.6 %
ZChess : 2312 22 25 705 46.5 % 2337 23.4 %
Bringer : 2292 31 32 401 45.4 % 2324 20.0 %
Arasan : 2286 45 37 239 50.4 % 2283 19.7 %
GnuChess4 : 2268 75 75 79 45.6 % 2299 12.7 %
LambChop : 2265 41 34 291 38.3 % 2347 19.6 %
Yace : 2257 44 46 195 57.7 % 2203 20.0 %
InmiChess3 : 2247 67 71 89 57.3 % 2196 15.7 %
Ant : 2244 34 36 339 57.1 % 2194 17.4 %
ExChess3 : 2196 28 38 417 62.8 % 2104 13.4 %
Knightx : 2189 97 83 54 50.9 % 2182 16.7 %
InmiChess2 : 2187 96 76 59 34.7 % 2297 18.6 %
Dragon : 2165 89 46 122 26.2 % 2345 11.5 %
Fortress : 2160 71 80 77 59.1 % 2097 14.3 %
GnuChess5 : 2132 51 66 126 63.9 % 2033 16.7 %
LDBlanche : 2116 73 44 140 28.6 % 2275 17.1 %
Amyan : 2086 64 58 122 44.3 % 2126 8.2 %
Cilian : 2036 52 44 182 38.7 % 2116 18.1 %
SSEChess : 1997 82 49 115 29.1 % 2152 14.8 %
Averno : 1987 77 54 120 38.3 % 2070 3.3 %
Sjeng : 1972 54 44 177 38.4 % 2054 15.8 %
RivalChess : 1917 127 113 32 39.1 % 1995 15.6 %
Freyr : 1885 123 166 28 60.7 % 1810 0.0 %
DChess : 1883 77 59 105 39.5 % 1957 6.7 %
Monik : 1858 74 81 68 45.6 % 1888 20.6 %
Faile : 1830 150 73 50 24.0 % 2030 8.0 %
Chessterfield: 1805 84 20 456 14.7 % 2110 3.9 %
Snail : 1805 222 97 28 16.1 % 2092 10.7 %
TSCP : 1764 154 75 51 26.5 % 1942 2.0 %
Zephyr : 1724 274 150 16 21.9 % 1945 6.2 %
Crux : 1719 0 0 4 0.0 % 2319 0.0 %
Noonian : 1683 209 66 49 12.2 % 2025 8.2 %
Ozwald : 1641 129 129 28 50.0 % 1641 14.3 %
Storm : 1613 331 104 24 12.5 % 1951 0.0 %
Golem01 : 1303 254 78 34 5.9 % 1785 11.8 %
Need games at G/60 for:
RivalChess, Freyr, Faile, Snail, TSCP, Zephyr, Crux, Noonian, Ozwald, Storm, Golem01.
Can You explain what is this Battle of the Crowns? No info available in your
ftp...
thanks Jouni
Jouni
 

Re: Battle of the Crowns -- calibration results so far

Postby Dann Corbit » 15 Jun 2000, 06:43

Geschrieben von: / Posted by: Dann Corbit at 15 June 2000 07:43:40:
Als Antwort auf: / As an answer to: Re: Battle of the Crowns -- calibration results so far geschrieben von: / posted by: Jouni at 15 June 2000 07:14:15:
[snip]
Can You explain what is this Battle of the Crowns? No info available in your
ftp...
The idea came from Volker's Chess Engines overview page. He ranks each engine by strength, going from one star for the weakest ones to four stars for the strongest ones.
By his own admission, the strengths are only guesses based upon his observations. In other words, we all have some pretty good ideas about how strong the freely available winboard engines are, but we don't really know for sure -- not scientifically.
My idea for "The Battle of the Crowns" was to first calibrate the programs with thousands of games at G/60 or slower time control on hardware that is at least PII 300 MHz (or the equivalent by time handicapping).
After each program has been calibrated to withing +/1 100 ELO, we will form four starting groups of 1-4 stars by the ELO average ranking. Within each of these groups, we will play a round robin tournament. Within the round robin tournament, each program will play each opponent in its category 4 games, two as white and two as black at G/90.
At the conclusion of the tournament, the winner of each group will be promoted to the higher group and the loser demoted to the lower group (except the very best and the very worst, which have nowhere to go).
That's the basic idea. If you would like any more details, please let me know.
There are dozens of posts on that topic here in the Winboard forum. If you look in the Archives section, you can read a lot more about it. The calibration has been going on for a couple of months now.


m
Dann Corbit
 

Re: Battle of the Crowns -- calibration results so far

Postby Dann Corbit » 15 Jun 2000, 06:48

Geschrieben von: / Posted by: Dann Corbit at 15 June 2000 07:48:17:
Als Antwort auf: / As an answer to: Re: Battle of the Crowns -- calibration results so far geschrieben von: / posted by: Pete Galati at 15 June 2000 06:07:17:
Sorry if this has been gone over, but how many people are actually running these games in this tournament?
Look at SOS way up there. How many games are sceduled to be played?

Program Elo + - Games Score Av.Op. Draws
------------- ---- --- --- ----- -------- ------- -------
Crafty : 2499 17 24 977 68.4 % 2365 24.6 %
SOS : 2430 32 36 348 60.2 % 2358 20.4 %
LGoliath : 2416 23 19 827 53.3 % 2393 26.5 %
Comet : 2414 19 18 1105 55.9 % 2372 24.6 %
AnMon : 2395 26 22 670 52.4 % 2378 25.1 %
Phalanx : 2385 20 24 796 49.4 % 2390 19.6 %
TCBishop : 2359 23 26 627 46.7 % 2382 21.5 %
Gromit3 : 2351 60 51 131 51.1 % 2343 19.8 %
Gromit2 : 2344 25 27 545 44.1 % 2385 24.4 %
Francesca : 2325 28 27 486 40.8 % 2389 25.7 %
Amy : 2316 35 62 215 70.0 % 2169 11.6 %

{snip}
Err... Not sure exactly. I run 4 machines, and several others are helping by sending me games by email. I also gather games from other contests that are run at G/60 or slower from other people's contests and add these into the calibration data. That's how I came up with 7000 games (14,000 hour's worth of computer time!)
The calibration will continue until the ELO +/- is within a tolerance of +/- 100 ELO. All of the upper crust programs are finished. There are only about ten or so lower level programs that need more games. I think that Crux and Freyr have the stiff-arm in the list and will rise greatly when more games come in. Both are (I think) a lot stronger than their current position shows. But I could be wrong. That's why we go ahead and run the games.
After calibration is finished, there will be four divisions of 11-12 programs each. Within each division, each program will play 4 games against each of the other programs -- two as white and two as black.


My FTP site
Dann Corbit
 

Re: Battle of the Crowns -- calibration results so far

Postby Dann Corbit » 15 Jun 2000, 07:14

Geschrieben von: / Posted by: Dann Corbit at 15 June 2000 08:14:26:
Als Antwort auf: / As an answer to: Battle of the Crowns -- calibration results so far geschrieben von: / posted by: Dann Corbit at 15 June 2000 02:56:46:
[snip]
Need games at G/60 for:
RivalChess, Freyr, Faile, Snail, TSCP, Zephyr, Crux, Noonian, Ozwald, Storm, Golem01.
And ColChess!!!
Need games for ColChess!
Pretty please, down on my hands and knees.


My FTP site
Dann Corbit
 

Re: Battle of the Crowns -- calibration results so far

Postby Mogens Larsen » 15 Jun 2000, 07:53

Geschrieben von: / Posted by: Mogens Larsen at 15 June 2000 08:53:07:
Als Antwort auf: / As an answer to: Re: Battle of the Crowns -- calibration results so far geschrieben von: / posted by: Dann Corbit at 15 June 2000 08:14:26:
And ColChess!!!
Need games for ColChess!
Pretty please, down on my hands and knees.
Currently I have:
Sjeng - Colchess: 2 games.
Averno - Colchess: 4 games.
Planned:
Zephyr - ColChess: ?
Crux - ColChess: ?
Due to the timecontrol of Game/90 it's a very slow process. I also have to finish round 4 of my Superwb tournament before I change computer. Hopefully this will happen today and then it's all Colchess. Hopefully others are conducting games with ColChess as well.
Best wishes...
Mogens
Mogens Larsen
 

Re: Battle of the Crowns -- calibration results so far

Postby Gábor Szõts » 15 Jun 2000, 08:37

Geschrieben von: / Posted by: Gábor Szõts at 15 June 2000 09:37:09:
Als Antwort auf: / As an answer to: Re: Battle of the Crowns -- calibration results so far geschrieben von: / posted by: Mogens Larsen at 15 June 2000 08:53:07:
Hopefully others are conducting games with ColChess as well.
Yesterday I ran a ColChess-Crux game (CC won). I'm planning to play three more games today. I'll send them tomorrow. More to follow, only I don't know at what pace.
Gábor Szõts
 

Re: Battle of the Crowns -- calibration results so far

Postby Colin Frayn » 15 Jun 2000, 09:48

Geschrieben von: / Posted by: Colin Frayn at 15 June 2000 10:48:57:
Als Antwort auf: / As an answer to: Re: Battle of the Crowns -- calibration results so far geschrieben von: / posted by: Gábor Szõts at 15 June 2000 09:37:09:
Hopefully others are conducting games with ColChess as well.
Yesterday I ran a ColChess-Crux game (CC won).
Great news! I just finished running two matches;
ColChess - Amy (WLD 0-11-1) :(
ColChess - Sjeng (WLD 7-5-4) :)
Just started ColChess - Faile. Shouldn't take too long ;)
Dann - I'll send you the results when I'm done.
Cheers,
Col

ColChess Homepage
Colin Frayn
 

Re: Battle of the Crowns -- calibration results so far

Postby Aaron » 15 Jun 2000, 11:31

Geschrieben von: / Posted by: Aaron at 15 June 2000 12:31:24:
Als Antwort auf: / As an answer to: Re: Battle of the Crowns -- calibration results so far geschrieben von: / posted by: Dann Corbit at 15 June 2000 07:48:17:
Program Elo + - Games Score Av.Op. Draws
------------- ---- --- --- ----- -------- ------- -------
Crafty : 2499 17 24 977 68.4 % 2365 24.6 %
SOS : 2430 32 36 348 60.2 % 2358 20.4 %
LGoliath : 2416 23 19 827 53.3 % 2393 26.5 %
Comet : 2414 19 18 1105 55.9 % 2372 24.6 %
AnMon : 2395 26 22 670 52.4 % 2378 25.1 %
Phalanx : 2385 20 24 796 49.4 % 2390 19.6 %
TCBishop : 2359 23 26 627 46.7 % 2382 21.5 %
Gromit3 : 2351 60 51 131 51.1 % 2343 19.8 %
Gromit2 : 2344 25 27 545 44.1 % 2385 24.4 %
Francesca : 2325 28 27 486 40.8 % 2389 25.7 %
Amy : 2316 35 62 215 70.0 % 2169 11.6 %
Nice to see my own list (based on vague intution and results seen) of "top 10" corresponds with what you have (I count gromit3 and 2 has one engine..)...
More interesting to me is to see the next group..Hmm GNUchess4 is quite good..Bringer isn't a surprise, it has being in the "almost there" position for some time..Same for Arasan..
Yace's performance at longer time controls isn't nearly quite as impressive as it's Blitz performance , I think..Where it belongs to the top ranked programs..i.e 4 crowns..At G/60 it's merely 3 crowns?
Aaron
 

Re: Battle of the Crowns -- calibration results so far

Postby Jon Dart » 15 Jun 2000, 15:01

Geschrieben von: / Posted by: Jon Dart at 15 June 2000 16:01:06:
Als Antwort auf: / As an answer to: Re: Battle of the Crowns -- calibration results so far geschrieben von: / posted by: Dann Corbit at 15 June 2000 07:48:17:
I think that Crux and Freyr have the stiff-arm in the list and will rise >greatly when more games come in. Both are (I think) a lot stronger than their >current position shows. But I could be wrong. That's why we go ahead and run >the games.
I am running some games with Freyr and will provide them to Dann soon.
It does seem to be quite a respectable engine. Last I checked it was
+2 =1 -2 against Arasan, but that's a small sample of games. It lost a
game to Gnuchess 4. I will run some more Arasan and Gnuchess games
today.
--Jon
Jon Dart
 

Re: Battle of the Crowns -- calibration results so far

Postby Pete Galati » 15 Jun 2000, 16:31

Geschrieben von: / Posted by: Pete Galati at 15 June 2000 17:31:44:
Als Antwort auf: / As an answer to: Re: Battle of the Crowns -- calibration results so far geschrieben von: / posted by: Jon Dart at 15 June 2000 16:01:06:
I think that Crux and Freyr have the stiff-arm in the list and will rise >greatly when more games come in. Both are (I think) a lot stronger than their >current position shows. But I could be wrong. That's why we go ahead and run >the games.
I am running some games with Freyr and will provide them to Dann soon.
It does seem to be quite a respectable engine. Last I checked it was
+2 =1 -2 against Arasan, but that's a small sample of games. It lost a
game to Gnuchess 4. I will run some more Arasan and Gnuchess games
today.
--Jon
GNUchess 4 is sort of the Rodney Dangerfield of computer Chess. I sure wish someone had given it the ability to resign at some point before it stopped being developed. But it's a tough program, it can win some games, but it won't win that many.
Pete
Pete Galati
 


Return to Archive (Old Parsimony Forum)

Who is online

Users browsing this forum: Majestic-12 [Bot] and 46 guests