Battle of the Crowns -- calibration runs (early indications

[Dann Corbit]

Geschrieben von: / Posted by: Dann Corbit at 03 May 2000 23:44:22:
Program Elo + - Games Score Av.Op. Draws
1 Crafty : 2814 142 239 16 84.4 % 2521 18.8 %
2 SOS : 2767 73 109 57 69.3 % 2625 15.8 %
3 Comet : 2728 64 75 86 48.8 % 2736 18.6 %
4 LGoliath : 2722 106 118 36 47.2 % 2742 16.7 %
5 Dragon : 2620 320 378 3 83.3 % 2340 33.3 %
6 TCBishop : 2564 109 118 33 43.9 % 2606 21.2 %
7 GnuChess5 : 2553 130 395 8 93.8 % 2083 12.5 %
8 AnMon : 2481 60 74 96 62.5 % 2392 16.7 %
9 Amy : 2474 56 68 112 61.6 % 2392 16.1 %
10 Arasan : 2378 53 46 167 51.8 % 2366 19.8 %
11 Bringer : 2357 66 60 101 54.5 % 2326 21.8 %
12 Ant : 2342 52 47 160 54.1 % 2314 21.9 %
13 EXchess : 2329 81 111 39 46.2 % 2356 35.9 %
14 ZChess : 2326 42 76 143 73.4 % 2150 16.8 %
15 Knightx : 2233 364 273 8 18.8 % 2488 12.5 %
16 Sjeng : 2165 252 192 12 25.0 % 2356 16.7 %
17 LDBlanche : 2124 153 86 36 20.8 % 2356 19.4 %
18 Amyan : 2069 53 50 175 50.3 % 2067 5.7 %
19 SSEChess : 2028 315 118 20 12.5 % 2366 5.0 %
20 Averno : 1891 165 64 59 20.3 % 2128 3.4 %
21 DChess : 1740 0 0 10 0.0 % 2340 0.0 %
22 Skaki : 1641 223 166 12 58.3 % 1583 33.3 %
23 TSCP : 1569 254 45 86 5.8 % 2053 4.7 %
24 Monik : 1304 0 0 5 0.0 % 1904 0.0 %
25 Pierre : 1304 0 0 6 0.0 % 1904 0.0 %
Games : 743 (finished)
White Wins : 336 (45.2 %)
Black Wins : 289 (38.9 %)
Draws : 118 (15.9 %)
Unfinished : 13
White Perf. : 53.2 %
Black Perf. : 46.8 %

ECO A = 81 Games (10.9 %)
ECO B = 174 Games (23.4 %)
ECO C = 340 Games (45.8 %)
ECO D = 117 Games (15.7 %)
ECO E = 31 Games ( 4.2 %)
The ECO C preponderance is surprising to me. I wonder if that statistic will hold?


My FTP site

[Mogens Larsen]

Geschrieben von: / Posted by: Mogens Larsen at 03 May 2000 23:59:18:
Als Antwort auf: / As an answer to: Battle of the Crowns -- calibration runs (early indications courtesy of Elostat) geschrieben von: / posted by: Dann Corbit at 03 May 2000 23:44:22:

Program Elo + - Games Score Av.Op. Draws
1 Crafty : 2814 142 239 16 84.4 % 2521 18.8 %
2 SOS : 2767 73 109 57 69.3 % 2625 15.8 %
ECO A = 81 Games (10.9 %)
ECO B = 174 Games (23.4 %)
ECO C = 340 Games (45.8 %)
ECO D = 117 Games (15.7 %)
ECO E = 31 Games ( 4.2 %)
The ECO C preponderance is surprising to me. I wonder if that statistic will hold?

This looks nice. I hope SOS can hang on.

Is there a lot of Ruy Lopez? That would be my guess, since a lot of programs use it with black and white.
Best wishes...
Mogens

[Dann Corbit]

Geschrieben von: / Posted by: Dann Corbit at 04 May 2000 00:09:41:
Als Antwort auf: / As an answer to: Re: Battle of the Crowns -- calibration runs (early indications courtesy of Elostat) geschrieben von: / posted by: Mogens Larsen at 03 May 2000 23:59:18:

Program Elo + - Games Score Av.Op. Draws
1 Crafty : 2814 142 239 16 84.4 % 2521 18.8 %
2 SOS : 2767 73 109 57 69.3 % 2625 15.8 %
ECO A = 81 Games (10.9 %)
ECO B = 174 Games (23.4 %)
ECO C = 340 Games (45.8 %)
ECO D = 117 Games (15.7 %)
ECO E = 31 Games ( 4.2 %)
The ECO C preponderance is surprising to me. I wonder if that statistic will hold?

This looks nice. I hope SOS can hang on.

Is there a lot of Ruy Lopez? That would be my guess, since a lot of programs use it with black and white.

Here are the biggest fragment files, split out by ECO by program extract:
4,040 C02.pgn
4,544 C13.pgn
5,876 C92.pgn
6,515 C47.pgn
6,534 C78.pgn
6,874 C10.pgn
8,063 C09.pgn
8,889 C07.pgn
9,246 C49.pgn
9,354 C11.pgn
9,989 C26.pgn
10,287 C42.pgn
10,368 C14.pgn
11,809 C40.pgn
12,006 C15.pgn
12,548 C29.pgn
12,959 C28.pgn
14,534 C43.pgn
14,723 C65.pgn
16,136 C24.pgn
19,215 C22.pgn
49,802 C00.pgn
Looks like C00 French has the lion's share [AnMon and Ant both seem to love this opening], with C22 Center game (some bookless programs mindlessly reapeated this one) and C24 Bishop's opening favored by Zchess.


My FTP site

[Mogens Larsen]

Geschrieben von: / Posted by: Mogens Larsen at 04 May 2000 01:29:59:
Als Antwort auf: / As an answer to: Re: Battle of the Crowns -- calibration runs (early indications courtesy of Elostat) geschrieben von: / posted by: Dann Corbit at 04 May 2000 00:09:41:

Here are the biggest fragment files, split out by ECO by program extract:
4,040 C02.pgn
4,544 C13.pgn
5,876 C92.pgn
6,515 C47.pgn
6,534 C78.pgn
6,874 C10.pgn
8,063 C09.pgn
8,889 C07.pgn
9,246 C49.pgn
9,354 C11.pgn
9,989 C26.pgn
10,287 C42.pgn
10,368 C14.pgn
11,809 C40.pgn
12,006 C15.pgn
12,548 C29.pgn
12,959 C28.pgn
14,534 C43.pgn
14,723 C65.pgn
16,136 C24.pgn
19,215 C22.pgn
49,802 C00.pgn
Looks like C00 French has the lion's share [AnMon and Ant both seem to love this opening], with C22 Center game (some bookless programs mindlessly reapeated this one) and C24 Bishop's opening favored by Zchess.

If you add all the French variations then it's quite significant. I think it's a good choice for black even though things might get a little closed. Thanks for the info.
Best wishes...
Mogens

[pete]

Geschrieben von: / Posted by: pete at 04 May 2000 02:05:50:
Als Antwort auf: / As an answer to: Battle of the Crowns -- calibration runs (early indications courtesy of Elostat) geschrieben von: / posted by: Dann Corbit at 03 May 2000 23:44:22:

Program Elo + - Games Score Av.Op. Draws
1 Crafty : 2814 142 239 16 84.4 % 2521 18.8 %
2 SOS : 2767 73 109 57 69.3 % 2625 15.8 %
3 Comet : 2728 64 75 86 48.8 % 2736 18.6 %
4 LGoliath : 2722 106 118 36 47.2 % 2742 16.7 %
5 Dragon : 2620 320 378 3 83.3 % 2340 33.3 %
6 TCBishop : 2564 109 118 33 43.9 % 2606 21.2 %
7 GnuChess5 : 2553 130 395 8 93.8 % 2083 12.5 %
8 AnMon : 2481 60 74 96 62.5 % 2392 16.7 %
9 Amy : 2474 56 68 112 61.6 % 2392 16.1 %
10 Arasan : 2378 53 46 167 51.8 % 2366 19.8 %
11 Bringer : 2357 66 60 101 54.5 % 2326 21.8 %
12 Ant : 2342 52 47 160 54.1 % 2314 21.9 %
13 EXchess : 2329 81 111 39 46.2 % 2356 35.9 %
14 ZChess : 2326 42 76 143 73.4 % 2150 16.8 %
15 Knightx : 2233 364 273 8 18.8 % 2488 12.5 %
16 Sjeng : 2165 252 192 12 25.0 % 2356 16.7 %
17 LDBlanche : 2124 153 86 36 20.8 % 2356 19.4 %
18 Amyan : 2069 53 50 175 50.3 % 2067 5.7 %
19 SSEChess : 2028 315 118 20 12.5 % 2366 5.0 %
20 Averno : 1891 165 64 59 20.3 % 2128 3.4 %
21 DChess : 1740 0 0 10 0.0 % 2340 0.0 %
22 Skaki : 1641 223 166 12 58.3 % 1583 33.3 %
23 TSCP : 1569 254 45 86 5.8 % 2053 4.7 %
24 Monik : 1304 0 0 5 0.0 % 1904 0.0 %
25 Pierre : 1304 0 0 6 0.0 % 1904 0.0 %

This rating list looked quite screwed to me at first sight , Dann .
Then I took a closer look : Gnu is 2553 ( +130 / -395 ) )
Interesting also how high some of the really provisional ratings are ie crafty

[Dann Corbit]

Geschrieben von: / Posted by: Dann Corbit at 04 May 2000 02:16:11:
Als Antwort auf: / As an answer to: Re: Battle of the Crowns -- calibration runs (early indications courtesy of Elostat) geschrieben von: / posted by: pete at 04 May 2000 02:05:50:

Program Elo + - Games Score Av.Op. Draws
1 Crafty : 2814 142 239 16 84.4 % 2521 18.8 %
2 SOS : 2767 73 109 57 69.3 % 2625 15.8 %
3 Comet : 2728 64 75 86 48.8 % 2736 18.6 %
4 LGoliath : 2722 106 118 36 47.2 % 2742 16.7 %
5 Dragon : 2620 320 378 3 83.3 % 2340 33.3 %
6 TCBishop : 2564 109 118 33 43.9 % 2606 21.2 %
7 GnuChess5 : 2553 130 395 8 93.8 % 2083 12.5 %
8 AnMon : 2481 60 74 96 62.5 % 2392 16.7 %
9 Amy : 2474 56 68 112 61.6 % 2392 16.1 %
10 Arasan : 2378 53 46 167 51.8 % 2366 19.8 %
11 Bringer : 2357 66 60 101 54.5 % 2326 21.8 %
12 Ant : 2342 52 47 160 54.1 % 2314 21.9 %
13 EXchess : 2329 81 111 39 46.2 % 2356 35.9 %
14 ZChess : 2326 42 76 143 73.4 % 2150 16.8 %
15 Knightx : 2233 364 273 8 18.8 % 2488 12.5 %
16 Sjeng : 2165 252 192 12 25.0 % 2356 16.7 %
17 LDBlanche : 2124 153 86 36 20.8 % 2356 19.4 %
18 Amyan : 2069 53 50 175 50.3 % 2067 5.7 %
19 SSEChess : 2028 315 118 20 12.5 % 2366 5.0 %
20 Averno : 1891 165 64 59 20.3 % 2128 3.4 %
21 DChess : 1740 0 0 10 0.0 % 2340 0.0 %
22 Skaki : 1641 223 166 12 58.3 % 1583 33.3 %
23 TSCP : 1569 254 45 86 5.8 % 2053 4.7 %
24 Monik : 1304 0 0 5 0.0 % 1904 0.0 %
25 Pierre : 1304 0 0 6 0.0 % 1904 0.0 %

This rating list looked quite screwed to me at first sight , Dann .
Then I took a closer look : Gnu is 2553 ( +130 / -395 ) )
Interesting also how high some of the really provisional ratings are ie crafty

Nothing surprising about it really. Consider how few the number of games is.
If you look at the deltas (as you mention) they are rather huge.


My FTP site

[Jon Dart]

Geschrieben von: / Posted by: Jon Dart at 04 May 2000 06:08:20:
Als Antwort auf: / As an answer to: Re: Battle of the Crowns -- calibration runs (early indications courtesy of Elostat) geschrieben von: / posted by: Mogens Larsen at 04 May 2000 01:29:59:

If you add all the French variations then it's quite significant. I think it's a good choice for black even though things might get a little closed. Thanks for the info.
Best wishes...
Mogens

I don't think the French in general is bad, but I do think there
are some variations of it that are just bad for Black. E.g.
e4 e6 d4 d5 Nc3 Bb4 e5 b6 is one.
e4 e6 d4 d5 Nc3 Bb4 e5 c5 a3 Ba5 is another - GMs have used
this as Black but computers do not handle it well. I recently put
these lines into "avoid" status in Arasan's book - but they
are still in the 5.3 version that is used in these tests.
--Jon

[Dann Corbit]

Geschrieben von: / Posted by: Dann Corbit at 04 May 2000 08:42:13:
Als Antwort auf: / As an answer to: Re: Battle of the Crowns -- calibration runs (early indications courtesy of Elostat) geschrieben von: / posted by: Jon Dart at 04 May 2000 06:08:20:

If you add all the French variations then it's quite significant. I think it's a good choice for black even though things might get a little closed. Thanks for the info.
Best wishes...
Mogens

I don't think the French in general is bad, but I do think there
are some variations of it that are just bad for Black. E.g.
e4 e6 d4 d5 Nc3 Bb4 e5 b6 is one.
e4 e6 d4 d5 Nc3 Bb4 e5 c5 a3 Ba5 is another - GMs have used
this as Black but computers do not handle it well. I recently put
these lines into "avoid" status in Arasan's book - but they
are still in the 5.3 version that is used in these tests.

When the contest begins in earnest, I plan to solicit input from all of the authors for optimal setup of their programs.
You can supply any book you like for the contest with a caveat that it must be available to the general public also.

my FTP site

[Aaron]

Geschrieben von: / Posted by: Aaron at 04 May 2000 10:18:45:
Als Antwort auf: / As an answer to: Re: Battle of the Crowns -- calibration runs (early indications courtesy of Elostat) geschrieben von: / posted by: Dann Corbit at 04 May 2000 02:16:11:

Program Elo + - Games Score Av.Op. Draws
1 Crafty : 2814 142 239 16 84.4 % 2521 18.8 %
2 SOS : 2767 73 109 57 69.3 % 2625 15.8 %
3 Comet : 2728 64 75 86 48.8 % 2736 18.6 %
4 LGoliath : 2722 106 118 36 47.2 % 2742 16.7 %
5 Dragon : 2620 320 378 3 83.3 % 2340 33.3 %
6 TCBishop : 2564 109 118 33 43.9 % 2606 21.2 %
7 GnuChess5 : 2553 130 395 8 93.8 % 2083 12.5 %
8 AnMon : 2481 60 74 96 62.5 % 2392 16.7 %
9 Amy : 2474 56 68 112 61.6 % 2392 16.1 %
10 Arasan : 2378 53 46 167 51.8 % 2366 19.8 %
11 Bringer : 2357 66 60 101 54.5 % 2326 21.8 %
12 Ant : 2342 52 47 160 54.1 % 2314 21.9 %
13 EXchess : 2329 81 111 39 46.2 % 2356 35.9 %
14 ZChess : 2326 42 76 143 73.4 % 2150 16.8 %
15 Knightx : 2233 364 273 8 18.8 % 2488 12.5 %
16 Sjeng : 2165 252 192 12 25.0 % 2356 16.7 %
17 LDBlanche : 2124 153 86 36 20.8 % 2356 19.4 %
18 Amyan : 2069 53 50 175 50.3 % 2067 5.7 %
19 SSEChess : 2028 315 118 20 12.5 % 2366 5.0 %
20 Averno : 1891 165 64 59 20.3 % 2128 3.4 %
21 DChess : 1740 0 0 10 0.0 % 2340 0.0 %
22 Skaki : 1641 223 166 12 58.3 % 1583 33.3 %
23 TSCP : 1569 254 45 86 5.8 % 2053 4.7 %
24 Monik : 1304 0 0 5 0.0 % 1904 0.0 %
25 Pierre : 1304 0 0 6 0.0 % 1904 0.0 %

This rating list looked quite screwed to me at first sight , Dann .
Then I took a closer look : Gnu is 2553 ( +130 / -395 ) )
Interesting also how high some of the really provisional ratings are ie >>crafty

Nothing surprising about it really. Consider how few the number of games is.
If you look at the deltas (as you mention) they are rather huge.

Dragon is even worse..Only 3 games? Interesting how certain engines have so many games while others have relatively few..
Comet looks quite impressive. Despite a respectable number of games it still has a high rating..Good results against Crafty?
SOS hasn't played Crafty much? It's average Opposition elo is low compared to the others..

Amy,Anmon Arsan,Amyan,Ant etc also have played quite a few..Hmm..Are you testing the programs starting with "A" first? Then again Zchess and TSCP have played a lot as well..
How many games each are you planning to run each?

[Dann Corbit]

Geschrieben von: / Posted by: Dann Corbit at 04 May 2000 19:17:46:
Als Antwort auf: / As an answer to: Re: Battle of the Crowns -- calibration runs (early indications courtesy of Elostat) geschrieben von: / posted by: Aaron at 04 May 2000 10:18:45:

Program Elo + - Games Score Av.Op. Draws
1 Crafty : 2814 142 239 16 84.4 % 2521 18.8 %
2 SOS : 2767 73 109 57 69.3 % 2625 15.8 %
3 Comet : 2728 64 75 86 48.8 % 2736 18.6 %
4 LGoliath : 2722 106 118 36 47.2 % 2742 16.7 %
5 Dragon : 2620 320 378 3 83.3 % 2340 33.3 %
6 TCBishop : 2564 109 118 33 43.9 % 2606 21.2 %
7 GnuChess5 : 2553 130 395 8 93.8 % 2083 12.5 %
8 AnMon : 2481 60 74 96 62.5 % 2392 16.7 %
9 Amy : 2474 56 68 112 61.6 % 2392 16.1 %
10 Arasan : 2378 53 46 167 51.8 % 2366 19.8 %
11 Bringer : 2357 66 60 101 54.5 % 2326 21.8 %
12 Ant : 2342 52 47 160 54.1 % 2314 21.9 %
13 EXchess : 2329 81 111 39 46.2 % 2356 35.9 %
14 ZChess : 2326 42 76 143 73.4 % 2150 16.8 %
15 Knightx : 2233 364 273 8 18.8 % 2488 12.5 %
16 Sjeng : 2165 252 192 12 25.0 % 2356 16.7 %
17 LDBlanche : 2124 153 86 36 20.8 % 2356 19.4 %
18 Amyan : 2069 53 50 175 50.3 % 2067 5.7 %
19 SSEChess : 2028 315 118 20 12.5 % 2366 5.0 %
20 Averno : 1891 165 64 59 20.3 % 2128 3.4 %
21 DChess : 1740 0 0 10 0.0 % 2340 0.0 %
22 Skaki : 1641 223 166 12 58.3 % 1583 33.3 %
23 TSCP : 1569 254 45 86 5.8 % 2053 4.7 %
24 Monik : 1304 0 0 5 0.0 % 1904 0.0 %
25 Pierre : 1304 0 0 6 0.0 % 1904 0.0 %

This rating list looked quite screwed to me at first sight , Dann .
Then I took a closer look : Gnu is 2553 ( +130 / -395 ) )
Interesting also how high some of the really provisional ratings are ie >>crafty

Nothing surprising about it really. Consider how few the number of games is.
If you look at the deltas (as you mention) they are rather huge.

Dragon is even worse..Only 3 games? Interesting how certain engines have so many games while others have relatively few..
Comet looks quite impressive. Despite a respectable number of games it still has a high rating..Good results against Crafty?
SOS hasn't played Crafty much? It's average Opposition elo is low compared to the others..
Amy,Anmon Arsan,Amyan,Ant etc also have played quite a few..Hmm..Are you testing the programs starting with "A" first? Then again Zchess and TSCP have played a lot as well..
How many games each are you planning to run each?

To some degree it is rather arbitrary so far.
Don't remember who.
I have played very few crafty games period.
I am sort of going along alphabetically. But as the early letters get lots of opponents, some of the later ones get games with them.
I'll have a couple thousand games or more when I am done. I also intent to incorporate the work of others, using slow time control games to give additional input to the quality of the assessments. Once I have them approximately ranked, the contest will begin in earnest. So far the only purpose of the tests is to get a real grasp on approximately how strong the programs are without resorting to guesswork.

My FTP site

[Mogens Larsen]

Geschrieben von: / Posted by: Mogens Larsen at 05 May 2000 08:21:37:
Als Antwort auf: / As an answer to: Re: Battle of the Crowns -- calibration runs (early indications courtesy of Elostat) geschrieben von: / posted by: Jon Dart at 04 May 2000 06:08:20:

If you add all the French variations then it's quite significant. I think it's a good choice for black even though things might get a little closed. Thanks for the info.
Best wishes...
Mogens

I don't think the French in general is bad, but I do think there
are some variations of it that are just bad for Black. E.g.
e4 e6 d4 d5 Nc3 Bb4 e5 b6 is one.
e4 e6 d4 d5 Nc3 Bb4 e5 c5 a3 Ba5 is another - GMs have used
this as Black but computers do not handle it well. I recently put
these lines into "avoid" status in Arasan's book - but they
are still in the 5.3 version that is used in these tests.
--Jon

If you have a new or altered book that you won't mind sharing then please let me know.
Sincerely,
Mogens

[Aaron]

Geschrieben von: / Posted by: Aaron at 05 May 2000 14:01:50:
Als Antwort auf: / As an answer to: Re: Battle of the Crowns -- calibration runs (early indications courtesy of Elostat) geschrieben von: / posted by: Dann Corbit at 04 May 2000 19:17:46:

Program Elo + - Games Score Av.Op. Draws
1 Crafty : 2814 142 239 16 84.4 % 2521 18.8 %
2 SOS : 2767 73 109 57 69.3 % 2625 15.8 %
3 Comet : 2728 64 75 86 48.8 % 2736 18.6 %
4 LGoliath : 2722 106 118 36 47.2 % 2742 16.7 %
5 Dragon : 2620 320 378 3 83.3 % 2340 33.3 %
6 TCBishop : 2564 109 118 33 43.9 % 2606 21.2 %
7 GnuChess5 : 2553 130 395 8 93.8 % 2083 12.5 %
8 AnMon : 2481 60 74 96 62.5 % 2392 16.7 %
9 Amy : 2474 56 68 112 61.6 % 2392 16.1 %
10 Arasan : 2378 53 46 167 51.8 % 2366 19.8 %
11 Bringer : 2357 66 60 101 54.5 % 2326 21.8 %
12 Ant : 2342 52 47 160 54.1 % 2314 21.9 %
13 EXchess : 2329 81 111 39 46.2 % 2356 35.9 %
14 ZChess : 2326 42 76 143 73.4 % 2150 16.8 %
15 Knightx : 2233 364 273 8 18.8 % 2488 12.5 %
16 Sjeng : 2165 252 192 12 25.0 % 2356 16.7 %
17 LDBlanche : 2124 153 86 36 20.8 % 2356 19.4 %
18 Amyan : 2069 53 50 175 50.3 % 2067 5.7 %
19 SSEChess : 2028 315 118 20 12.5 % 2366 5.0 %
20 Averno : 1891 165 64 59 20.3 % 2128 3.4 %
21 DChess : 1740 0 0 10 0.0 % 2340 0.0 %
22 Skaki : 1641 223 166 12 58.3 % 1583 33.3 %
23 TSCP : 1569 254 45 86 5.8 % 2053 4.7 %
24 Monik : 1304 0 0 5 0.0 % 1904 0.0 %
25 Pierre : 1304 0 0 6 0.0 % 1904 0.0 %

I'll have a couple thousand games or more when I am done. I also intent to >incorporate the work of others, using slow time control games to give >additional input to the quality of the assessments.

For a moment I thought that you were going to be like that Swedish group which rates the commericals..
Why can't we do something like that? The problem i see is that people tend to have different computer setups, and not everyone has 2 computers to autoplay..
Still it could work..Combining efforts to create a combined rating list for winboard programs instead of everyone coming up with various ratings lists..
Has this being tried before?

Oh.Are those games above played with ponder on , 2 computers?

[Dann Corbit]

Geschrieben von: / Posted by: Dann Corbit at 05 May 2000 18:33:21:
Als Antwort auf: / As an answer to: Re: Battle of the Crowns -- calibration runs (early indications courtesy of Elostat) geschrieben von: / posted by: Aaron at 05 May 2000 14:01:50:
[snip]

For a moment I thought that you were going to be like that Swedish group which rates the commericals..
Why can't we do something like that? The problem i see is that people tend to have different computer setups, and not everyone has 2 computers to autoplay..
Still it could work..Combining efforts to create a combined rating list for winboard programs instead of everyone coming up with various ratings lists..
Has this being tried before?

Oh.Are those games above played with ponder on , 2 computers?

No. That's too ambitious for me.
As long as the configuration is specified carefully and the setup properly controlled, spelled out, and tested, I don't think that is a problem. It is a resource issue.
No. They are ponder-off, one computer. In the actual contest, I will probably switch to 2 computers and 40/2. The entire gist of the experiment is to find some sort of handle on the true strength of the program. Unfortunately, if I do change time control, I am making an unwise extrapolation.


My FTP site