SWCR after 54.560 games ...

[Frank Quisinsky]

Hi there,

yesterday a still running tourney ended. 4.960 games produced by 6 new engines.

Code: Select all

092. x64 Equinox 0.83              NEW      Started: 06.09.2010 private status
091  x64 WB Crafty 23.3 JA NP      TEST     Started: 06.09.2010 without ponder
090. x64 Cipollino 3.25            NEW      Started: 06.09.2010 private status
089. x64 Fire 1.31                 NEW      Started: 06.09.2010 IPP family
088. x64 Houdini 1.03a             NEW      Started: 06.09.2010 IPP family
087. x64 Loop M1-T                 Update   Started: 06.09.2010

I started yesterday a new SWCR-64 tourney with 4 new updates:

Code: Select all

095. x64 GullChess 1.0a            NEW      Started: 26.09.2010
094. x64 IvanHoe B52aC             NEW      Started: 26.09.2010 IPP family
093. x64 FireBird 1.1 Dr. Deeb I   NEW      Started: 26.09.2010 IPP family, setting
060. w32 Hiarcs 13.1               Again    Started: 26.09.2010 (for SWCR-64)

If you have interest you can follow in the next 12-13 days the LIVE table on my webpage:

Frank's Chess Page, SWCR
http://www.amateurschach.de

Game databases and all rating lists are updated yesterday.

Here is the actual SWCR with all played games:

Code: Select all

After 54.560 games, 30 different engines, 92 versions included
ELO are fixed to 2.800 for Shredder 12 32bit. All other engines are derived from that fixed number.
Calculated with:  Bayesian 0056

The separate columns give the (mean) Elo performance, the + and - margins of error given with 95 % confidence,
the number of finished games, the relative score given in percentages, the average opponent ELO and finally the
relative number of draws for each program.

01. IPP Houdini 1.03a x64        2.948 22 21  920 79% 2.718 29%
02. Rybka 4 x64                  2.939 18 18 1400 81% 2.690 28% +44 to 32bit
03. Stockfish 1.8.0 JA x64       2.906 20 20 1040 76% 2.701 31% +43 to 32bit
--. Rybka 3 x64                  2.905 23 22  840 78% 2.687 28% +45 to 32bit
--. IPP Fire 1.31 x64            2.900 21 20  920 75% 2.720 36%
--. Stockfish 1.7.1 JA x64       2.896 19 18 1120 76% 2.706 34% +21 to 32bit
--. Rybka 4 w32                  2.895 20 20  960 76% 2.703 32%
--. Stockfish 1.7.1 JA w32       2.875 18 18 1200 75% 2.690 31%
--. Stockfish 1.8.0 JA w32       2.863 20 20  920 74% 2.697 37%
--. Rybka 3 w32                  2.860 16 16 1520 74% 2.689 31%
04. Critter 0.80 x64             2.838 19 19 1040 68% 2.703 32% +24 to 32bit
05. Naum 4.2 x64                 2.832 15 15 1640 68% 2.700 35% + 8 to 32bit
--. Stockfish 1.6.3 JA w32       2.827 18 18 1080 71% 2.680 36%
--. Naum 4.2 w32                 2.824 16 16 1440 68% 2.701 36%
--. Critter 0.80 w32             2.814 21 20  800 66% 2.703 39%
--. Naum 4.1 w32                 2.811 20 19  920 68% 2.686 35%
--. Critter 0.70 x64             2.806 20 20  880 65% 2.701 38% +28 to 32bit
--. Stockfish 1.6.0 JA w32       2.802 19 19  960 68% 2.687 39%
06. Komodo 1.2 JA x64            2.800 16 16 1400 65% 2.694 39% +39 to 32bit
06. Shredder 12 w32              2.800 10 10 3600 64% 2.700 36%
--. Komodo 1.0 JA x64            2.791 20 20  840 64% 2.693 40% +75 to 32bit
--. Shredder 12 x64              2.788 18 18 1080 61% 2.707 35% -12 to 32bit
--. Naum 4.0 w32                 2.785 19 18  960 65% 2.682 38%
--. Critter 0.70 w32             2.778 19 19  920 61% 2.703 39%
08. Deep Fritz 12 w32            2.777 15 14 1520 61% 2.704 42% 32bit only
--. Komodo 1.2 JA w32            2.761 18 18  960 58% 2.709 42%
--. Fritz 12 w32                 2.746 16 16 1160 59% 2.688 44% 32bit only
09. Spark 0.5 x64                2.741 16 16 1400 57% 2.696 37% +27 to 32bit
10. Thinker 5.4d Inert x64       2.737 14 14 1640 55% 2.702 39% +26 to 32bit
11. Hiarcs 13.1 w32              2.735 18 18  960 54% 2.710 42% 32bit only
--. Stockfish 1.5.1 JA w32       2.731 19 19  840 59% 2.672 43%
12. Zappa Mexico II x64          2.723 14 14 1640 53% 2.702 40% +61 to 32bit
--. Spark 0.4 x64                2.717 20 20  840 53% 2.696 40% +31 to 32bit
--. Komodo 1.0 JA w32            2.716 16 16 1200 53% 2.695 40%
--. Spark 0.5 w32                2.714 18 18  960 51% 2.711 41%
13. Protector 1.3.4 JA x64       2.712 15 15 1600 52% 2.703 37% +33 to 32bit
--. Thinker 5.4d Inert w32       2.711 13 13 2000 51% 2.703 42%
14. Fruit 09_07_05 x64           2.706 15 15 1640 50% 2.703 34%
--. Critter 0.60 x64             2.700 20 20  840 50% 2.697 38% +12 to 32bit
--. Doch 1.3.4 JA w32            2.690 19 19  840 51% 2.687 44%
--. Critter 0.60 w32             2.688 19 19  920 50% 2.688 39%
15. Sjeng WC-2008 x64            2.687 14 15 1640 48% 2.703 35% +22 to 32bit
15. Hannibal 1.0a x64            2.687 18 18 1040 47% 2.709 35% +47 to 32bit
--. Spark 0.4 w32                2.686 19 19  880 49% 2.692 42%
17. Junior 11.2 w32              2.685 19 19  960 46% 2.712 33%
--. Protector 1.3.5 x64          2.684 20 20  840 47% 2.707 39% +14 to 32bit
--. Junior 11.2 x64              2.681 16 16 1400 48% 2.697 30% - 4 to 32bit
--. Protector 1.3.4 JA w32       2.679 16 16 1360 46% 2.707 39%
18. Onno 1.2.70 x64              2.677 16 16 1400 47% 2.698 38% +41 to 32bit
--. Cyclone xTreme Wrath w32     2.676 17 17 1080 47% 2.698 41% 32bit only
--. Protector 1.3.2 w32          2.674 16 16 1160 47% 2.695 41%
--. Onno 1.1.1 x64               2.671 20 20  840 46% 2.698 40% +40 to 32bit
--. Protector 1.3.5 JA w32       2.670 19 20  840 44% 2.713 42%
--. Junior 2010 w32              2.669 16 16 1240 47% 2.691 36%
--. Hiarcs 12.1 w32              2.667 19 19  880 47% 2.689 41% 32bit only
--. Protector 1.3.1b w32         2.667 19 19  840 47% 2.690 42%
--. Doch 1.2 JA w32              2.666 19 19  840 48% 2.679 40%
--. Sjeng WC-2008 w32            2.665 13 13 2000 44% 2.704 37%
--. Hiarcs 12.1 Sharpen PV w32   2.665 16 16 1280 45% 2.699 39% 32bit only
--. Zappa Mexico II w32          2.662 13 13 2000 44% 2.704 42%
--. Spark 0.3a w32               2.656 17 17 1120 44% 2.698 41%
--. Doch 09.980 JA w32           2.652 19 19  840 46% 2.676 42%
--. Junior 11.1a x64             2.648 20 20  840 43% 2.700 32% + 3 to 32bit
--. Junior 11.1a w32             2.645 18 18  960 44% 2.691 36%
--. Spark 0.3 w32                2.643 19 19  880 43% 2.691 42%
--. Hannibal 1.0a w32            2.640 20 20  840 39% 2.715 36%
--. Onno 1.2.70 w32              2.636 18 19  960 39% 2.714 40%
--. Onno 1.1.1 w32               2.631 15 14 1520 40% 2.695 41%
19. Loop 2007 x64                2.630 15 16 1440 40% 2.703 37% +16 to 32bit
--. Loop M1-T x64                2.628 19 20  920 36% 2.731 36%
--. Fruit 05/11/03 w32           2.617 13 13 2000 37% 2.705 41% 32bit only
--. Loop 13.6 w32                2.615 15 15 1520 38% 2.696 39%
--. Loop 2007 w32                2.614 19 19  960 35% 2.715 36%
--. Critter 0.52b w32            2.614 18 18 1040 38% 2.699 37%
--. Twisted Logic 20100131x x64  2.611 18 18 1120 35% 2.716 32% +37 to 32bit
--. Glaurung 2.2 JA w32          2.610 17 18 1080 37% 2.700 36%
20. Ktulu 9.03 w32               2.609 15 15 1520 36% 2.708 30% 32bit only
21. Umko 1.0 x64                 2.607 18 18 1040 35% 2.712 37% (ponder not possible)
22. SmarThink 1.20 x64           2.602 15 15 1640 36% 2.705 34% + 5 to 32bit
23. Equinox 0.83 x64             2.598 19 20  960 32% 2.728 33%
--. SmarThink 1.20 w32           2.597 13 13 2000 34% 2.705 37%
24. Crafty 23.3 JA x64           2.592 18 19 1040 34% 2.713 35%
--. Twisted Logic 20100131x w32  2.574 15 15 1600 32% 2.707 30%
25. Spike 1.2 Turin w32          2.574 16 16 1480 31% 2.708 34% 32bit only
--. Cipollino 3.25 x64           2.565 20 21  920 28% 2.734 30%
26. BugChess2 1.7 x64            2.563 21 22  800 29% 2.719 33%
27. Scorpio 2.6 JA x64           2.555 18 19 1120 28% 2.718 32%
--. Crafty 23.2 JA x64           2.555 18 19 1120 28% 2.718 30%
28. Chronos 1.99 x64             2.552 18 18 1120 27% 2.718 33% (ponder not possible)
--. Crafty 23.3 JA x64 NP        2.546 21 21  920 25% 2.735 30% (Test: Without ponder)
29. Daydreamer 1.75 JA x64       2.521 19 19 1120 24% 2.719 30%
30. Tornado 3.6.7 x64            2.479 24 24  800 19% 2.723 24%

Date range            = 2009.10.26 - 2010.09.26
Average White ELO     = 2.702,21
Average Black ELO     = 2.702,21
Maximum ELO distance  = 460
Average ELO distance  = 116.707
Number of White wins  = 19859 (36.4%)
Number of draws       = 19971 (36.6%)
Number of Black wins  = 14730 (27.0%)
White score           = 54.7%
Black score           = 45.3%
ECO A games           = 15628 (28.6%)
ECO B games           = 11351 (20.8%)
ECO C games           =  7488 (13.7%)
ECO D games           = 12925 (23.7%)
ECO E games           =  7168 (13.1%)

Code: Select all

Conditions:
Windows XP Prof. x64 Edition
SClassic 4, Fritz 11 GUI
S12: Random-Book 4.1
F12: Perfect 2010

40 moves in 10 minutes
ponder = on
each engine = 1 core
resign / learning = off
hashtables = 256Mb
4TB Nalimov, GaviotaB., egbb,
ShredderB., TotalB. & TripleB.

Intel® Core™2 Quad
4x Q9550, 2,83GHz

Have fun with SWCR!

Best
Frank

[Frank Quisinsky]

Hi there,

here are the results of the 11.040 SWCR-64 games tourney:
Interesting is the Crafty 23.3 JA x64 ponder test. Two Crafty versions played. One version without ponder in a ponder = on toruney.

Result: 44 ELO different !!
With other words ... if Umko 1.0 x64 can play with ponder the ELO could be around 2.650 !!

Code: Select all

After 11.040 games, 24 engines included
ELO are fixed to 2.800 for Shredder 12 32bit. All other engines are derived from that fixed number.
Calculated with:  Bayesian 0056

The separate columns give the (mean) Elo performance, the + and - margins of error given with 95 % confidence,
the number of finished games, the relative score given in percentages, the average opponent ELO and finally the
relative number of draws for each program.

01. IPP Houdini 1.03a x64        2.945 21 21  920 79% 2.715 29%
02. Rybka 4 x64                  2.936 21 20  920 79% 2.716 31%
03. Stockfish 1.8.0 JA x64       2.902 20 20  920 74% 2.717 33%
04. IPP Fire 1.31 x64            2.898 20 20  920 75% 2.717 36%
05. Critter 0.80 x64             2.836 19 19  920 66% 2.720 33%
06. Naum 4.2 x64                 2.826 19 19  920 65% 2.721 38%
07. Shredder 12 w32              2.800 19 19  920 61% 2.722 36%
08. Komodo 1.2 JA x64            2.797 19 18  920 61% 2.722 41%
09. Spark 0.5 x64                2.738 18 18  920 52% 2.724 39%
10. Thinker 5.4d Inert x64       2.737 18 18  920 52% 2.724 39%
11. Zappa Mexico II x64          2.720 18 18  920 49% 2.725 41%
12. Protector 1.3.4 JA x64       2.713 18 18  920 49% 2.726 39%
13. Fruit 09_07_05 x64           2.698 18 19  920 46% 2.726 35%
14. Sjeng WC-2008 x64            2.691 19 19  920 45% 2.726 34%
15. Junior 11.2 x64              2.689 19 19  920 45% 2.727 30%
16. Hannibal 1.0a x64            2.684 18 19  920 44% 2.727 35%
17. Onno 1.2.70 x64              2.673 18 19  920 42% 2.727 39%
18. Loop M1-T x64                2.626 19 19  920 36% 2.729 36%
19. Umko 1.0 x64                 2.605 19 19  920 33% 2.730 36% (ponder not possible)
20. SmarThink 1.20 x64           2.595 19 19  920 32% 2.731 34%
21. Equinox 0.83 x64             2.594 19 20  920 32% 2.731 33%
22. Crafty 23.3 JA x64           2.588 19 20  920 31% 2.731 34%
23. Cipollino 3.25 x64           2.563 20 20  920 28% 2.732 30%
24. Crafty 23.3 JA x64 NP        2.544 20 20  920 25% 2.733 30% (Test: Without ponder)

Date range            = 2010.05.12 - 2010.09.26
Average White ELO     = 2.724,92
Average Black ELO     = 2.724,92 = Category 19
Maximum ELO distance  = 401
Average ELO distance  = 139.232
Number of White wins  =  3957 (35.8%)
Number of draws       =  3874 (35.1%)
Number of Black wins  =  3209 (29.1%)
White score           = 53.4%
Black score           = 46.6%
ECO A games           =  4397 (39.8%)
ECO B games           =  1939 (17.6%)
ECO C games           =  1506 (13.6%)
ECO D games           =  2306 (20.9%)
ECO E games           =   892 ( 8.1%)

Best
Frank

[beachknight]

Code: Select all

Date range            = 2010.05.12 - 2010.09.26
Average White ELO     = 2.724,92
Average Black ELO     = 2.724,92 = Category 19
Maximum ELO distance  = 401
Average ELO distance  = 139.232
Number of White wins  =  3957 (35.8%)
Number of draws       =  3874 (35.1%)
Number of Black wins  =  3209 (29.1%)
White score           = 53.4%
Black score           = 46.6%
ECO A games           =  4397 (39.8%)
ECO B games           =  1939 (17.6%)
ECO C games           =  1506 (13.6%)
ECO D games           =  2306 (20.9%)
ECO E games           =   892 ( 8.1%)

Hi Frank,

How do you output those stats with bayeselo 0056?

TIA,

Best,

[Dann Corbit]

Hi there,

here are the results of the 11.040 SWCR-64 games tourney:
Interesting is the Crafty 23.3 JA x64 ponder test. Two Crafty versions played. One version without ponder in a ponder = on toruney.

Result: 44 ELO different !!
With other words ... if Umko 1.0 x64 can play with ponder the ELO could be around 2.650 !!

That assumes that ponder is implemented as well as it is in crafty.
For instance, if the ponder hit rate is lower, then the benefit is also lower.
In general, though, you are approximately doubling the think time so the result is rather unsurprising. Still, it is an excellent idea to test it, and so thank you for that.

Code: Select all

After 11.040 games, 24 engines included
ELO are fixed to 2.800 for Shredder 12 32bit. All other engines are derived from that fixed number.
Calculated with:  Bayesian 0056

The separate columns give the (mean) Elo performance, the + and - margins of error given with 95 % confidence,
the number of finished games, the relative score given in percentages, the average opponent ELO and finally the
relative number of draws for each program.

01. IPP Houdini 1.03a x64        2.945 21 21  920 79% 2.715 29%
02. Rybka 4 x64                  2.936 21 20  920 79% 2.716 31%
03. Stockfish 1.8.0 JA x64       2.902 20 20  920 74% 2.717 33%
04. IPP Fire 1.31 x64            2.898 20 20  920 75% 2.717 36%
05. Critter 0.80 x64             2.836 19 19  920 66% 2.720 33%
06. Naum 4.2 x64                 2.826 19 19  920 65% 2.721 38%
07. Shredder 12 w32              2.800 19 19  920 61% 2.722 36%
08. Komodo 1.2 JA x64            2.797 19 18  920 61% 2.722 41%
09. Spark 0.5 x64                2.738 18 18  920 52% 2.724 39%
10. Thinker 5.4d Inert x64       2.737 18 18  920 52% 2.724 39%
11. Zappa Mexico II x64          2.720 18 18  920 49% 2.725 41%
12. Protector 1.3.4 JA x64       2.713 18 18  920 49% 2.726 39%
13. Fruit 09_07_05 x64           2.698 18 19  920 46% 2.726 35%
14. Sjeng WC-2008 x64            2.691 19 19  920 45% 2.726 34%
15. Junior 11.2 x64              2.689 19 19  920 45% 2.727 30%
16. Hannibal 1.0a x64            2.684 18 19  920 44% 2.727 35%
17. Onno 1.2.70 x64              2.673 18 19  920 42% 2.727 39%
18. Loop M1-T x64                2.626 19 19  920 36% 2.729 36%
19. Umko 1.0 x64                 2.605 19 19  920 33% 2.730 36% (ponder not possible)
20. SmarThink 1.20 x64           2.595 19 19  920 32% 2.731 34%
21. Equinox 0.83 x64             2.594 19 20  920 32% 2.731 33%
22. Crafty 23.3 JA x64           2.588 19 20  920 31% 2.731 34%
23. Cipollino 3.25 x64           2.563 20 20  920 28% 2.732 30%
24. Crafty 23.3 JA x64 NP        2.544 20 20  920 25% 2.733 30% (Test: Without ponder)

Date range            = 2010.05.12 - 2010.09.26
Average White ELO     = 2.724,92
Average Black ELO     = 2.724,92 = Category 19
Maximum ELO distance  = 401
Average ELO distance  = 139.232
Number of White wins  =  3957 (35.8%)
Number of draws       =  3874 (35.1%)
Number of Black wins  =  3209 (29.1%)
White score           = 53.4%
Black score           = 46.6%
ECO A games           =  4397 (39.8%)
ECO B games           =  1939 (17.6%)
ECO C games           =  1506 (13.6%)
ECO D games           =  2306 (20.9%)
ECO E games           =   892 ( 8.1%)

Best
Frank

[Frank Quisinsky]

Hi Harun,

easy ...
Created with "summary" by Norm Pollock.
Can be found in the PGN tool collection by Norm.

I used the important information from the output of summary only.

Norm created a fantastic EPD and PGN tool collection.
I like it so much ... great piece of software !!

Best
Frank

[Frank Quisinsky]

Hi Dann,

do you think I should make again such a test with a second engine. Only possible with WB engines in Shredder Classic 4 GUI tournaments. So I can make it with Thinker 5.4d Inert.

But later, so many things to do with testing.
At the moment I test different IPP Familie engines. All need a little bit time because 1 game in SWCR is running around 40 minutes with my conditions.

Best
Frank

[bob]

Hi there,

here are the results of the 11.040 SWCR-64 games tourney:
Interesting is the Crafty 23.3 JA x64 ponder test. Two Crafty versions played. One version without ponder in a ponder = on toruney.

Result: 44 ELO different !!
With other words ... if Umko 1.0 x64 can play with ponder the ELO could be around 2.650 !!

That assumes that ponder is implemented as well as it is in crafty.
For instance, if the ponder hit rate is lower, then the benefit is also lower.
In general, though, you are approximately doubling the think time so the result is rather unsurprising. Still, it is an excellent idea to test it, and so thank you for that.

I figure about 50% ponder hits. But about 1/2 of those will already be correctly predicted by your opponent. Which means you save no time. I think that roughly 50% time savings is closer to the truth. In ACM chess tournaments, we have often seen the "ponder hit" issue occur in streaks. You incorrectly predict, I do a search and play a different move than you expect. Your ponder time is wasted. When you get my surprise move, I am pondering while you search. When you make your move, there is a good chance I have an instant reply ready, and we repeat this for several moves. Then you play a surprise, I have to search and you correctly predict and ponder for a few moves, and then the cycle repeats. About 1/2 of the theoretical savings becomes realizable savings because of this.

As far as "the quality of pondering" goes, that is really not a "ponder" issue, but a quality of search/evaluation issue. A good program vs a bad program is going to see poor ponder hit rates because the opponent is often making sub-optimal moves while we are predicting good moves. About the only thing Crafty does that some do not is that there is _never_ a case where we are not pondering, unless we pondered and found a forced mate. If there is no predicted move to ponder, Crafty will find one, either thru the hash table or thru a search for the opponent's side. But it will ponder and not sit.

Code: Select all

After 11.040 games, 24 engines included
ELO are fixed to 2.800 for Shredder 12 32bit. All other engines are derived from that fixed number.
Calculated with:  Bayesian 0056

The separate columns give the (mean) Elo performance, the + and - margins of error given with 95 % confidence,
the number of finished games, the relative score given in percentages, the average opponent ELO and finally the
relative number of draws for each program.

01. IPP Houdini 1.03a x64        2.945 21 21  920 79% 2.715 29%
02. Rybka 4 x64                  2.936 21 20  920 79% 2.716 31%
03. Stockfish 1.8.0 JA x64       2.902 20 20  920 74% 2.717 33%
04. IPP Fire 1.31 x64            2.898 20 20  920 75% 2.717 36%
05. Critter 0.80 x64             2.836 19 19  920 66% 2.720 33%
06. Naum 4.2 x64                 2.826 19 19  920 65% 2.721 38%
07. Shredder 12 w32              2.800 19 19  920 61% 2.722 36%
08. Komodo 1.2 JA x64            2.797 19 18  920 61% 2.722 41%
09. Spark 0.5 x64                2.738 18 18  920 52% 2.724 39%
10. Thinker 5.4d Inert x64       2.737 18 18  920 52% 2.724 39%
11. Zappa Mexico II x64          2.720 18 18  920 49% 2.725 41%
12. Protector 1.3.4 JA x64       2.713 18 18  920 49% 2.726 39%
13. Fruit 09_07_05 x64           2.698 18 19  920 46% 2.726 35%
14. Sjeng WC-2008 x64            2.691 19 19  920 45% 2.726 34%
15. Junior 11.2 x64              2.689 19 19  920 45% 2.727 30%
16. Hannibal 1.0a x64            2.684 18 19  920 44% 2.727 35%
17. Onno 1.2.70 x64              2.673 18 19  920 42% 2.727 39%
18. Loop M1-T x64                2.626 19 19  920 36% 2.729 36%
19. Umko 1.0 x64                 2.605 19 19  920 33% 2.730 36% (ponder not possible)
20. SmarThink 1.20 x64           2.595 19 19  920 32% 2.731 34%
21. Equinox 0.83 x64             2.594 19 20  920 32% 2.731 33%
22. Crafty 23.3 JA x64           2.588 19 20  920 31% 2.731 34%
23. Cipollino 3.25 x64           2.563 20 20  920 28% 2.732 30%
24. Crafty 23.3 JA x64 NP        2.544 20 20  920 25% 2.733 30% (Test: Without ponder)

Date range            = 2010.05.12 - 2010.09.26
Average White ELO     = 2.724,92
Average Black ELO     = 2.724,92 = Category 19
Maximum ELO distance  = 401
Average ELO distance  = 139.232
Number of White wins  =  3957 (35.8%)
Number of draws       =  3874 (35.1%)
Number of Black wins  =  3209 (29.1%)
White score           = 53.4%
Black score           = 46.6%
ECO A games           =  4397 (39.8%)
ECO B games           =  1939 (17.6%)
ECO C games           =  1506 (13.6%)
ECO D games           =  2306 (20.9%)
ECO E games           =   892 ( 8.1%)

Best
Frank

[Dann Corbit]

Hi Dann,

do you think I should make again such a test with a second engine. Only possible with WB engines in Shredder Classic 4 GUI tournaments. So I can make it with Thinker 5.4d Inert.

But later, so many things to do with testing.
At the moment I test different IPP Familie engines. All need a little bit time because 1 game in SWCR is running around 40 minutes with my conditions.

Best
Frank

It would be interesting to see if the same trend holds for more than one engine, but whatever is interesting to you is what I would test if I were you.

[beachknight]

Hi Harun,

easy ...
Created with "summary" by Norm Pollock.
Can be found in the PGN tool collection by Norm.

I used the important information from the output of summary only.

Norm created a fantastic EPD and PGN tool collection.
I like it so much ... great piece of software !!

Best
Frank

Thanks, Frank.

I should have that collection of, then, 40 tools.

Will try summary.

Best,

[Frank Quisinsky]

Hi Dann,

thats right, but I have fun if I can visit the games on my PCs on my TV. All SWCR machines goes to my TV via HDMI. So I have for channels more

Honest, all what we need is a reality TV-Channel for computer chess via sattelite. This would be a nice new project for the computer chess community. Do you know a person which have enough and don't need here money?

A little bit statistic is for me interesting too, like that.
I think I should test a second engine without ponder in a ponder = on tourney. I will make it, if I have a little bit more time for it.

If you like you can add the SWCR databases in your game collection.

Best
Frank