|
< Earlier Kibitzing · PAGE 25 OF 79 ·
Later Kibitzing> |
| Jun-12-06 | | technical draw: How many here know that the toothbrush was invented in Pensacola? |
|
| Jun-12-06 | | Dionyseus: CEGT has updated their 40/4 blitz rating list, and Rybka 2.0 with 2 cpu is destroying everything and broke the 3000 barrier, 53 elo higher than Rybka 1.2f!
http://www.husvankempen.de/nunn/40_... |
|
| Jun-13-06 | | RandomVisitor: 1 Rybka 2.0 Beta x64 2CPU 3002
2 Rybka 1.2f 64-bit 2949
3 Rybka 1.1 64-bit 2941
4 Rybka 1.01 Beta 13x 64-bit 2903
5 Rybka 1.2f 32-bit 2902
6 Rybka 1.0 Beta 64-bit 2897
7 Rybka 1.1 32-bit 2883
8 Deep Shredder 10 x64 2CPU 2873
9 Rybka 1.01 Beta 13x 32-bit 2847
10 Zap!Chess Paderborn x64 2CPU 2838
11 Hiarcs X50 UCI 2831
12 Deep Fritz 7 2CPU 2822
13 Rybka 1.0 Beta 32-bit 2820
14 Toga II 1.2 Beta2a e26 2817 |
|
| Jun-13-06 | | notyetagm: <Dionyseus> I really like that CEGT site. Thanks for pointing it out. |
|
| Jun-13-06 | | notyetagm: <RV> Damn! The top 7(!) slots are all Rybkas. Holy criminy! |
|
| Jun-13-06 | | blingice: <RandomVisitor> Do you know how these tests are conducted? I'm not one of the dubious ones, but <cu8sfan> is helping me to send an email to http://chessengines.vs243093.vserve... to see how they conduct their tests. |
|
| Jun-13-06 | | RandomVisitor: Conditions
Time control and hash:
CEGT Games are medium time control games 40/40 repeated and blitz time control games 40/4 repeated. The meaning of 40/40 is 40 moves in 40 minutes and another 40 minutes for moves 41 to 80 and so on. For CEGT Blitz 40/4 meaning 40 moves in 4 minutes repeated. Given the different hardware from testers we agreed to adapt to 2 Ghz Pentium CPU. Some examples: for a machine Athlon64 3500+ this comes down to 40 moves in 18 minutes. A tester with Pentium 800 Mhz has to give full two hours for every 40 moves.
Hash given is usually 256 MB for each engine. Very few testers who have less RAM available are allowed to give 128 MB.
Deep versions: Deep Shredder 9. Deep Fritz 8, Deep Junior 9 and others are tested on dual machines using 2 CPU´s and 512 MB hash. There is an exception for Junior 9.003 using only 256 MB, because there seem to occur bugs when giving 512 MB to this one. Books:
In the first months of CEGT all Nunn Suite 1 and 2 positons were used and also many from Noomen Select. Currently we use mainly books like 8move.ctg. remis.ctg, Perfect books, Powerbooks, Master Elect and Arena books mainly by Harry Schnapp. We have started now to use in a bigger extend a testsuite with 220 positions by Harry Schnapp. Thanks to Harry for this one! Tablebases:
Most Testers use 5 men EGTB. Some use only 4 men. Testers using 5 men give 32 MB EGTB hash. Testers using 4 men give 16 MB EGTB hash. GUI´s:
All testers use one or more different GUI´s. Most used are Shredder 9 GUI, Arena and Shredder Classic GUI. Chess Partner GUI and Winboard can also be used. Not used are buggy GUI´s like Fritz 9, Fritz 8 with server update, known buggy UCI.dll´s. Adjudications:
Testers and GUI´s are allowed to adjudicate totally won or drawn games Benching:
to adapt the different hardware of testers in CEGT to a standard (currently 40/40 and 40/4 on P4 2Ghz reference machine from Uschi) we perform a benchmark with a Bryan Hofmann Crafty compile. Bryan Hofmann and Johan Havegheer also calculated
the according table for time controls to use with different CPU´s. Charles sends both (compile and table) out to new testers. Just put the exe in an empty folder (do not include a Crafty.rc or a book) and make sure that you only have the necessary tasks running in the background - best perform a reboot beforehand.
Then click the exe and type in bench at the command line. Wait for around 40 to 120 seconds. A logfile will be created in the folder. Amongst other values at the bottom you will find the seconds needed and can just compare the time you have to give
with your CPU in order to adapt. This is a repeated time control! In ChessBase GUI´s for example you can give 0 (zero) for all values for second and third time control and this way the first time control like for example 40/24 is always repeated.
For Blitz 40/4 just divide by ten. |
|
| Jun-13-06 | | RandomVisitor: There appears to be some kind of Benchmark testing that goes on in CEGT testing so that all computers appear to run at 40/40 time control on a 2GHz machine. A slower machine would run at 40moves/60minutes, say, in order to simulate a faster machine running at 40/40. |
|
| Jun-13-06 | | blingice: <RV> So does the testing seem consistant with the optimal performance and continuity of the experiment that is so desired? |
|
| Jun-13-06 | | RandomVisitor: <blingice>say what? |
|
| Jun-13-06 | | notyetagm: <RV> LOL! Sounds like <LMAJ> gibberish. :-) |
|
| Jun-13-06 | | Dionyseus: <blingice> Yes |
|
| Jun-13-06 | | notyetagm: <RV> You're a member now! Great. Welcome, my green scaley friend. :-)
|
|
| Jun-13-06 | | cotdt: i bet Rybka won't survive in a correspondence chess match against nickel arno! |
|
| Jun-13-06 | | blingice: <RV> Hahaha, as I previewed the message, I considered clarifying it, but I thought it would be funnier to leave it like that. Clarified, I meant do these tests (to you) seem conducted in a manner to maintain the best performance of the individual engines? Also, are these tests conducted in a consistent manner (all variables equal except for engines)? Read what I wrote again, you'll probably understand it perfectly fine now :) |
|
| Jun-13-06 | | RandomVisitor: <blingice>I would say that 40/40 testing is only part of the answer. 40/120 testing (now underway) would answer the question of how the engine would perform under tournament conditions. It is said that Rybka performs better than rated against the weaker engines, and less than rated against the top competing engines, such as Shredder, Junior, Zap!. I think that CEGT has done a great job trying to make the tests reflect the best efforts of each engine. |
|
| Jun-14-06 | | Rybka: Hi , does any one here knows that is SSDF is using the 64 bit versions of rybka or not and why there is no multiprocessor version of any engine in the SSDF list ?.
i think because they use much old hardware like 1200 athlon.
I think with 64 bit version rybka has almost 40 points higher than 32 bit version , so this SSDF list is not showing the proper strength of The RYBKA engine. |
|
| Jun-14-06 | | Rybka: < cotdt>:< i bet Rybka won't survive in a correspondence chess match against nickel arno! >. Yes Rybka won't survive if nickel uses rybka 2 as a assistance engine for his correspondece match , which i think he should definately use because in such a match how could u don't use an approx 3000 elo (rybka) engines help .
If any one reads the article of arno nickel on the chess base its clear that this man arno nickel is very biase . it seems that chess base is paying him for such articles , in his articles he tries as much as he can to ignore the rybka's strength and continue to insist that fritz and shreeder are better than rybka. and he also said that when shredder 10 will come in the market then no one on the playchess server is uses the rybka engine, which i think that has proved wrong now because 90 percent people on play chess server are still using rybka engine .
One thing about chessbase , i am reading their news daily from almost three years and they are very punctual in praising their engines when ever SSDF list comes and their engines top the ssdf list but in feb this year when fruit topped the ssdf list chess base has not write any article about the SSDF ratings and again now when rybka topped the list by recored 75 points they agian have not mentioned any thing about it.
Infact they ridiculisly saying the Deepfritz 9 the strongest engine.which is 150+ points below than rybka 2 how ridiculous is that. |
|
| Jun-14-06 | | gus inn: good points <Rybka> .E.g. a friend of mine is a betatester of Rybka.And here
Rybka is beating Shredder and Junior , big time ! |
|
| Jun-14-06 | | blingice: Ok, when Arno Nickel plays against a computer, why is he allowed to use a computer?! And how do you define "assistance"? If I played in a computer-assisting tournament, it wouldn't be assisting me, it'd be playing for me! Couldn't Nickel just turn on Rybka (or whatever computer he's using) and leave it until he has to move? I don't think I have as much respect for Nickel's accomplishments, as he isn't winning purely in his own skill. |
|
| Jun-14-06 | | RandomVisitor: <blingice><I don't think I have as much respect for Nickel's accomplishments, as he isn't winning purely in his own skill.>The skill comes in deciding on a move to play when there are several available that are all close in score. A machine would just select the highest-scored move, but there may be reasons to select a move that is not scored highest. By calculating deep variations on the possible best moves you can out-think just about any machine and at least draw under best play. This is what Arno Nickel does when he plays correspondence chess with computers, and I respect his skills in that area. |
|
| Jun-14-06 | | ughaibu: I doubt Nickel selects from moves chosen by the computer, he'll use the computer to check for unexpected difficulties involved in his own plans. |
|
| Jun-14-06 | | blingice: I suppose I'd have to know his procedure to really understand it. I just think it's awkward to use a computer to beat a computer and then say "I won". |
|
| Jun-14-06 | | cotdt: Arno Nickel just uses the computer for tactical calculations and blunder-check. If he followed the computer's moves, he wouldn't have won against Hydra, and indeed many of the moves that computers recommend are would have drawn in a won position, or perhaps even lost. |
|
Jun-14-06
 | | LIFE Master AJ: < <So now where are you <LIFE Master AJ> ? Rybka is 141 points higher than ur junior 9 . take it , that's the reality. Rybka = 2931
Junior= 2790 >
# 1.) Right here.
# 2.) Those ratings are BS.
(If there were any grounds to them, Rybka would have performed better at the WCC. Either Junior is under-rated, Rybka's rating is over-infalted ... or maybe it is just a little of both.) |
|
 |
|
|
< Earlier Kibitzing · PAGE 25 OF 79 ·
Later Kibitzing> |
Kibitzing
