Go back
the best chess engine is rated 3100 elo !

the best chess engine is rated 3100 elo !

Only Chess

Vote Up
Vote Down

Originally posted by DeepThought
I left Crafty on all night and it says after 37,900 seconds that it evaluate the position at +5.57 at a ply depth of 20, now what this means is that there is no way of forcing checkmate after 20 moves.
20 ply is ten moves. A ply is a half-move.

Vote Up
Vote Down

Originally posted by Wulebgr
Okay. I think I'll order Botvinnik's book, as it seems interesting. Alas, as I cannot read Russian, so it will be the Springer Verlag edition.

Botvinnik's position is a good test position, as most engines initially see equality. Postmodernist 10.10 was still at +0.34 after three minutes, for example.

Forced mate in 17 (or there about) is beyond the hor ...[text shortened]... ware, although several pruning algorithms in common use today should bring them fairly close.
You can't go wrong buying this book, but a word of caution:
buy it if and only if you have the math foundations to understand it (it is way to dense) and if you are really interested in computer chess algorithms.
If not, don't waste your money. It may be useful only for computer programmers.

Cheers

- J

Vote Up
Vote Down

Originally posted by HFRorbis
look at this list :
http://computerchess.org.uk/ccrl/4040/

Rybka 2.1 which has the highest rating is rated 3100 elo
When I first saw this rating I couldn't believe it.
It is incredible that a chess engine overcomes 3000 elo
At last chess engines have become better than the best chess players
Could Kasparov beat it ? Don't think so
I already mentioned this subject a month ago.

Vote Up
Vote Down

Rybka or Rykba or however its called is not 3100. That is some marketing gimmick. How many world championships has it won?

Vote Up
Vote Down

Originally posted by z00t
Rybka or Rykba or however its called is not 3100. That is some marketing gimmick. How many world championships has it won?
Just because Rybka has neither won any championship nor beaten any GMIs doesn't mean that it is not rated 3100
look at this list,first Rybka = 3012 (40 moves/2 hours tournament)

1 Rybka 2.1c 64 2CPU 3012 34 33 300 74.5 % 2826 35.0 %
2 Rybka 1.2f 64-bit 2951 30 29 350 70.6 % 2799 38.3 %
3 Zap!Chess Paderborn 64 2CPU 2870 28 28 350 53.3 % 2847 42.6 %
4 Deep Shredder 10 x64 2CPU 2854 27 27 373 49.5 % 2858 41.0 %
5 Deep Junior 10 2CPU 2824 32 32 300 50.3 % 2821 33.3 %
6 Spike 1.2 Turin 2CPU 2822 28 28 350 50.3 % 2820 40.6 %
7 Toga II 1.2.1 2816 25 25 450 46.9 % 2838 40.4 %
8 Fruit 2.2 2777 28 29 356 40.9 % 2842 38.5 %
9 Hiarcs 10 2770 28 28 350 44.3 % 2810 42.9 %
10 Spike 1.2 Turin 2761 31 31 300 48.0 % 2775 40.0 %
11 Hiarcs X50 UCI 2759 25 25 450 43.1 % 2807 40.0 %
12 Fritz 9 2756 29 29 350 43.1 % 2804 38.3 %
13 Ktulu 8 2748 30 30 333 45.8 % 2778 37.5 %
14 Glaurung 1.2 2CPU 2741 31 31 300 44.8 % 2777 38.3 %
link here :http://www.husvankempen.de/nunn/40_120_ratinglist/ratinglist/rangliste.html

1 edit
Vote Up
Vote Down

Originally posted by obliteration
I already mentioned this subject a month ago.
link please

Vote Up
Vote Down

Originally posted by HFRorbis
Just because Rybka has neither won any championship nor beaten any GMIs doesn't mean that it is not rated 3100
look at this list,first Rybka = 3012 (40 moves/2 hours tournament)

1 Rybka 2.1c 64 2CPU 3012 34 33 300 74.5 % 2826 35.0 %
2 Rybka 1.2f 64-bit 2951 30 29 350 70.6 % 2799 38.3 %
3 Zap!Chess Paderborn 64 2CPU 2870 28 28 350 53.3 % 2847 42.6 % ...[text shortened]... 8.3 %
link here :http://www.husvankempen.de/nunn/40_120_ratinglist/ratinglist/rangliste.html
That's an isolated list containing only 14 engines. While yes Rybka can claim to have an ELO of 3100 that is meaningless to anyone who knows what ELO means.

Vote Up
Vote Down

Originally posted by XanthosNZ
That's an isolated list containing only 14 engines. While yes Rybka can claim to have an ELO of 3100 that is meaningless to anyone who knows what ELO means.
Back in 1996 (I think) the number two rated player in the world was actually not very strong - he rigged the system and beat many, many lower rated players. His story is insane (He killed somebody...) if anybody can remember his name.

2 edits
Vote Up
Vote Down

Originally posted by cmsMaster
Back in 1996 (I think) the number two rated player in the world was actually not very strong - he rigged the system and beat many, many lower rated players. His story is insane (He killed somebody...) if anybody can remember his name.
Claude Bloodgood had a high USCF rating, and I think a respectable correspondence rating, but never had a FIDE rating (and, thus, was not ranked in the world)

You can see his games, and his master rating at http://www.chessgames.com/perl/chessplayer?pid=72632

He did achieve a 2702 USCF rating, 2nd in U.S. behind Gata Kamsky, and qualified for the U.S. Championship (until they modified the rules).

Vote Up
Vote Down

I have no doubt that computers in the near future will easily beat anyone in the world (if they haven't already). The speed and efficiency of computers is increasing exponentially and there is no limit in sight. The human brain can't improve much, so it was simply a matter of time. Chess is a mathematical problem in the end; this is what computers do best. This is not to say chess is doomed. Cars can easily outspeed humans, yet we still have foot races, horse races, etc. Far more problematical is the effect of computers on human chess. They have all but assumed a solo role of preparation. Databases stored in computers have augmented this role. I recently read in an interview of John Nunn whether he was going to write another MCO and he said there wasn't much need anymore because even amateurs use databases in opening preparation. Also, heard Short say about an opening not used by grandmasters anymore because it was "all fritzed out." Recently two players in a U.S. open in Philadelphia were caught using ( i should say alleged) computers. This was a cash prize situation. It isn't whether computers are better than the best humans it's whether we can separate the two so that we can have a fair contest in chess.

1 edit
Vote Up
Vote Down

Originally posted by HFRorbis
Just because Rybka has neither won any championship nor beaten any GMIs doesn't mean that it is not rated 3100
look at this list,first Rybka = 3012 (40 moves/2 hours tournament)
......

link here :http://www.husvankempen.de/nunn/40_120_ratinglist/ratinglist/rangliste.html
I do not recognise husvankempen.de. Tell them to bring rybka to the next world computer championships where it can play under tournament conditions. The likes of shredder/hydra/juniour are genuinely strong programs because of their feats against other top human/computer opponents.

I can organise a mickey mouse competition and claim some unknown upstart as "champ" but where is the recognition from the chess community.
If the programmers of rykba were serious, they would go about achieving results in open tounament conditions.

Vote Up
Vote Down

Originally posted by z00t
I do not recognise husvankempen.de.
The leading rating lists for chess engines are http://web.telia.com/~u85924109/ssdf/ and http://wbec-ridderkerk.nl/html/BayesianElo_ed12.htm

Rybka tops both.


I agree that it needs more testing against top GMs.

Vote Up
Vote Down

Originally posted by Marinkatomb
A commercial version is due for release...
Where did you read that? It is not on Hydras website.

Vote Up
Vote Down

Originally posted by z00t
Where did you read that? It is not on Hydras website.
Its already available on www.rybkachess.com😀

Vote Up
Vote Down

Who the hell cares!! Surely it's better to concentrate on your own game and try to seek what improvement you can rather than worrying about super GM'S and computers that calculate moves with such vast vision that we're never likely to understand it throughout the duration of our pathetic lowly rated little lives.

Cookies help us deliver our Services. By using our Services or clicking I agree, you agree to our use of cookies. Learn More.