G.M. MILOV VADIM versus Rybka 3 Playchess Top Challenger

Discussion of anything and everything relating to chess playing software and machines.

Moderators: hgm, Rebel, chrisw

Uri
Posts: 473
Joined: Thu Dec 27, 2007 9:34 pm

Re: G.M. MILOV VADIM versus Rybka 3 Playchess Top Challenger

Post by Uri »

This match proves that Rybka 3 and a top GM are about equally strong, the top GM being slightly stronger i think. If the games were without odds for the GM, i believe the match would have ended with a 4-4 draw.
User avatar
Ovyron
Posts: 4556
Joined: Tue Jul 03, 2007 4:30 am

Re: G.M. MILOV VADIM versus Rybka 3 Playchess Top Challenger

Post by Ovyron »

swami wrote:Near 3300 in some lists would seem exaggerated and meaningless, Imo.
Actually, Larry estimated that the engines with ~2700 rating on the lists are about right when compared to human ratings, but that engines above that strength are increasingly getting overrated due to ELO formula flaws or its implementation in the programs used to calculate them like ELOStat or Bayeselo (after enough games.)

Similarly, engines under the ~2700 rating are increasingly getting underrated, such as a 2500 rated engine would perform better than that against humans. I don't have exact figures or quotes but that was basically the idea.

The lists would then not be fixed by just subtracting 200 rating from them, something else needs to be done, but having computer rating lists that are equivalent to human rating lists seems possible.
User avatar
Ovyron
Posts: 4556
Joined: Tue Jul 03, 2007 4:30 am

Re: G.M. MILOV VADIM versus Rybka 3 Playchess Top Challenger

Post by Ovyron »

After Milov's match, Larry has just stated that:

"I would say that in human terms Rybka did not quite live up to 3100 or better expectations; maybe 3050 might be realistic."
Terry McCracken
Posts: 16465
Joined: Wed Aug 01, 2007 4:16 am
Location: Canada

Re: G.M. MILOV VADIM versus Rybka 3 Playchess Top Challenger

Post by Terry McCracken »

Dr.Ex wrote:
Terry McCracken wrote:
swami wrote:
Terry McCracken wrote:
swami wrote:
Jouni wrote:Milov is leading 4-3! Note, that he got one draw without handicap also. I have feeling, that Rybka is not playing near 3200 level at all...

Jouni
Milov is 2700+, So...without any such handicap, Rybka would surely play at >3200.
A pawn isn't worth anywhere near 500 points.
Add Pawn + Opening book + Time + Hardware.

Definitely worth 500 elo.
It has an opening book for these positons, it has the hardware and plenty of time.
Great opening book:

[D]rnbq1rk1/ppp1b1pp/3ppn2/8/3PPP2/2N2N2/PPP3PP/R1BQKB1R w

For White? Yes! :lol:
Terry McCracken
Posts: 16465
Joined: Wed Aug 01, 2007 4:16 am
Location: Canada

Re: G.M. MILOV VADIM versus Rybka 3 Playchess Top Challenger

Post by Terry McCracken »

Uri Blass wrote:
Terry McCracken wrote:
swami wrote:
Jouni wrote:Milov is leading 4-3! Note, that he got one draw without handicap also. I have feeling, that Rybka is not playing near 3200 level at all...

Jouni
Milov is 2700+, So...without any such handicap, Rybka would surely play at >3200.
A pawn isn't worth anywhere near 500 points.
It is dependent on the playing strength of the players.
The value of the pawn is higher when the players are stronger.

Uri
I realize that Uri. In fact if you give me a Good Center pawn for nothing I may turn that pawn into a piece and it will be worth ~350 rating points till finally 1000 rating points and eventually Mate with infinite value.

Terry
Uri Blass
Posts: 10297
Joined: Thu Mar 09, 2006 12:37 am
Location: Tel-Aviv Israel

Re: G.M. MILOV VADIM versus Rybka 3 Playchess Top Challenger

Post by Uri Blass »

ArmyBridge wrote:Mmmmmm, so why do all of you guys are surprised? Milov has a lot of time to check Rybka weekaness playing with this stupid odds ...let me 1 or 2 months to prepare for a handicap match and surley I can beat Rybka too, and I´m just a patzer < 2000 elo, so why congrats Milov???? :roll:
I dont see reasons to still schedule this Man-Machine matchs ,is a waste of time, what does exactly wanted prove larry in this match? even the Rybka book for this match was wrong.. If Rybka is far better than Milov in tactics why dont put in the book wild´s positions??? :? I just hope that Hydra team bring to the light its news monsters and we will see a real fight!!!
Regards
GM's already failed to win similiar match against rybka when rybka started without a pawn.

I do not see a reason to believe that patzer can beat rybka too with similiar conditions(2 games with black and 6 games with pawn or exchange odd)

I have fide rating of 2035 (I expect fide rating of 2049 in the next update) and I believe that I cannot beat rybka with pawn or exchange odds.

Uri
Uri Blass
Posts: 10297
Joined: Thu Mar 09, 2006 12:37 am
Location: Tel-Aviv Israel

Re: G.M. MILOV VADIM versus Rybka 3 Playchess Top Challenger

Post by Uri Blass »

Uri wrote:This match proves that Rybka 3 and a top GM are about equally strong, the top GM being slightly stronger i think. If the games were without odds for the GM, i believe the match would have ended with a 4-4 draw.
No basis for your words.

The facts are that milov played exactly 2 games with white and hardly got one draw in them.

Rybka proved herself to be clearly better in equal conditions.
You cn claim that the GM did not play well but the same happened also in previous match when rybka got 6 wins and 2 draws with black against another strong GM(not super GM but GM with almost rating of 2600) and it is something that normal 2700 player cannot do.

All the evidence of the results suggest that computer programs are significantly stronger than humans and even programs that are significantly weaker than rybka on slower hardware could get draws with kramnik and kasparov but for some reason some people insist that human are still equal or better.

I do not understand why they do it.

Uri
User avatar
Dr.Wael Deeb
Posts: 9773
Joined: Wed Mar 08, 2006 8:44 pm
Location: Amman,Jordan

Re: G.M. MILOV VADIM versus Rybka 3 Playchess Top Challenger

Post by Dr.Wael Deeb »

Uri Blass wrote:
Uri wrote:This match proves that Rybka 3 and a top GM are about equally strong, the top GM being slightly stronger i think. If the games were without odds for the GM, i believe the match would have ended with a 4-4 draw.
No basis for your words.

The facts are that milov played exactly 2 games with white and hardly got one draw in them.

Rybka proved herself to be clearly better in equal conditions.
You cn claim that the GM did not play well but the same happened also in previous match when rybka got 6 wins and 2 draws with black against another strong GM(not super GM but GM with almost rating of 2600) and it is something that normal 2700 player cannot do.

All the evidence of the results suggest that computer programs are significantly stronger than humans and even programs that are significantly weaker than rybka on slower hardware could get draws with kramnik and kasparov but for some reason some people insist that human are still equal or better.

I do not understand why they do it.


Uri
Thanks Uri....I also don't understand,but they still do it and they don't provide any kind of proof,just empty and worthless statements....
_No one can hit as hard as life.But it ain’t about how hard you can hit.It’s about how hard you can get hit and keep moving forward.How much you can take and keep moving forward….
User avatar
M ANSARI
Posts: 3707
Joined: Thu Mar 16, 2006 7:10 pm

Re: G.M. MILOV VADIM versus Rybka 3 Playchess Top Challenger

Post by M ANSARI »

Uri wrote:This match proves that Rybka 3 and a top GM are about equally strong, the top GM being slightly stronger i think. If the games were without odds for the GM, i believe the match would have ended with a 4-4 draw.
I really doubt that. Playing Rybka when she is lost at move 1 is very different than playing Rybka on an even plate. I think the human would have no chance and would be happy to get a draw or two.
Marc MP

Re: G.M. MILOV VADIM versus Rybka 3 Playchess Top Challenger

Post by Marc MP »

Dr.Wael Deeb wrote:
Uri Blass wrote:
Uri wrote:This match proves that Rybka 3 and a top GM are about equally strong, the top GM being slightly stronger i think. If the games were without odds for the GM, i believe the match would have ended with a 4-4 draw.
No basis for your words.

The facts are that milov played exactly 2 games with white and hardly got one draw in them.

Rybka proved herself to be clearly better in equal conditions.
You cn claim that the GM did not play well but the same happened also in previous match when rybka got 6 wins and 2 draws with black against another strong GM(not super GM but GM with almost rating of 2600) and it is something that normal 2700 player cannot do.

All the evidence of the results suggest that computer programs are significantly stronger than humans and even programs that are significantly weaker than rybka on slower hardware could get draws with kramnik and kasparov but for some reason some people insist that human are still equal or better.

I do not understand why they do it.


Uri
Thanks Uri....I also don't understand,but they still do it and they don't provide any kind of proof,just empty and worthless statements....
Right Doc. Just like your statement (your prediction before the match) that the human will be crushed like a bug...

Worthless... as you say.