Not logged inRybka Chess Community Forum
Up Topic Rybka Support & Discussion / Rybka Discussion / Possible handicap for rybka match
- - By kaustubh (**) Date 2007-06-27 05:47
I think the human opponent can be given plain paper to write his variation and analysis during the match or an extra board where he can try his variations .
Parent - - By Uri Blass (*****) Date 2007-06-27 07:59
This is not a significant advantage for the strong GM's.
Parent - By Uly (Gold) Date 2007-06-27 08:16
I agree with Uri.

Some grandmasters can play at the same strength even by playing blindfold chess, and since that's seeing no board at all, I can't see how an extra board would help (Though it may be very helpful to weaker players).

Now, don't let Rybka see the board and see what happens ;) (This last statement is not meant to be taken seriously ¬_¬)
Parent - By VP (*) Date 2007-06-28 03:41
I agree as well.
A Grandmaster does not have to play moves on the board to be able to actually "see" them- although the strength does drop while playing blindfold chess.
An extra board and analysis may be acceptable for correspondence chess, not timed competition, where time is of essence.
Parent - - By Lee Ma Hong (**) Date 2007-06-27 23:26
If you want to see a handicap match, why not simply a match between a GM with Deep Fritz or Deep Junior on a C2D machine versus pure-engine Rybka?
Parent - - By VP (*) Date 2007-06-28 03:37
Are you really serious? A GM supported by a computer would thrash Rybka.
Let's wait first for the Computer- Man event to be over before drawing conclusions.
Regards
Parent - - By Leto (***) Date 2007-06-28 03:42
I don't believe that.  Maybe that'd be right if the GM is supported by Rybka, but if it's any other engine I doubt it would be of any help to the GM.  Even if the GM is supported by Rybka, would he really be listening to the suggestions of Rybka or would he rather trust himself?  If he trusts Rybka more than himself, is the GM really playing or is it just another Rybka vs Rybka match? 
Parent - By VP (*) Date 2007-06-28 04:07
Rybka may be very good, but it is only because of its superb search and evaluation.
Anand with a Fritz on 32 bit CPU would most definitely beat a standalone 8 core 64 bit Rybka, simply because he plays better positional chess than Rybka, and any human errors/ blunders would be cut down by Fritz.

Regards,
Parent - - By Lee Ma Hong (**) Date 2007-06-28 03:43
There are people in this forum who really believe Rybka is an invincible god. My suggestion was given with sarcasm.
Parent - By VP (*) Date 2007-06-28 04:04
Toche

Thanks for the clarification!
Parent - By turbojuice1122 (Gold) Date 2007-06-28 11:58
Actually, I think your suggestion is nonetheless very good for the following reason: Rybka would probably lose such a match, and what better way to find its weaknesses and try to correct them?
Parent - - By Roland Rösler (****) Date 2007-06-28 04:09
Look to the freestyle. Pure Rybka is trashed by centaurs with Elo less than 2200.
Parent - By turbojuice1122 (Gold) Date 2007-06-28 12:00
Yes, but not by centaurs with elo over 2500.  Nakamura, for example, has always performed quite poorly in these tournaments.  Only centaur teams, with both knowledgeable "engine attendants" in addition to grandmasters performed fairly well.  Even in this past tournament, pure Rybka trashed single GM plus engine.
Parent - - By Roland Rösler (****) Date 2007-06-28 05:10
Vas don´t like matches, where Rybka can lose. But this match can be although good marketing. The GM will be surprised by the good moves of Rybka, which he and Fritz (Junior) don´t have suggested. After the match, which he surely wins, he will damn this poorly engines (Fritz, Junior), which have no ideas and will admire his smart opponent. So smart, that he often has to recalculate his evaluation. If this GM is an Top-Five GM (Anand or Kramnik for example; not Topalov or Aronian, they are convinced, I believe), it will be a good marketing clou from Rybka side. But these people are to expensive! And it would break my heart, if even Rybka (or Convekta) give one dollar to the reigning WM (I don´t tell you, what I´m thinking about him). But Anand would be okay; but too expensive. But why not?
Challenge to Anand from Rybka side! 20.000 $ for a 6 round centaur match (assisted by a C2D PC with a CB-engine of his choice) against pure Rybka engine (hardware, engine, opening book in their choice). Classical time control (120/40, 60/20, 60 rest) and no rest (decision in the time).
It´s sure, that the money is lost for the Rybka side! But they can make 2 points in this match, I believe. Great performance!
PS: 20.000 $ is much money, but eventually not enough to challenge Anand in this way. Okay, then forget it
PS2: Adams played a match against Hydra for less money; I guess. I like Adams and his play. He crushes the reigning WM, when he wants to do it (and he wants it very often)!
Parent - - By Uri Blass (*****) Date 2007-06-28 05:22
I can add that I think that rybka has also chances to win team of human and Fritz in normal match.

remember that even team of Vas and rybka with white could not beat rybka in the last round of the free style so
I doubt if GM+Fritz is stronger than Rybka in normal chess.

Uri
Parent - - By Roland Rösler (****) Date 2007-06-28 07:41
You also have to remember this: In five years (or ten?) Anand will be beaten by a programm, we have in our mobile phone or wrist watch!
PS: Not in centaur mode, but in OTB mode.
Parent - - By Uri Blass (*****) Date 2007-06-28 08:06
I think that it can happen in less than 5 years and I do not see the relevance of it for this discussion.

Uri
Parent - - By gala.martin (**) Date 2007-06-28 08:33
I think Anand cost much more than 20k$; still, he is supposed to be one cheap for his ranking (I guess Topalov would cost even more). From my point of view, the best handicap against a top GM is a time handicap (eg. rybka 5+10 and GM 60+15).

Although I can defeat rybka running on my laptop at knight odd, I start thinking than -in a short time- knight odd could be reasonable for IM level players. Rybka would mainly need a very well cooked book,  and some anti-human improvements (ok, maybe also a good hardware). But my impression is that even knight odd can be challenging at a good level.

It would be nice to have also handicapped positions in the rybkaIII book.
Parent - By lkaufman (*****) Date 2007-06-29 02:45
     In actual blitz play (straight 5' with no increment), Rybka 2.3.2 on a quad has at least demonstrated equality at knight odds with myself and my son (with varied openings stipulated). I am a bit weaker in blitz than the average IM due to my age, while my son, a strong FM with one IM norm and three near-misses, is also probably in the low-IM range in blitz strength. But with even a couple seconds added increment, I think we would become solid favorites. A knight is quite a large handicap. But at pawn and move (meaning Rybka plays without f7), I would already back her against any human in blitz, even Anand.
Parent - - By genorb (**) Date 2007-06-28 09:02
remember that even team of Vas and rybka with white could not beat rybka in the last round of the free style so I doubt if GM+Fritz is stronger than Rybka in normal chess.

you draw strong conclusion from 1 game, we would need much more ;) ...
Parent - - By Uri Blass (*****) Date 2007-06-28 12:12
I did not draw strong conclusions and I only claim that it is not clear that rybka is going to lose against human+fritz.
The point is that even pure rybka against human+rybka is not a sure loss for pure rybka.

Uri
Parent - By genorb (**) Date 2007-06-28 15:21
The point is that even pure rybka against human+rybka is not a sure loss for pure rybka.

Yes but you need examples to say that, but perhaps there is enough examples on playchess... But GM+Rybka against Rybka alone, I do not know if there is so much examples to draw any conclusion.

But this remark was also done to tease you and it works as expected :) .
Parent - - By VP (*) Date 2007-06-28 23:30
In chess, sometimes a player rated 2300 would be able to draw ( or even defeat) a 2400 rated player.
And, with due respect, there is a difference in class between a top GM like Anand (or Kramnik or Topalov) and Vas, when it comes to OTB chess.
Uri, apples should be compared to apples.

When we are talking about human+engine v/s Rybka, we definitely do not mean any human, but a top GM.
Never underestimate them.
Parent - - By turbojuice1122 (Gold) Date 2007-06-28 23:43
I again bring forth the example of Nakamura, who has always played poorly in Freestyle events, in spite of being rated around 2660 for one of them.
Parent - - By VP (*) Date 2007-06-29 00:06 Edited 2007-06-29 00:28
Was he playing as a human, or was he using a Computer as well? I think he was playing with minimal computer assistance, if any.
I know an Indian GM tried to enter the freestyle without any computer assistance and was outplayed.

From an article in chessbase by Correspondence GM Arno Nickel

Can Grandmasters play Freestyle?
Two or three tournaments back it would have been a sensation, if grandmasters like Hikaru Nakamura (FIDE Elo 2651) failed to achieve more than 50 percent in the Freestyle Main Tournament. Nowadays any experienced Freestyle player or even kibitz is no longer surprised. Players like him will always play beautiful games, with lots of ideas, but they are vulnerable when it comes to deep tactics on the board, especially in a rapid game 60m + 15s. Although they have Fritz or other advisors at their disposal, they will always try to force their own moves and only realize that something went wrong with their plan when it’s already too late. Someone asked “Star Wars” (H. Nakamura) in the chat whether he uses Rybka, which is the most popular Freestyle engine running as UCI engine under the Fritz or ChessBase. His answer: “I use my brain, because its better than Rybka on 6/7 days of the week.”

Now, he was being arrogant when he made that claim and was thoroughly outplayed.

Or have you forgotten, when in the third freestyle,  Krasenkow helped team Rajlich win an endgame, which Rybka declared as a draw and even Vas could not see a win.
Then, the human GM (and current world no 42) led the way.
Do you think a world no 1 would not know how to convert such an advantage?
BTW, Namamura (2663) is very close to Krasenkow(2661) in FIDE rating.

While I am not saying that Rybka playing at full strength is not going to win against Anand, I think that Anand with a computer would be more than just a handful.
Have we forgotten that he won three Advanced Chess tournaments in Leon, against the best GMs assisted by Chess engines?
Parent - By turbojuice1122 (Gold) Date 2007-06-29 04:26
Krasenkow already had experience working with the Rybka team, and they are a very experienced centaur team.  The world top 10 are not experienced or knowledgeable centaurs (except for Anand), and none of them would currently do well by themselves with a high-speed computer in a Freestyle tournament.  However, they would probably do fairly well after a bit of practice, but even after that, I would place my money on team Rajlich against a team consisting of Topalov and Mamedyarov, but not against a team consisting of Anand and an experienced engine technician.
Parent - - By lkaufman (*****) Date 2007-06-29 02:48
There is nothing wrong with such a match, it might be about even (depending on the choice of GM), but it is not a handicap match between man and machine, but between engine and engine, since Deep Fritz or Deep Junior would have to be considered the stronger player on the team and hence the main opponent.
Parent - By turbojuice1122 (Gold) Date 2007-06-29 04:29
I was thinking earlier that a match that would be quite interesting would be Rybka vs. a top 10 GM who is allowed to use Fritz 8 on a single-processor computer.  Here, you have a computer program that would be at the level or lower than a top GM--it would be able to help in blunder prevention and also in analyzing tactical lines, while the GM would need to work out the positional details.
Up Topic Rybka Support & Discussion / Rybka Discussion / Possible handicap for rybka match

Powered by mwForum 2.27.4 © 1999-2012 Markus Wichitill