Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Congratulations RYBKA on winning IPCCC, says forum's patzer A. Steen

Author: A. Steen

Date: 06:18:03 12/30/05

Go up one level in this thread


On December 30, 2005 at 08:51:19, Sune Larsson wrote:

>On December 30, 2005 at 07:20:46, A. Steen wrote:
>
>>Unlike most of the crowd, at all stages I was confident Rybka would not lose
>>this game:
>>
>>http://wwwcs.uni-paderborn.de/~IPCCC/IPCCC2005b/pgn7.html
>>
>>[Event "IPCCC"]
>>[Site "Paderborn"]
>>[Date "2005.12.30"]
>>[Round "7"]
>>[White "Ikarus V0.36FR7a SMP"]
>>[Black "Rybka Beta w64 1.0"]
>>[Result "*"]
>>[ECO "B00"]
>>[TimeControl "7200"]
>>1. e4 Nc6 2. d4 d5 3. e5 Bf5 4. Nf3 e6 5. Bd3 Nge7 6. Bg5 Be4 7. c3 h6 8. Bxe7
>>Bxe7 9. Bxe4 dxe4 10. Nfd2 Qd7 11. Nxe4 O-O-O 12. Qe2 f6 13. f4 f5 14. Nf2 g5
>>15. O-O Rhg8 16. Kh1 g4 17. Nd1 h5 18. Nd2 h4 19. Ne3 h3 20. gxh3 gxh3 21. Rf3
>>Rh8 22. Qf1 Bf8 23. Rxh3 Bh6 24. Qf3 Ne7 25. Rg1 Kb8 26. Rg2 Qa4 27. Rh5 Qxa2
>>28. Qh3 Ng8 29. Ndf1 c6 30. Rg6 Re8 31. Rgxh6 Nxh6 32. Rxh6 Rhg8 33. Qh5 Qxb2
>>34. Qf7 Qxc3 35. Rxe6 Ref8 36. Qe7 Qxd4 37. Qd6+ Qxd6 38. Rxd6 a5 39. Ng3 Rd8
>>40. Ngxf5 a4 41. Nc2 Kc7 42. Rxd8 Rxd8 43. Nd6 Ra8 44. Kg2 a3 45. h3 a2 *
>>40. Ngxf5 a4 41. Nc2 Kc7 42. Rxd8 Rxd8 43. Nd6 Ra8 44. Kg2 *
>
>
>     27.a3! and it's a whole different ballgame....

Correct, Mr Larsson, though the exclamation mark is quite unnecessary IMO.

But ... possibly slightly better was, on white's previous move, the prophylactic
26. b3.

[d]1k1r3r/pppqn3/4p2b/4Pp2/3P1P2/1PP1NQ1R/P2N3P/6RK b - - 0 26

where if 26. ... a5 then 27. a4 and then if 27. ... b5 28. axb Qxb5 29. Ndc4 and
white wins.  Not that I suggest engines would want either 26. ... a5 or 27. ...
b5.

Why let the lady in to the Q-side at all?  With the WK where it is, queening the
h pawn is a very, very diffuse and long-term goal/threat, and there is time for
the BK to gobble and return.  Hence 26. b3

There is another difference between 26. b3 and 27. a3 - the one that makes me
slightly prefer the former - but I leave it to you to find out.

>     First pawn sac is okey - as so the position
>     after 15.-Rhg8.

It is even.

>     Then some queer moves followed - resulting in
>     pawn sac n:o 2 - and black has really no comp.

Queer, yes.  Tal's calling cards were queer too.  And often he had no
compensation - but he won the point all the same.  Rybka, like Misha, knows his
opponents are in some ways patzers, especially when the clock is ticking.

As I wrote below:
"So the theoretically unsound sac delivers the full point more often than not,
until the next generation of opponents evolves."
>
>     But why shall everything be "correct" and
>     scientific verifiable??

Clearly, you have understood what many others here did not. :-)

Rybka is a pragmatic OTB playing engine, amply equipped with the poison that
some GM or other might have said somewhere that Fruit lacked.

> Boring boring said the child and smiled...

and later collected the full point, leaving the local heroes fulminating about
why their masterpieces do not deliver.

Unfair old world, isn't it?

> ;-)

;-)

>     /S
>
>
>
>>
>>[d]r7/1pk5/2pN4/4P3/5P2/p7/2N3KP/8 w - - 0 45
>>
>>After the obvious 44. ... a2, Ikarus has absolutely no winning chances and would
>>be astoundingly lucky to draw (a power outage, maybe?). I stopped looking.
>>
>>While the above game is very far from a fine example of what Rybka is capable of
>>(it is more of what poor Ikarus was hoodwinked into permitting), I could thank a
>>few of the forum's "chess experts" for the amusing fantasies and wrong prognoses
>>they wrote about this game along the way.
>>
>>Black obtained three supported united passed pawns on the Q-side while the White
>>King was so magnificently placed on h1.  Rybka's opponents do not properly know
>>how to handle the very unbalanced positions that arise from Rybka's Tal-like
>>sacs.  So the theoretically unsound sac delivers the full point more often than
>>not, until the next generation of opponents evolves.
>>
>>My general evaluation of Rybka:
>>Much stronger than overly-staid Fruit, psychotic Shredder, berserk-monster
>>Fritz.
>>
>>'Extraordinarily dynamic middlegame: unbelievable Rybka, a killer grandmaster.'
>>
>>A. Steen strongly recommends Rybka -
>>* to patzers, to watch/study it play their other engines, and hopefully learn at
>>least a little therefrom; and
>>* to others, to play it themselves.
>>
>>Congratulations, Mr Rajlich. You clearly have (at least?) one super-GM on your
>>team.  :-)



This page took 0 seconds to execute

Last modified: Thu, 15 Apr 21 08:11:13 -0700

Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.