Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Deep Blue's Strength

Author: Dave Gomboc

Date: 12:34:44 04/05/99

Go up one level in this thread


On April 05, 1999 at 04:57:48, Adnan wrote:

>On April 04, 1999 at 20:49:59, Dave Gomboc wrote:
>
>
>>Second, this whole thing might be a case of what Bent Larsen calls "long
>>variation, wrong variation"!  I spent another fifteen minutes or so on 20.a3 Ba5
>>21.b4 Bc7 22.Bxh7 Kxh7 23.Ng5+ Kg6 (and if anyone out there thinks that my last
>>move seems stupid, there is a new edition of "The Art of Attack in Chess" by
>>Vukovic out, so feel free to check out the chapter on the Classic Bishop
>>Sacrifice :-) 24.Qg4 f5 25.h5+ Kh6 26.Qh4 e5 27.Nf7+ Kh7 28.Nxd8 Qxd8 29.h6 Nd5
>>30.Qxd8 Bxd8 and I'd say Black even has a slight advantage.  True, there are
>>many potential places for White to improve.  But seeing as I have been
>>procrastinating from doing homework for a while, and because you are the one
>>trying to prove that White has a crushing attack after 22.Bxh7+, I'll give you
>>the opportunity to come up with improvements for White.
>
>Wrong!! In fact, I would call your 28. Nxd8 stupid, really! You are attacking.
>Why did you take the rook on d8???

What is wrong?  I said there are many potential places for White to improve.  Of
course, move 28 is one of them.  I had looked at one much earlier:

24.g4 dc!? 25.Qc2+ Nf5 26.Rce1 Kh6 27.Rxe6 fe 28.Rxe6 g6 29.gf Ne5 30.Qe4 Kh5
(30...Qd5 31.Re7, but maybe 30...Re8!?) 31.de Qd1+ 32.Kh2 Qg4 33.Rxg6 Rh8
(33...Qxh4+ 34.Qxh4 Kxh4 35. Ne4 Rg8 36.Nf6 Rg6 37.fg) 34.f3 Qxh4+ 35.Qxh4+ Kxh4
36.Ne6

I said I looked for 15 minutes, did you expect a definitive assessment?  I'm not
a 2800 player, I can't provide one.  But I'm sure you've figured that out by
now.

>Play 28. dxe5 and black is busted. You have threats like e6 and h6.
>
>28. One variation might be: 28...dxc4 29. h6 Ng6 30. Qh5
>
>and black is lost.

This looks like it works.  There are some earlier Black tries to deviate, but
they don't seem to be saving the game.

>That was simple.

I guess that means you are a stronger player than I am.  Or you have a strong
computer program to analyze with.  Or whatever.  In any case, there's no need to
be insulting.

>Believe whatever you want, but Deep Blue in 97 was not better than 2700 at >best, or more accurately 2650.

This game was from the first match, wasn't it?  I don't recall making a claim
that it was better than 2700 at that time.  Recently somebody asked how strong
Deep Blue was, and I said "2800.", but I was talking about the newer version.
This is analogous to someone asking me how strong Rebel is, I would say "2600",
because I assume they mean Rebel 10, not any of Rebel 6 through 9.

Dave Gomboc



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.