Author: Robert Hyatt
Date: 06:29:12 11/17/03
Go up one level in this thread
On November 17, 2003 at 04:40:44, martin fierz wrote: >On November 16, 2003 at 19:45:47, Robert Hyatt wrote: > >>On November 16, 2003 at 19:23:14, Uri Blass wrote: >> >>>On November 16, 2003 at 19:12:42, rait wrote: >>> >>>>got wrong-wery closed (not suiteble) opening variation and then had no good >>>>strategy if any... >>> >>>I will ask the question in a different way. >>> >>>What is the move that changed the theoretical result of the game from draw to >>>win for kasparov? >>> >>>I think that the answer is going to be we do not know because unfortunately or >>>firtunately we did not solve chess. >>> >>>Uri >> >>I can think of a couple. the early a6. The f4 move that really crystallized >>the pawn structure. Pushing a5 isolating that pawn. Once the pawn structure >>was defined, it became imperative that black advance on the kingside as white >>owned the queen-side. Black fiddled while Rome burned. > >objectively, i disagree completely: >1) the early ...a6 was played by kasparov himself in his infamous loss against >huzman earlier this year. That's ok, but in this case, who cares what _Kasparov_ played. DF played into a line that it wasn't capable of understanding. Ergo, it was a bad move, and a bad plan. >2) i don't know what "f4" move you are referring to. that was never played by >either side... Sorry, a typo. e4. Building the chain e4-d5-c6-b7 which was locked for pretty much all time. >3) isolating the a5 pawn is *not* a problem as long as you know what to do >afterwards. e.g. in the french winawer, white often goes a4, and black can >always go and gobble that pawn. in return, white pushes his entire kingside down >the board and has more than enough compensation. Wasn't going to happen here. Again, my point. a5 is ok if you know how to play it. Fritz did not. Therefore a5 was a blunder, as the pawn was lost with zero compensation. You are talking from a human perspective on the game. I am talking from the apparent perspective of the program playing the black pieces. > >GM yermolinsky was saying that he actually liked black's position after 10 >moves. he believed that black had good chances with such a kingside pawn >advance. Again, who cares what he thinks? He is looking at it from a human perspective. Advancing the king-side pawns would be a natural plan. But _not_ to Fritz, as we all saw. Therefore, if the program has no hope of finding the right plan and exercising it, then playing into such a position should give every move leading to the position a ??. > >but....: >of course, fritz had no idea that it had to do that, so in a way you are right - That is my point. That is the part about creating an opening book that is so very difficult. You _must_ play into variations you can reasonably be expected to handle. And you must avoid those variations that lead to positions you are going to wreck. >getting into this interesting position was equivalent to losing for fritz. but >it has nothing to do with the objective evaluation of that position. Objective evaluations are meaningless here. You have to evaluate from the perspective of the two players. White knew what was needed. Black had no clue. The basic conclusion was that black was busted, even if a human GM could have won with black. > >i was already surprised that they chose ...a6 - that often leads to such closed >positions. i've asked this before, but i don't remember getting an answer: how >good is mr. kure at chess? > He is very good at preparing openings against computers. However, when you step into GM competition, you enter a different world. That might have been find vs computers, but against a wiley human, it was a mistake. >cheers > martin
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.