Smyslov_Fan wrote on 02/08/12 at 15:48:39:


There is no doubt in my mind that the computer had Bh3 already played for it. My engines can't find Bh3 though they are sure that a3 & Be4 wins.

At depth 32 Critter 1.4a is unable to find Bh3 with endgame tablebases installed. Time to try Houdini 2.0c (which I am sure will do no better).

Vladimir wrote on 02/08/12 at 00:34:07:

Well, lucky you! Less than five minutes.. 
And at what depth in plys? 

trw wrote on 02/07/12 at 19:02:01:


This used to be the case just a few years ago. However, this position and move are considered not too much of a challenge anymore.

On my computer, both Critter and Houdini find 47...Bh3 in less than five minutes with default settings and no endgame tablebases.

Shirov in his DVD on the Grünfeld stated that Bh3 is the only winning move in this position and that all other moves only lead to a draw. He also said that no engine could spot this move when he checked it last time, iirc he said that engines tend to dismiss certain moves right at the beginning of their calculation so that there isn't a chance that Bh3 is spotted. But not sure about that.

I came upon this diagram viewing Roman's video. He showed a game where he beat a computer in this kind of position maybe about 10 years ago with black. Now my houdini at depth 21 shows 0.55 for white. I think it is very hard for white to beat black if he just sits still.

@ Markovich: I support your opinion. As an old-timer I used to analyse myself. The engines help me only to find tactics. My own evaluation of this position before starting my engines was: white is better. That's why I went to search for tactics (first Ne4-f6+, next g5-g6). And yet, Master Om says it right: Qe3 and Qxh6 is very, very hard to find without an engine... Mhmm..and yes, too much exclams, as well as question marks. 
@ Master OM: Great article!.. I always liked chess players that are strong enough to admit they were wrong. Please, let me cite your words: "This particular game has given me insight that how a human is always superior in finding deep tactics involving positional sacrifices." and "I learnt an important lesson. Until and Unless a human guides an chessprogram to a concrete plan they are useless." ...Aaah, one more thing - never change your mind! If you start an attack (or go for a queens' exchange) then do it! Do it no matter what! A wrong plan is always better than changing your mind. 
Edit: Master Om, I think you use your engines too much. This way they don't give you the freedom to choose the move you feel is right. Put your intuition to a test! You check up and verify the moves they give you over and over.. There's no fun in this approach I think. Next time just do your own move and give it to the engines! After all, they're helpers, not players. 

Nice yeah, but do we really need an engine to lead us to this?  These to me appear to be fairly stock ideas, and White's win seems relatively evident in the diagram, though not, perhaps, precisely how to go about it.  Does g6 really rate a double exclam?  Exclams are cheap these days, I guess.

You could say in words very easily why White's position is much better than Black's.  I'm not trying to pose as a chess genius or anything.