Author: Francois Bertin
Date: 06:03:57 09/09/98
Go up one level in this thread
On September 09, 1998 at 06:15:09, Amir Ban wrote: >On September 08, 1998 at 21:23:43, Francois Bertin wrote: > >> >> 6r1/p2Pbk2/P4p2/1B3P2/2K1p1p1/8/8/3R4 b - - >> >> This position happened in a correspondence game I played recently. >>I played Rg5, letting White promote to queen, and then sacrificed my bishop >>on it, after which White took with rook. I was hoping that my two passed >>pawns on the kingside would give me good chances. That was quite >>speculative, to say the least and I had to concede victory to my >>opponent about 15 moves later. >> > >How was the game lost ? I thought too I had at least a draw because of the far advanced passers, but somehow I badly misplayed after 43...Rg5 and got myself in a hopeless position with K and B plus a lonely passed pawn on a7, against K, B and R. I saw no way of doing something with the pawn and so resigned :-/ >> However, when I put this position in Rebel 9, I was surprised to see >>that the program liked the move and even gave a small advantage to Black!? >> >> I run Rebel on a P-90 with only 8 Mg of memory, so I would like to >>know if it would stick to the sacrifice and its evaluation when running >>on a more powerful system with much more memory for hash tables. And what >>would other programs like Fritz, Hiarc and al. play? >> > >Rg5 looks like a good move, giving black the advantage. It should be good enough >to draw, at least. I have: > >1... Rg5 2.d8=Q Bxd8 3.Rxd8 g3 and black looks dangerous because white doesn't >have time to capture on a7. Black can easily opt for a draw by taking the f5 & >a6 pawns. > >Maybe (1... Rg5) 2.Bc6 is safer for white, but after 2... Rxf5 is doesn't look >good either. > >1...g3 doesn't win for black because of 2.Kd4, getting the king nearer with >tempo because of the threat Bc4+. It looks good enough to draw, though. > >Amir
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.