Author: Howard Exner
Date: 14:07:55 01/29/98
Go up one level in this thread
On January 29, 1998 at 07:19:05, Amir Ban wrote: >On January 28, 1998 at 13:45:07, Howard Exner wrote: > >>Posted by Amir Ban on January 27, 1998 at 11:53:23: * >> > >> >>*No. 336 (Rxd4): >> >>*I find Bxe5! as a free gain of a pawn. I don't know what the point of >>*Rxd4 is. >> >>*Recommend: Change key to Bxe5. >> >>#335 >>Book: 1... Rxd4 2.Rxd4 Be5 3.Rd8 f5 4.Kh3 Rxf2 5.Rd3 Kh5 0-1 >>To avoid this line I think white will have to drop all the pawns >>which is also hopeless. Is there a defense for white? The mate theme >>is pretty. >> > >This one's 335 of course. In the line above black mustn't play 5.Rd3 but >let the f-pawn go. Yes, we agree on this. A sample line might be 5.Rg8+ Kf6 6.Ra8 Bf4 and f3 will fall and black keeps the three pawns. >Is the win of R+B+3P vs R+R easy ? Should compare it >to the position after 1...Bxe5 2.Rxe5 Rxd4+ 3.Kf3. I evaluate the two >positions about the same. Which of these is actually won ? Here is a quote from Yuri Averbach from his endgame book "Rook vs Minor Piece Endings". He states,"in contrast to the middlegame,in the endgame a bishop and two pawns are normally stronger than a rook." In this position there is an extra rook for both sides and an extra pawn, making three pawns up and all connected. From my judgement the win here would be more effortless than the two rook ending with one pawn up. I'm not sure black could even win from this position since white's king is in a favourable spot. I wonder what chess programs set on autoplay would come up with for both positions? > > > >>*No. 354 (Qxf3): >> >>*This one's unclear. 1...Qxf3 2.Nxf3 Rxh3+ 3.Kg1 Rxf3 is a nice position >>*for black but I don't see a decisive line. This may be meant as a >>*positional sacrifice. It doesn't look inferior to any other >>possibility, >>*which may be a basis for accepting it. >> >>*Recommend: Investigate. >> >>Book continues with 3 white replys: Nc4, Qd1 and Qh2 which all fail. >>Rebel 8 suggests Qb4 with the following position: >>5rk1/2p3pp/p1p5/2P1p3/1Q1pPn2/1P3r2/PN6/R5K1 b - - id "Pos 354 - >>ECM.EPD" >>Now here black should avoid a draw sequence and play h5! which >>gives his king some breathing room so that the rook on f8 can >>be lifted. So for now it looks ok to play Qxf3 but programs may >>play this to draw initially. >> > >I got as far as Qb4 h5 but remained unconvinced. I have a positive score >here, but I don't see a forced win and it's not clear how long black can >maintain the threats and contain the white pieces. If he can't he may >lose. Are you looking at the line where white plays Qc4+ Kh7 and Qxa6? In that case Rf6 should win. I can't see any white defense for now. What are you finding for white? > > >>*No. 366 (Rxd8): >> >>*I think this one's wrong. 1.Rxd8 Qxd8 2.Qh4 Kg8 3.Rxf6 white gets two >>*minor pieces for a rook, but will be sorry he did after 3...Qb6 which >>*gives black a strong counterattack. >> >>*Recommend: Discard. >> >>The book only considers Bc4 instead of Kg8. >>As white I would continue your above line with 4.h6! Qxb2(what else?) >>5.Rh6! f6 6.Rxh7 1-0 but there might be some better play for black >>after h6. If this holds then Rxd8 is good. >> > >I suppose you mean 4.h3? Yes > >Then try 4...Rc6. All continuations are unclear: 5.Rh6 gxh6 6.Bf6 Qe3 or >5.Nd1 Rxc2 or 5.Qg3 Kf8 or 5.Bh6 gxh6. The quiet Kh2 still makes Qxb2 unplayable. Black can try Rec8 and now Rebel 8 suggests Rf2 while I was trying Bd3 as another possibility. In any case the line is quite involved and maybe beyond the scope of what the programmers here are looking for in a test suite. > >>*No. 371 (Rxd5): >> >>*I think this is based on faulty defense: 1.Rxd5 cxd5 2.c6 Bc8 3.Ba6 Ka8 >>*and black is not in trouble. >> >>*Recommend: Discard. >> >>This is a long Nolot kind of position unless there is a flaw in the >>following book >>line: 1.Rxd5 cxd5 2.c6 Bc8 3.Ba6 Ka8 4.Nd4!! Qg4+ 5. Kb1 Qe3 6.Bb7+ Kb8 >>7. Rd1 Be7 8.Qb5 1-0. Be7 is poor but I can't see a defense at this >>point. Is there >>a better defense earlier on though? >> > >How about 5...Qg4 ? Bb7 wins violently, despite the many checks for black and material gain. Ka1 also looks like a win. Rxd5 is a move that I doubt will be found by computers > >>*No. 381 (e5): >> >>*I choose fxe6 (Bxe6 Nd4 or fxe6 e5) which evaluates about the same as >>*e5. Both options are not decisive in my opinion. >> >>*Recommend: Discard. >> >>Book: 1.e5 dxe5 2.Ne4 Qb6 3.Nd6+ Ke7 4.Nxe5! fxe5 5.f6! bxf6 >>6. Rxf6 and according to the book white will mate or win big material. >>If this is all forced then e5 looks stronger than fxe6. >> > >This works well with the black queen on b6, but if 2...Qc6 the same >continuation doesn't work well. Black can play Rcf8 (best) or even Kxf6. After Qc6 this might transpose to your original exf6. From your original exf6 Bxe6 do you now score Nd4 as better than e5?
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.