Author: Ricardo Gibert
Date: 15:50:24 03/11/00
Go up one level in this thread
On March 10, 2000 at 16:35:49, Robert Hyatt wrote: >On March 10, 2000 at 15:11:38, Mike S. wrote: > >>On March 10, 2000 at 13:48:10, G. R. Morton wrote: >> >>>(...) >>> >>>Black : pawns at h7,b6 and h5, king at d6, rook at c7. >>>White : pawns at a2,b2, and d3, king at g2, rook at f7. White to move. >>> >>[D]8/2r2R1p/1p1k4/7p/8/3P4/PP4K1/8 w - - >>>The average club player can quickly see that RxR?? gives an easy win for black, >>>since after the exchange, black simply uses his h pawn to draw the white king >>>away from the queen side on which black pushes through. (...) >> >>Is the position you gave correct? I have tested with Shredder 3 (without TB's), >>Crafty and M-Chess 7.1, all play Rxc7 and think white is winning after 1...Kxc7 >>2.a4! Does this move appear in the book you mentioned? I think, the previous >>move ...Rc8-c7 might be a blunder. >>Astonishing also, that the programs want to play 2...b5 (?!) for black - I >>haven't really analysed and do not understand, but 2...h4 3.b4 looks like RxR is >>ok (?). >> >>Regards, >>M.Scheidl >> >>Permanent Brain: >>http://members.surfeu.at/MScheidl > > >At reasonable depths, I get Rf6: > > (2) 13 1:14 0.45 1. Rf6+ Kc5 2. d4+ Kb5 3. Kf3 h4 4. > b3 Rc2 5. a4+ Ka6 6. Rh6 Rc3+ 7. Ke4 > h3 8. b4 Kb7 9. d5 > >How long did you search with crafty? > >at very shallow depths, I also get rf6. In the 'middle' I get Rxc7 but as it >goes deeper it decides the distant passers are a problem... although I don't see >why white can't play Rxc7, a4, b4 and a5, and at least turn >this into a draw?? Your idea is very close to correct. 1.Rxc7 Kxc7 2.a4 Kd6 3.b4 followed by Kg2-h3xh4-h5xh6, etc. Black cannot play Kd6-d5-d4xd3, since a5 would then Queen. An easy win, but White must hold off on a5 until the proper moment. Your idea to play a5 quickly is worth discussing. This draws. When 2 passed pawns are separated by 1 or 3 files, unless they are advanced too far, they can be stopped, but not captured (in general). They are "self-sustaining". When separated by 2 files, they must succeed in queening or they will be stopped _and_ captured. They are not "self-sustaining". Thanks to the presence of the b-pawns on the Q-side, the pawn configuration is "self-sustaining" there so White can win just has long as White holds back in playing a5, which eliminates the b-pawns. Of course, the above assumes the passed pawns are not assisted by another unit, which is the relevant case here. 1.Rxc7 Kxc7 2.a4 Kd6 3.b4 Kd5 4.a5 bxa5 5.bxa5 Kc5 6.d4+ Kb5 7.d5 Kxa5=. At first, it seems Black can improve by 3...Kc6, since 4.a5 bxa5 5.Kb5 draws more easily, but in reality, it discourages White from playing the mistake 4.a5 and find the correct idea instead. A trivial example of how a sub-optimal move can encourage an error. Another reply to your post gives 1.Rxc7 Kxc7 2.a4 b5. Of course this loses as White gets 2 passed pawns separated by 1 file. These are "self-sustaining". They can be stopped, but not captured and white wins as in the 1st paragraph above. Care must be exercised in the case when the passed pawns are separated by 2 files, since this is an all of nothing case. The analysis of this position is very easy for a human armed with a small set of ideas. Any IM or GM would take less than 10 secs to realize all of the above.
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.