Author: Eduard Nemeth
Date: 16:06:13 09/04/01
Go up one level in this thread
On September 04, 2001 at 17:34:43, Ricardo Gibert wrote: >On September 04, 2001 at 16:35:05, Eduard Nemeth wrote: > >>On September 04, 2001 at 16:03:04, Sune Larsson wrote: >> >>>On September 04, 2001 at 15:37:40, Eduard Nemeth wrote: >>> >>>>First the game: >>>> >>>>Nsubuga,G (2260) - Scott,D [B01] >>>>US Open Framingham USA (9), 2001 >>>> >>>>1.e4 d5 2.exd5 Sf6 3.Sf3 Lg4 4.Le2 Dxd5 5.d4 e6 6.Le3 Le7 7.0-0 c6 8.c4 Da5 >>>>9.Sc3 0-0 10.a3 Td8 11.b4 Dc7 12.Se5 Lxe2 13.Dxe2 a5 14.Sd3 axb4 15.axb4 Sa6 >>>>16.Tab1 Dc8 17.Tfc1 Sc7 18.h3 Sce8 19.d5 Sc7 20.dxc6 bxc6 21.Se5 Sa6 22.b5 cxb5 >>>>23.cxb5 La3 24.Sc6 Te8 25.Sd5 Sxd5 26.Se7+ Lxe7 27.Txc8 Texc8 28.bxa6 Sc3 29.Dc2 >>>>Sxb1 30.Dxb1 Txa6 31.Db7 Ta1+ 32.Kh2 Ld6+ 33.g3 Taa8 34.h4 Tab8 35.Da7 Tc6 36.h5 >>>>h6 37.Kg2 Tbc8 38.Da4 Lc5 39.Lf4 Lb6 40.Db5 Ld4 41.g4 Tc5 42.Dd7 e5 43.Lg3 Tc2 >>>>44.Kh3 T8c7 45.Df5 Tb2 46.g5 Lxf2 47.Dxe5 hxg5 48.Dxb2 Lxg3 49.Kxg3 Tc4 50.Db5 >>>>Tf4 51.Dxg5 Tf6 52.De5 Te6 53.Dd4 Th6 54.Dd8+ Kh7 55.Dd5 Te6 56.Df5+ Kg8 57.Dg5 >>>>Kf8 58.Dd8+ Te8 59.Dd7 f6 60.Kf4 Te5 61.Dd8+ Kf7 62.Dh8 Tg5 63.Dh7 Kf8 64.Ke4 >>>>Te5+ 65.Kd4 Tg5 66.Ke4 Ke8 67.Dg8+ Ke7 68.Kd4 Tg4+ 69.Kd5 Tg5+ 70.Kc6 Tg1 71.h6 >>>>Tc1+ 72.Kb5 Tb1+ 73.Kc4 Tc1+ 74.Kb3 gxh6 75.Dg7+ Ke6 76.Dxh6 Tc5 77.Kb4 Te5 >>>>78.Kc4 Kf7 79.Dh7+ Ke8 80.Dg6+ Ke7 81.Dg7+ Ke6 82.Df8 Kf5 83.Kd4 Ke6 84.Dc8+ Ke7 >>>>85.Dg8 Tg5 86.Dh7+ Kf8 87.Ke4 Te5+ 88.Kf4 Tg5 89.Da7 Te5 90.Db7 Tg5 91.Db8+ Kf7 >>>>92.Dc7+ Kf8 93.Dd8+ Kf7 94.Dd7+ Kf8 95.De6 Kg7 96.De7+ Kg6 97.Df8 Tf5+ 98.Ke4 >>>>Te5+ 99.Kd4 Kf5 100.Kc3 Ke6 101.Kb2 Tg5 102.Ka1 Te5 103.Dc8+ Kf7 104.Dg4 Ta5+ >>>>105.Kb2 Tb5+ 106.Kc3 Te5 107.Kd4 Tg5 108.Dd7+ Kf8 109.Ke4 Te5+ 110.Kf4 Tg5 >>>>111.De6 Kg7 112.De7+ Kg6 113.De4+ Kf7 114.Dh7+ Kf8 115.Dc7 ½-½ >>>> >>>>position after move 77: >>>> >>>>[D]8/8/4kp1Q/4r3/1K6/8/8/8 w - - 0 78 >>>> >>>>For a chess program very easy! >>> >>> Yes, but still no news. A friend of mine is still ashamed of not >>> having been able to win a Q vs R ending several years ago! And >>> he had around ELO 2400 at that time... Then Walter Browne failed >>> his first try vs a computer some years ago - so... ;-) >>> >>> Sune >> >>:-) >> >>For few years I played in a "Simultan" against a Grandmaster, and last 10 >>moves me played (german "bereits Blitz!") blitz. In this posiition I give >>him my hand (I resigned.): >> >>(German: hätte ich etwas weiterspielen sollen - it was blitz? ;-) >> >>Grandmaster - Eduard, after move 66 >> >>[D]8/8/4r3/3pk3/8/3K1pQ1/8/8 b - - 0 67 >> >>:) >> >>Eduard > > >The ending KQKR is much harder to win than most players expect. Both GM Miles & >GM Browne have failed to win it against "perfect" computer play. It was once >thought such endings were easy and resignable, but computer analysis has changed >that viewpoint. > >Therefore a reasonable conclusion is your resignation is a blunder, though >perhaps a defensible decision due to your not knowing the effective methods of >defence. Only against a computer program armed with EGTBs is immediate >resignation okay. For a computer, it is a piece of cake. > >There is a way of defending called the 3rd rank defence, which you can get into >very easily in the position you give. The method of defence was unforeseen by >human players, which was discovered by computer analysis. > >A long time ago when the results of computer analysis were relatively new, a >friend of mine, an NM, got a hold of some of the Thompson EGTBs. He used them to >learn how to play KQKR perfectly. A nontrivial task. > >He did this for 2 reasons: (1) He thought by learning one type of really ending >well, it would afford some insight into playing endings more generally. (2) He >had the quite realizable fantasy of reaching this ending against an IM or GM and >stealing a draw. > >He told me about a trip to St. Johns Canada where there was a major tournament. >There, he ran into a guy who was challenging strong players to try to win this >ending against his computer. They almost always failed. Then he ran into my >friend and discovered to his astonishment that he could consistently beat the >computer in the minimum number of moves. > >All this should give you an idea of how difficult the ending actually is. When >the defending side has a pawn or more, the difficulty of winning is only >compounded. > I cannot all this understand, sorry. This endgame with GM Brown I know! Here is this game: [D]K3r3/8/5k2/Q7/8/8/8/8 w - - 0 1 Walter Browne - Thompson-Datenbank M&T-298 Endspiel, 1982 1.Kb7 Te7+ 2.Kc6 Te6+ 3.Kd7 Te7+ 4.Kd8 Te4 5.Dc5 Te5 6.Dd4 Kf5 7.Kd7 Te4 8.Dd3 Kf4 9.Kd6 Te3 10.Dd4+ Te4 11.Df2+ Kg4 12.Kd5 Te8 13.Df6 Te3 14.Kd4 Tf3 15.Dg6+ Kf4 16.Dg2 Ta3 17.Dc6 Ta1 18.Dc7+ Kf5 19.Dc2+ Ke6 20.Dd2 Ta7 21.Db4 Te7 22.Ke4 Kf6+ 23.Kf4 Ke6 24.Dd4 Tf7+ 25.Ke4 Tf6 26.Dd5+ Ke7 27.Ke5 Th6 28.Db7+ Kd8 29.Df7 Tc6 30.Kd5 Tb6 31.Kc5 Ta6 32.Dc4 Tf6 33.Dh4 Ke7 34.Kd5 Kf7 35.Ke5 Te6+ 36.Kf5 Td6 37.Dc4+ Ke7 38.Ke5 Th6 39.Dc7+ Kf8 40.Kf5 Ke8 41.Dc1 Td6 42.Dc8+ Ke7 43.Dc7+ Td7 44.Dc5+ Kd8 45.Ke6 Tb7 Line But, after 10 days played Brown again and win in a another endgame! (german: nach 10 Tagen spielte GM Brown wieder, und konnte gegen die Datenbank gewinnen, aber sehr knapp - er schaffte es, kurz vor dem 50. Zug den Turm zu gewinnen. Quelle: "Schach am PC" von F.A. Friedel und D. Steinwender) Eduard
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.