Author: A. Steen
Date: 15:52:30 11/20/05
Go up one level in this thread
On November 20, 2005 at 17:41:42, Uri Blass wrote: >On November 20, 2005 at 15:41:30, A. Steen wrote: > >>On November 20, 2005 at 13:39:30, Uri Blass wrote: >>>I do not understand 26...Rag8 >>>Even I could easily avoid that move against a computer because it is easy to see >>>that it gives a pawn. >>> >>>[Event "II Man vs Machine, Bilbao"] >>>[Site "?"] >>>[Date "2005.11.20"] >>>[Round "?"] >>>[White "Hydra"] >>>[Black "Kasimyanov"] >>>[Result "*"] >>>[Annotator "Hob"] >>>[PlyCount "59"] >>>[EventDate "2005.??.??"] >>> >>>1. d4 Nf6 2. c4 g6 3. Nc3 Bg7 4. e4 d6 5. Nf3 O-O 6. Be2 e5 7. Be3 Na6 8. O-O >>>c6 9. d5 Ng4 10. Bg5 f6 11. Bh4 c5 12. Ne1 Nh6 13. a3 Bd7 14. Nd3 g5 15. Bg3 >>>Qe7 16. f3 f5 17. Bf2 f4 18. b4 b6 19. h3 Nf7 20. Rb1 h5 21. Rb2 Nh6 22. Be1 >>>Bf6 23. Nf2 Kh8 24. Qd3 Rg8 25. Nb5 Rg6 26. Bd1 Rag8 27. Nxa7 g4 28. fxg4 Bh4 >>>29. Nc6 Qg7 30. Bc3 hxg4 31.Nxg4 Nxg4 32.Bxg4 Bxg4 33.Rxf4 Bh5 34.Qf1 Bg3 >>>35.Rf7 Qh6 36.Bd2* >>> >>> >>>Uri >> >> >>[D]r6k/p2bq3/np1p1brn/1NpPp1pp/1PP1Pp2/P2Q1P1P/1R3NP1/3BBRK1 b - - 0 26 >> >>Uri, with great respect, in this position 26 .. Ra-g8 is a fine move and made at >>the perfect time (in my opinion). White's pieces (especially the N on b5) are >>not well-placed for a defence on the K-side, and white has not yet begun a pawn >>push on the Q-side. To delay is not wise, white may regroup and bring his >>misplaced N back. But see the last para before replying. >> >>After grabbing the P the WN is badly out of play, and black can attack strongly >>on the K-side. >> >>[D]6rk/N2bq3/np1p1brn/2pPp1pp/1PP1Pp2/P2Q1P1P/1R3NP1/3BBRK1 b - - 0 27 >> >>Speaking only from the perspective of a human player, here I think 27 .. Q-h7 >>(or after 27 .. g4 as actually played) and black is fine even against strong >>computers. After his K-side pawn push, the h-file is forced open and then BQ can >>eventually occupy h4 or h3. White may have to fight for the draw. >> >>In the game played, I think Kasimdzhanov misplaced the Q on the g-file on move >>29. Its place is on the h-file (h7), or maybe on f6 to menace h4. >> >>But even after this inaccuracy, black had equality, up to this point IMO- >> >>[D]6rk/6q1/npNp2r1/2pPp2b/1PP1PR1b/P1B4P/1R4P1/5QK1 b - - 0 34 >> >>Now all black has to do is retreat with 34 .. B-f6 (obviously not e7 or d8 else >>white can exchange his compromised knight), and next move correct his error >>(from move 29) and put the Q on the h-file. White is in a bind, and can only >>draw this. >> >>But unfortunately Kasimdzhanov blundered here with 34 .. B-g3?, at which point >>it is 1-0 in my opinion. HE ALLOWED THE WR TO ENTER HIS 7th RANK BY THIS >>BLUNDER!! And also activated the Q-side WN, which can enter quickly via the 7th >>rank. >> >>[D]6rk/6q1/npNp2r1/2pPp2b/1PP1PR2/P1B3bP/1R4P1/5QK1 w - - 0 35 >> >>Hydra of course found 35 R-f7, and now the human could safely resign. I don't >>have the moves after 36 B-d2 but I would predict Hydra will exchange and push a >>pawn storm through, with its R on the 7th it has nothing to fear on the K-side. >> >>Now please set it up with the programs and show me I am wrong in my human >>overview. >> >>Uri, given that K. was not able to handle the tactics properly, you too are also >>right that K. should have avoided them by keeping his R on the a-file and not >>playing 26. Ra-g8. :) >> >> >>-A.S. > >Fruit cannot see advantage for black by 26...Rag8 27.Nxa7 I did not say 26...Rag8 led to an advantage for black, only to a draw. And with the reputation of Hydra (5.5-0.5 against an alleged super-GM) a draw would be a sensible target. >Note that black could get equality and maybe more than it by other means. Such as what? By the move (26. .. Ra-g8) played, Kasimdzhanov has seized the initiative, wrong-footed the WN and developed a dangerous K-side attack that white can only draw against. Below I show your own (ie. Fruit's) analysis supports my position, not your position. :) >34...Bf6 probably lead to equality but draw does not justify 26...Rag8. There is no "probably" about 34...Bf6; white has no prospects of a win after it. >34...Bf6 35.Rf5 Qh6 36.Ra2 R8g7 37.Kh1 Bg4 38.Bd2 Qh4 39.Be1 Qh6 40.Bd2 > = (0.00) Depth: 16/40 00:01:49 69969kN Again, agreeing with me that 34. .. Bf6 is an obvious draw. No "probably". Now we come to your main point. >27...Qh7 28.Nc6 g4 29.hxg4 Bh4 30.Qc3 hxg4 31.fxg4 R8g7 32.Qh3 Bxf2+ 33.Bxf2 >Bxg4 34.Bxg4 Nxg4 35.Qxh7+ Rxh7 36.Re1 Rgh6 37.Kf1 Rh1+ 38.Ke2 R1h2 39.Kf3 Nxf2 >40.Rxf2 > ² (0.26) Depth: 19/53 00:23:11 798200kN I could ask how does a +0.26 evaluation help your argument that black has a weaker position? +0.26 evaluation in such a position, after deep analysis, means in my experience 'draw likely'. But there is a better method to refute you... I will go through the analysis you kindly gave. 27...Qh7 28.Nc6 g4 29.hxg4 Bh4 30.Qc3 hxg4 31.fxg4 Up to here, I agree. This is the position at this stage- [D]6rk/3b3q/npNp2rn/2pPp3/1PP1PpPb/P1Q5/1R3NP1/3BBRK1 b - - 0 31 Here your provided Fruit analysis has black play 31. .. R8-g7 and then white gets +0.26. But 31. .. R8-g7 is a ridiculous move, total nonsense, inhuman, robotic, unnecessary, pointless, insensible and a betrayal of the previous 5 moves. :) Please do not trust these annoFruited ejections. Instead correct, and sensible, is 31. .. Bxf2+, which draws immediately (or wins if white deviates at all!). Here is the position after 31. .. Bxf2+ [D]6rk/3b3q/npNp2rn/2pPp3/1PP1PpP1/P1Q5/1R3bP1/3BBRK1 w - - 0 32 If white recaptures with the bishop, it is an obvious win for black immediately, I am sure all computers will confirm that. 0-1. So white must play 32. Rxf2, it does not matter which rook, black's reply will be the same. First, say, the b-Rook, 32. Rbxf2. We get: [D]6rk/3b3q/npNp2rn/2pPp3/1PP1PpP1/P1Q5/5RP1/3BBRK1 b - - 0 32 And now 32. .. Nxg4 giving: [D]6rk/3b3q/npNp2r1/2pPp3/1PP1Ppn1/P1Q5/5RP1/3BBRK1 w - - 0 33 Hydra's reply (33. Bxg4) is 100% forced, because of the terrible threat of 33. .. Q-h3! to which White must topple his king. White must immediately capture the knight an put his bishop on g4 to prevent the BQ insertion on h3. After black recaptures with the rook we get- [D]6rk/3b3q/npNp4/2pPp3/1PP1Ppr1/P1Q5/5RP1/4BRK1 w - - 0 34 Now please put that position into your computer. If it does not show the game is drawn, totally drawn, dead drawn, stone-cold dead drawn, hopelessly obviously drawn, then please find a window and throw out the computer. :) In fact, white seems to have to sac his knight to get the draw! Last case, 32. Rfxf2, here is the position after 32 .. Nxg4 again- [D]6rk/3b3q/npNp2r1/2pPp3/1PP1Ppn1/P1Q5/1R3RP1/3BB1K1 w - - 0 33 Again, white is forced to play 33. BxN, (else easy 0-1!), so we see- [D]6rk/3b3q/npNp2r1/2pPp3/1PP1PpB1/P1Q5/1R3RP1/4B1K1 b - - 0 33 And now there is an excellent move for black, 33. .. R-h6! (you see why the Q had to be on the h-file?) (it also works after the double pawn exchange, makes no difference)- [D]6rk/3b3q/npNp3r/2pPp3/1PP1PpB1/P1Q5/1R3RP1/4B1K1 w - - 0 34 And now again white must sacrifice to prevent a heavy loss. He must remove the rook blocking his king's escape, and all he can get is a draw. Now please also put that position into your computer. If it does not show the game is drawn, totally drawn, dead drawn, stone-cold dead drawn, hopelessly obviously drawn, then please find a bucket of water and use it to cool the computer. :) My personal score against Fruit is now 75+%, while I find Fritz9 and Shredder9 still very hard to defeat. Uri, I hope you see your mistake(s), and agree I was correct. You may trust the machines too much, IMO, and are too dismissive of human moves. Above I have proven myself completely correct. So I will leave you with the last words on this issue. Best, A.S.
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.