< Earlier Kibitzing · PAGE 2 OF 2 ·
|Oct-10-09|| ||VincentL: For what it's worth, I tried the following:
Material is even.
White's three pawns on the fourth rank can participate in an attack on the black king.
With the doubled rooks, if the white queen can get to a8, it will cause damage (and may capture the rook on h8).
So, I think perhaps 21. Bxf6 gxf6
Now I want to play 22. b5.. but I think this line is now wrong.
Let us try fist 21 e5. Now 21.... Bxg2 22. exf6 is unpleasant, so I think black must
play 21. Bxe5, after which follows 22. Bxe5 dxe5
But this does not lead anywhere.
Perhaps white needs to get his rook to d1. The knight on d2 needs to move away.
What about 21. Bxf6 gxf6 22. c5 Now white will be able to move the knight to c4 threatening the queen.
22.....dxc4 23. Nc4
Black cannot play 23... Qb5 because after 24. Qxd6+ the black queen will be lost.
After 23.... Qc6 we have 24. b5 axb5 25. cxb5 Qc7. Now I don't see the way forward.
I am missing something.
|Oct-10-09|| ||gawain: <Al Wazir> You are right. The payoff is at move 27. |
If 27 Nc4 Qc7 then 28 Qxb7! is best because after 28...Qxb7 29 Nd6 forks the K and Q.
|Oct-10-09|| ||VincentL: Some right ideas, but move order effectively wrong, and so I did not get to the fork after 25. c6 and still later find 27. Nc4.|
Very difficult, I think, and there must be a lot more lines than this.
|Oct-10-09|| ||pts92: I am amazed of the deepness of Grandmaster's ideas. I never could have seen the variations arising from 21. e5. I thought of the move, but even the 2nd move of the variations gave me a hard time, what to say about the rest of 5-6-7 moves thought in advance by Salov :)) .|
|Oct-10-09|| ||VincentL: What happens if black plays 22... Qb5 ?|
|Oct-10-09|| ||TiTi: Although I can't explain, I found the opening move directly. Seems like a very intuitive move.|
|Oct-10-09|| ||cyclon: Very difficult.|
|Oct-10-09|| ||dzechiel: White to play. Material even. "Very Difficult."
It seems pretty obvious to me that white will be playing c5 in the near future. The question is, on move 21, or some time later? White could choose to exchange a pair of bishops first with 21 Bxf6.
After looking at the position a little longer, it appears to me that both c5 and e5 will be played, and you don't want the black knight to be able to occupy d5 once those moves have been made, so the exchange of bishops now looks to be off.
That means we start with
Originally I wanted to start with 21 c5, but it just wasn't working for me. This move looks a lot more promising. With this move, we attack the bishop on f6 AND open lines on the h1 to a8 diagonal, keeping the black queen off of c6.
Black can't really capture on e5 as after 21...dxe5 white gets to play 22 c5 and 23 c6 picking up a piece. Likewise, if 21...Bxe5 22 Bxe5 dxe5 23 c5 and 24 c6.
I also don't think black should try 21...Be7, as after 22 exd6 Bxd6 23 Bxg7 Rg8 24 c5 Qb5 25 Qxb5 axb5 26 Bxb7 Rxg7 27 c6 Ne5 28 c7 black is once again in a world of hurt.
Should black capture or move the queen? On
22...dxc5 23 Nc4
attacks the queen and there isn't really a good square to move to. After
Not 23...Qb5 24 Nd6+ and 25 Nxb5.
24 Nd6+ Ke7 25 Nxb7 Rxb7 26 Bxb7 Qxb7
and white goes up an exchange. If, back on move 22 black decides to move the queen instead, say
It still looks to me like white has lots of good stuff. I would like to finish this up, but it's midnight here (I got back late from the Angel - Red Sox game) and I'm running out of steam.
Time to check and see how this all went down.
Whoa! A lot more going on than I suspected.
|Oct-10-09|| ||moi: What is wrong with 21 ... Bxg2, eg
22 exf6 Bb7
23 fxg7 Rg8
24 and now? Black threatens Qxb4 or e5 or a strong attack along the diagonal a8-h1.
Did I miss something?
|Oct-10-09|| ||elgato: Black was exploiting the fact that the king hadn't castled, so that in any exchange of pieces the b8 rook couldn't leave the 8th row or the a8 rook would fall. Simple in theory, but I still missed it :(|
|Oct-10-09|| ||Samagonka: Without thinking too hard, I can see that the d6 pawn has a lot to do with the solution...|
|Oct-10-09|| ||lentil: No marks for me, except perhaps for charity. I am amused though. I did guess that B's previous move was ...Bf6 and did decide that 21. e5 was probably the move, taking advantage of the attack on the B/f6 to gain time to trade off the B/b7. No genius required. I also remembered LPDO, so I looked for loose pieces, but saw that only the deeply-hidden R/h8 was loose and discarded the idea. But then, sure enough, the combo exploited it: 27 Nc4! (the sting); B can't capture the N because of a check at a8, and the loose rook drops off. These guys don't operate in the realm of us normal (?) folk.|
|Oct-10-09|| ||patzer2: For today's Saturday puzzle solution, Salov's clearance sacrifice 21. e5!! discovers a winning attack on Karpov's Queenside after 21...Bxe5 22. c5! .|
<21. e5!! Bxe5>
21... dxe5 22. Rd3! Bc8
(22... Bxg2 23. Qxd7+ Kf8 24. b5 Rd8 25. Rxa6 ;
22... Qc7 23. Ne4 Bc8 24. Nd6+ Ke7 25. Rad1 Rd8 26. Ba3 Nb6 27. Nxc8+ Rdxc8 28. b5+ Ke8 29. Qxa6 Qxc4 30. Bd6 e4 31. Bxb8 Rxb8 32. Rd6 Qa4 33. Bxe4 Qxe4 34. Rxb6 Rd8 35. Rbd6 Ra8 36. Qb6 )
23. Ne4 Be7 24. Bxe5 Qxb4 25. Bxb8 Qxa4 26. Rxa4 Nxb8 27. c5 Bd7 28. Rad4 Bb5 29. Rd1 e5 30. Nd6+ Bxd6 31. Rxd6 .
<22. c5! Qc7>
22... Qb5 23. Qxb5! axb5 24. Bxe5!Bxg2 (24... Nxe5 25. Bxb7 ; 24...
dxe5 25. c6 ) 25. Bxd6 Bd5 (25... Rc8 26. Kxg2) 26. Bxb8 Nxb8 27. Ra5 Bc6 28. Nb1 Ke7 29. Nc3 ;
22... dxc5 23. Bxe5 Bxg2 24. Bxb8 .
<23. Bxe5 dxe5 24. Bxb7 Rxb7 25. c6 Nb6 26. Qxa6 Qxc6>
26... Rb8 27. Qa7! Nd5 28. Qc5 Qe7 29. Qxe7+ Kxe7 30. Ra7+ Kf6 31. b5 Rxb5 32. Ne4+ Kg6 33. Nd6 Rb6 34. Nxf7 Rc8 35. Nxe5+ Kf5 36. f4 Rbxc6 37. Rf7+ Nf6 38. Nxc6 Rxc6 39. Rxg7 .
<27. Nc4 Rb8 28. Nxb6 O-O 29. Nc4 Qe4 30. Re3 Qd5 31. Nxe5 Rfc8 32. Qd3 1-0>
|Oct-10-09|| ||patzer2: Another move worthy of note is Salov's 27. Nc4!, set up by 26. Qxa6!, which combines the pin, deflection and skewer tactical threats to overload and win a decisive piece.|
|Oct-10-09|| ||lost in space: I haven't got it within my normal time limit. I don't want to bore you with my lines; they started with 21. c5; no concrete outcome. I was aware, that 21. e5 is an alternative.|
|Oct-10-09|| ||David2009: I couldn't solve this one (Salov vs Karpov, 1998, 21?) within a reasonable time. <dzechiel> shows how the solution can be discovered logically from first principles. In the game I would probably have played Bxf6, then centralised a Rook.|
|Oct-10-09|| ||Bishoprick: Of course, this is Karpov past his prime. Once upon a time, it would have been a different story!|
|Oct-10-09|| ||moi: Ah ok: for those who are interested,
21 ... Bxg2 is wrong because of 22Kxg2
and the pawn on d6 has to go away or be taken, and white's c pawn becomes way to strong.
Maybe I overlooked a stronger continuation for white?
|Oct-10-09|| ||johnlspouge: Saturday (Very Difficult)
Salov vs Karpov, 1998 (21.?)
White to play and win.
Material: Even. The Black Ke8 has 3 legal moves, all dark squares. The White Qa4 pins Nd7 to Ke8. White has a triple battery Ra1, Ra3, and Qa4 threatening Pa6, which Bb7 protects. The loose White Bb2 faces Bf6, while the White Bg7 x-rays Bb7 through Pe4, suggesting a clearance with attack. The White Nd2 requires activation. The White Kg1 is secured from all checks except the pointless 21…Qxf2+.
Candidates (21.): e5, c5
21.e5 (threatening 22.exf6)
(1) Black can counter-attack:
21…Bxg2 22.exf6 Bc6 [or Bb7 or Ba8] [else, drop a B for a P]
White has won a P and has the advantage of superior development in a dangerous midgame with Bs of opposite color.
(2) Black can capture:
21…Bxe5 [dxe5 is similar] 22.Bxe5 dxe5
The White Pc4 is now free to advance.
23.c5 (threatening 24.c6, forking Bb7 and Nd7)
(2.1) 23…Qb5 24.Qxb5 axb5 25.Bxb7 Rxb7 26.c6
White wins N for 2P.
(2.2) 23…Qc7 [Qa7 or Qd8 are worse]
24.Bxb7 (threatening 25.c6, forking b7 and Nd7)
24…Rxb7 [Qxb7 25.c6 Nb6 26.cxb7 Nxa4 27.Rxa4 Rxb7 drops N for 2P]
25.c6 Nb6 [else, drop N for 2P]
(After 23…Qa7 or 23…Qd8, 26.c7+ would have won at least N for 2P.)
26.Qc2 (threatening 27.cxb7 Qxc2 28.b8=Q+)
The threat gives White a tempo to reinforce Pc6 with 27.Rxa6 28.b5. The connected passers Pb5 and Pc6 provide an overwhelming advantage.
At 18 plies, Toga prefers the game 26.Qxa6 (+2.48 P) to my 26.Qc2 (+1.98 P).
|Oct-10-09|| ||johnlspouge: Toga takes a long time to decide that 21.e5 is better than 21.Bxf6. Humans can improve near the end of the complete computer variations.|
[ply 15/46 time 01:08 value +0.95]
21.Bxf6 gxf6 22.Rd3 Bc6 23.Qa3 Qxb4 24.Qxb4 Rxb4 25.Rxa6 Bb7 26.Ra7 Rg8 27.f4 Nc5 28.Rxd6 Rb2 29.Bf3 Ke7 30.e5 fxe5 31.fxe5 f5 32.Bxb7 Rxb7
[ply 15/50 time 02:07 value +1.49]
21.e5 Bxe5 22.c5 Qc7 23.Bxe5 dxe5 24.Bxb7 Rxb7 25.c6 Nb6 26.Qxa6 Rb8 27.Qa7 Qc8 28.b5 0-0 29.Ne4 Nd5 30.Rc1 Rxb5 31.Nd6 Qb8 32.Nxb5 Qxb5
I never seriously considered 21.Bxf6, but like <dzechiel>, I spent about a minute sampling variations from 21.c5 before I decided that there was nothing there.
|Oct-10-09|| ||johnlspouge: Move order is important. My suggestion of 22.Bxf6 23.c5 (vs. the game variation 22.c5 23.Bxf6) permits 22.Bxg2, which leaves White with a bare winning advantage of +1.2 P.|
Psychologically, this reflects incomplete digestion of a puzzle position with human wetware. There is no reason to permit 22.Bxg2, so the choice of move order is a no-brainer, but I was disoriented by trying to analyze both 22...dxe5 and 22...Bxe5 simultaneously.
|Oct-10-09|| ||WhiteRook48: First I looked at 21 e5, 21 c5, and then decided 21 Qxd7+ |
|Oct-10-09|| ||wals: 21.c5 played out by Rybka 3 results in a draw because of insufficient material. White finishes King and Knight v King.|
|Oct-10-09|| ||Eduardo Leon: I didn't analyze too deeply, I just saw
which was pretty much self-suggesting, as <22. c5> is coming.
<21. ... xe5>
The other possibility, <21. ... dxe5> frees white from the task of deciding when to exchange the black squared bishops and accelerates black's death by one move.
Not <22. xe5? xg2! 23. xg2 b7+>, and black defends successfully.
<22. ... c7 23. xe5>
This is the right moment to exchange the black squared bishops. Now <23. ... xg2? 24. xd6 b7 25. c6> wins a piece.
<23. ... dxe5 24. xb7 xb7>
Not <24. ... xb7? 25. c6>.
<25. c6 b6 26. xa6>
Black can't capture the c6 pawn, because it would lead to the loss of a piece: <26. ... xc6? 27. c4!>. Therefore, white wins with his two passed pawns in the queenside.
|Oct-10-09|| ||newzild: Got it - after missing one earlier in the week and getting the move-order wrong in another.|
< Earlier Kibitzing · PAGE 2 OF 2 ·