|
< Earlier Kibitzing · PAGE 3 OF 3 ·
Later Kibitzing> |
| Nov-16-11 | | YouRang: I found the general idea of <36.Qxf6 gxf6 37.Rxe7+ Kg6 38.a7> with the idea of Bb5/Bc6. However, I couldn't get past black forcing a draw after <38...Qc8>.
click for larger viewIf we play the immediate 39.Bb5, then 39...Qc1+ 40.Bf1 Qa1 and how does white make progress? If we try to create luft for the king first with 39.g3, then 39...Bb5 40.h5 (waste a move) 41.Bb5 Qc1+ 42.Kg2 Qa3 43.Bc6 Qf3+ (perpetual). I needed the computer to inform me that the right approach is to realize that black is in a zugzwang situation, and the winning move is <39.Bf1!>, leaving black with no good choices.
click for larger view
Black doesn't want to move the queen, because he needs to guard a8 and hit c1 if the bishop moves. On the other hand, white now threatens Rc7! which forces the queen to move while blocking the c-file, which enables the Bb5-Bc6 idea. e.g.
if tries to shuffle his king with 39...Kh6, then 40.Rc7! Qa8 41.Bb5 threatening Bc6  . |
|
Nov-16-11
 | | chrisowen: Famous e3? On delivery it alleging in at pa5 lumberjack sacrifice log on qxf6 gf6 rxe7 kg6 white enamel strips brack of his tidy defence good form it is Vishy aw ear you kind end ritual I see vampire slayer nosher bet in? |
|
| Nov-16-11 | | polarx: 36.Qb6 is met by Bd4.
36. Qxf6 gxf6 37. Rxe7+ Kg6 38. a7 f4
and then? |
|
| Nov-16-11 | | morfishine: <YouRang> You found the move <38...Qc8> that made me scratch my head and think "Is this the right line?" [starting with 36.Qxf6] |
|
| Nov-16-11 | | ventricule: <YouRang> Thanks for the solution ! |
|
| Nov-16-11 | | morfishine: <polarx> On your comment <36.Qb6 is met by Bd4> I was temporarily hypnotized by that move, then came to my senses when I realized that White is safe after <37.Qxd4 Rxb7 38.axb7> and here 38...Qa1+ 39.Bf1 or 38...Qxb7 and white remains a piece-up |
|
| Nov-16-11 | | sevenseaman: After the opening skirmish;
<36. Qxf6 gxf6 37. Rxe7+ Kg6> this is how my best bout with Crafty went; <38. a7 f4 39. h4 e3 40. fxe3 fxe3 41. Ke1 f5 42. Ke2 f4 43. Kf3 Kf6 44. Re6+ kg7 45. Re7+ Kf6> (White cannot let go of a7)
<46. Re6+ Kg7 47. Re7+ Kf6 48. Re6+ Kf7 49. Re7+> and Crafty claims a draw by 3 repetitions. 
click for larger viewhttp://www.chessvideos.tv/endgame-t...
This Crafty link gives position after the opening skirmish ending with <37...Kg7> and White on play. Try if interested. You never know you may outwit Crafty. It can only be good news. |
|
| Nov-16-11 | | morfishine: <sevenseaman> Thanks for Ole Crafty! Maybe 36.Qb6 was best after all :) |
|
| Nov-16-11 | | YouRang: I see that I had a typo in my post above which screwed up the moves #s: <If we try to create luft for the king first with 39.g3, then 39...Bb5 40.h5 (waste a move) 41.Bb5 Qc1+ 42.Kg2 Qa3 43.Bc6 Qf3+ (perpetual).> Should be:
<If we try to create luft for the king first with 39.g3, then 39...h5 (waste a move) 40.Bb5 Qc1+ 41.Kg2 Qa3 42.Bc6 Qf3+ (perpetual).> Also, there's another wrinkle:
After <36.Qxf6 gxf6 37.Rxe7+ Kg6 38.a7 Qc8! 39.Bf1! h5 40.Rc7!>, black can try <40...Qe8!> [diagram]
click for larger view
Here, the reasonable looking continuation 41.Ba6 (planning Rc8) is probably a draw at best after 41...e3! White needs to find <41.Rc6!>, blocking the diagonal and preparing to move behind the a-pawn. - If 41...e3, then 42.Ra6 exf2 43.Kxf2 Qa8
- If 42...Qa8, then 42.Ra6 with no stopping Bb5/Bc6 |
|
Nov-16-11
 | | NM JRousselle: The correct move for the puzzle is 36 Qb6. Qf6 looks pretty, but gives Black undeserved chances. After Qb6, Black is pretty much forced to give up his bishop for the a pawn. |
|
| Nov-16-11 | | polarx: <morfishine> You are right. I agree 36.Qb6 seems to be the best move. I found that line but did not think it was the solution because I thought there must be some brilliant and crushing move by white. I also cannot figure out why f4 was not played. But we are talking Short and Anand (!). Maybe they know better. This puzzle is very weird. |
|
| Nov-16-11 | | soberknight: I saw the idea of Qf6 but could not see all the way to a forced win. |
|
| Nov-16-11 | | YouRang: <polarx: <morfishine> You are right. I agree 36.Qb6 seems to be the best move.> If you let an engine chew on it for a while, it may say that 36.Qb4 is best. But I agree that from a human perspective, the move that wins with the fewest sneaky subtleties is <36.Qb6> (not that it's entirely trivial). Still, I rather like the 36.Qxf6! solution, which isn't too terribly difficult if one recognizes that zugzwang tactics are part of the equation. |
|
| Nov-16-11 | | CHESSTTCAMPS: I preferred the clarity of 36.Qb6! over Short's sacrificial line. Crafty EGT doesn't offer much resistance. Try it from the puzzle position: http://www.chessvideos.tv/endgame-t... |
|
| Nov-16-11 | | TheBish: Short vs Anand, 1993 White to play (36.?) "Medium/Easy"
I believe my first instinct proves to be correct.
36. Qxf6! gxf6
This is a no-brainer; not 36...Rxb7 37. Qxf5+ g6 38. axb7 Qb8 (or 38...Qa1+ 39. Bf1 gxf5 40. b8=Q) 39. Qc8. 37. Rxe7+ Kg6 38. a7 Qc8
Or 38...f4 39. h3! (making luft) Kf5 40. Bb3 f3 41. Ba4 Qxd5 42. Re8 and the pawn will queen. 39. g3! Kg5 40. Bb5 Qc1+ 41. Kg2 and Black can't stop either Bb5-c6 or Re7-e8 followed by queening the pawn. |
|
| Nov-16-11 | | stst: No clear winning line.
The obvious QxB PxQ; RxR+ Kg8/g6; a7 gives W some tempo but still far.
Trying to smother the Bk Q might be another line: Qb6, if RxR; QxR QxQ; axQ and W got a pass P. So Bk won't go this line. With this, W got the control, and Bk Q has no good spot. Yet again the line is still long. |
|
| Nov-17-11 | | morfishine: <YouRang> On your comment: <Still, I rather like the 36.Qxf6! solution, which isn't too terribly difficult if one recognizes that zugzwang tactics are part of the equation> Zugzwang tactics are what I completely neglected as I worked thru 36.Qxf6. Very good and pointed |
|
| Nov-17-11 | | polarx: Sorry, but still no solution to 38...f4. <The Bish> Crafty plays the much more aggressive 39...e6 instead of 39....Kf5. |
|
| Nov-17-11 | | YouRang: <polarx: Sorry, but still no solution to 38...f4. <The Bish> Crafty plays the much more aggressive 39...e6 instead of 39....Kf5.> After <36.Qxf6! gxf6 37.Rxe7+ Kg6 38.a7 f4>, I think we continue with the main idea: <39.Bb5!> If <39...Qxd5>, then <40.h3> making luft, and black can't handle the triple threat of Rc7/Bc6/a8=Q or Re8/a8=Q or Ba6/Bb7/a8=Q. If <39...e3> then white simply sticks to the plan with <40.Bc6>. |
|
Nov-17-11
 | | Pawn and Two: Fritz indicates the given solution for this puzzle leads only to a draw. The indicated solution given by many kibitzers, which are also the moves played by Short are: 36.Qxf6 gxf6 37.Rxe7+ Kg6 38.a7. After these moves we have the following position: 
click for larger viewFritz indicates Black can obtain a draw in this position. Here is Fritz's analysis: (2.57) (30 ply) 38...e3! 39.fxe3 Qd8! 40.Kf2 Qa5 41.Kg3 Qe1+ 42.Kf3 Qa1 43.Bb5 h5!, (2.56) (30 ply) 44.h4 Qd1+ 45.Kf2 Qd2+ 46.Be2 Qa2 47.Rb7 Qa4 48.g3 Qa2 49.Rd7 Qa1 50.Rc7 Qa2 51.Kf1 Qa5. Fritz also finds a draw after: 38...h5! 39.g3 e3! 40.fxe3 Qc8!, (2.56) (30 ply) 41.h4 Qd8 42.Bb5 Qa5 43.Bf1 Qa2 44.Rb7 Qa4 45.Be2 Qa3 46.Re7 Qa5, or (2.56) (30 ply) 41.Be2 Qc1+ 42.Kf2 Qa3 44.h4 Qa2 45.Rb7 Qa5 46.Rd7 Qa4 47.Rc7 Qa5, or (2.56) (30 ply) 41.Bb5 Qc1+ 42.Kf2 Qb2+ 43.Be2 Qa3 44.h4 Qa2 45.Rb7 Qa5 46.Rd7 Qa4, or (2.56) (30 ply) 41.Bd3 Qc1+ 42.Kf2 Qa1 43.h3 Qa2+ 44.Be2 Qa4 45.h4 Qa5 46.Rd7 Qa4. After 38...h5!, if White plays Short's move 39.h4, Fritz still finds a draw: (2.56) (30 ply) 39.h4 e3! 40.fxe3 Qd8! 41.Kf2 Qa5 43.Kf3 Qa1 44.g3 Qd1+. A winning move for the puzzle position was 36.Qb6!. |
|
| Nov-17-11 | | YouRang: But in any case, proving that this line wins goes way beyond a typical "Wednesday" puzzle, IMO. |
|
| Nov-17-11 | | polarx: <YouRang> If you play those lines you will see it is not so simple. For example: <36.Qxf6! gxf6 37.Rxe7+ Kg6 38.a7 f4 39.Bb5, Qxd5.
40.h3, e3>
And then:
<41. Re8, e3xf2+
42.Kf1, Qxb4+
43.R2e, Qa6>
<41. Re8, e3xf2+
42.Kh1, Qd1+
43. Kh2, Qg1#>
<41. Re8, e3xf2+
42. Kxf2, Qd4+
43. Kf1, Qxa7>
<41. Rc7, Qd1+
42. Kh2, exf2
43..., Qg1#>
<41. Rc7, Qd1+
42. Bf1, e2
43. a8=Q, exf1=Q+
44. Kh2, Qh1#>
<41. Ba6, exf2
42. Kxf2, Qd4+
43. Kf1, Qa1+
44. Kf2, Qxa6>
etc.
Anyway, yes, way beyond a Wednesday puzzle. |
|
| Nov-17-11 | | YouRang: <Pawn and Two: Fritz indicates the given solution for this puzzle leads only to a draw.> You may be right, lol.
Nice work -- Until shown otherwise, I consider the 36.Qxf6?! "solution" to be busted. |
|
Nov-18-11
| | Pawn and Two: <YouRang> Thanks for your comment. I used Fritz 13 on my new computer, and it provided the analysis in an amazingly short amount of time. Fritz's evaluation actually went flat several ply before the 30 level. This was an unusual puzzle, expecially for a Wednesday level puzzle. After 36.Qxf6?! gxf6 37.Rxe7+ Kg6 38.a7, it should be, "Black to play and draw", but that would truly be an "Insane" level puzzle. On the other hand, the clear winning move 36.Qb6!, is probably not worthy of even a Wednesday level puzzle. |
|
| Nov-18-11 | | YouRang: <Pawn and Two><On the other hand, the clear winning move 36.Qb6!, is probably not worthy of even a Wednesday level puzzle.> I think this is one of those cases where, if I came across this position in a game, I would probably play 36.Qb6. But when presented as a puzzle, I looked for something "flashier", and found 36.Qxf6 (even though, as admitted above, I wasn't able to make it work myself). Interestingly, another solution may be 36.Qb4 -- this was the favorite of Houdini at sufficiently high ply, but the benefits of this move are not as obvious to the human mind. |
|
 |
|
|
< Earlier Kibitzing · PAGE 3 OF 3 ·
Later Kibitzing> |
Kibitzing
sac: 36.Qxf6