< Earlier Kibitzing · PAGE 4 OF 4 ·
Later Kibitzing> 
Aug2111   sevenseaman: <morfishine> Cheer up. It now seems your <37. Bh6> is the flavor of the day. See my latest analysis above and check it for holes if any I've overlooked. You might, after all, have rounded off the week nicely. 

Aug2111   patzer2: <stukkenjager> <36.Nf6+ gxf6 37.exf6 Qf8 (37...Kh8 38.Bh6!) 38.Bf4! game over> I agree you've found the strongest winning line, improving over the game continuation 38. Qg3+ with 38 Bf4! . However there's a little technique involved in trying to force the win from this point. Here's my attempt at playing out your improvement with Fritz: <36. Nf6+ gxf6 37. exf6 Qf8> 38. Bf4! Rc8 (38... Nd7 39. Qg3+! Kh7 40. Qh4! ) 39. Rxc8 Qxc8 40. Qxb5 Nc4 41. Kh2 Kh7 42. Qc5 Qa8 43. b5 Kg6 44. Qc7 Qa5 45. Qxa5 Nxa5 46. Bc7 Nb7 47. Be5 h4 48. b6 Kg5 49. f4+ Kh5 50. g3 hxg3+ 51. Kxg3 Na5 52. Bd6 Kg6 53. Be7 Kf5 54. h4 Nb7 55. h5 Na5 56. h6 Kg6 57. Bf8 Nb7 58. Kh4 Na5 59. f5+ exf5
60. Kg3 Kh7 61. Kf4 Kg6 62. Bg7 Nb7 63. Ke5 Na5 64. Kxd5 f4 65. Ke4 Nc4 66. b7 Nd6+ 67. Kxf4 Nxb7 68. Ke5 Na5 69. Kd6 Nb3 70. d5 Nd4 71. Ke7 Nf3 72. Kf8 Ng5 73. d6 Kh7 74. d7 Kg6 75. Kg8 Ne6 76. h7 Nd8 77. h8=Q . 

Aug2111   sevenseaman: <patzer2> A good hardworking piece of analysis in coordination with <stukkenjager>. Enjoyed playing it out. 

Aug2111   Marmot PFL: Didn't seem all that hard. 36 Nf6+ gf 37 ef Qf8 (37...h4 38 Qe3) 38 Qg3+ Kh8 39 Bh6 Qxh6 (Qg8 40 Bg7+ Kh7 41 Qd3 mate) 40 Qxb8+ Kh7. Now white moves the rook to a safe square, and wins a knight. I missed the cute rook sac 42 Qg3 but the rest was fairly routine. 

Aug2111   morfishine: <sevenseaman> Yep, gotta look at the positives. In my case (1) I got the first move right and (2) I was successful in refuting all of my flawed continuations! :) [and kudos to your 'eureka' getting thru to <40.Qxb8>] The main problem after 37.Bh6 was <37...h4> buying black just enough time to defend. It all makes sense now as <37.exf6> sets up the deflection shot <39.Bh6>. Really, a nice set of deflections all week. Good work <stukkenjager> and <patzer2> running this out 

Aug2111   Schach and Awe: To test my ideas, beginning with Nxf6, I loaded Rybka 4 to play the Black side. The game proceeded as follows ... 36. Nxf6 gxf6 to open up the Kingside
37. exf6 to create a thorn pawn, restricting the mobility of the K 37. .. Qf8 selected by Rybka as in the game, expecting white to reply with Bf4 attacking the R 38. Qg3+ Kh7 I could have elected to recover material with QxNb5, but after Nc4, the Q is under attack by the R, and I could see no future for White. 39. Qf3 Qh8 Attacking the hanging Rook pawn, and I suspect Rybka played Qh8 offering a QxQ exchange or more likely forcing my retreat (given I have already sacked a piece). 40. Bf4 R?? Declining the hanging pawn, attacking the Rook with the aim of driving it off of the cfile. 40. .. if Rd8, Qd3+ picking up the N and attacking Nb6, or ... 40. .. if Rc7, same sequence, but with more dire consequences for Black, as the K would be forced to g8, leading to a number of threats and attacks that include Qe8+ and back rank attacks involving White's R should N be convinced to vacate b6. Not a mate, and not anything as convincing as the text, but still winning in my opinion. 

Aug2111   LIFE Master AJ: I started at 9:49 AM, it was amazing to see that I quickly used over an hour on my analysis.
(Of course, in a tournament game, you almost never have this much time to ponder your moves, especially not at today's <modern, quicker> time limits.) However, the main idea (for me  esp. on a Sunday) is to take as much time as you can, work as hard as you can ... and see what you can come up with. click for larger viewWithin just a few minutes, I had quickly identified what HAD to be the first move ... it fits Kotov's definition perfectly. (Its a K+Q fork, AND its a check.) The only problem was making sure everything works ... 36.Nf6+!! gxf6; 37.exf6!,
This must be best, although at first, I started looking at 37.Bh6. Now White has some simple mating patterns, (Bh6, Bg7, & Qd2); which Black MUST prevent! [ Worse is: </= 37.Bh6?! h4!; This denies the WQ access to g3. 38.Qf3 f5; 39.Qh5 f6; ∞ ] 37...Qf8T; (Completely forced.)
This move is a necessity for Black, otherwise the White QB goes to h6. [ MUCH worse was: 37...Kh8?; 38.Bh6, " " (or) </= 37...h4?!; 38.Qe2! Qf8; 39.Qg4+ Kh7; 40.Qh5+ Kg8; 41.Bh6, " " and White is winning easily. ] 38.Qg3+ Kh8▢; This is also forced.
[ But not: </= 38...Kh7?; And now the simplest is: 39.Qd3+, and White wins back the Knight  with interest. " " ] 39.Bh6, This is not what I saw ...
[I THINK that better was: 39.Qe5!,
(This hits h5 while keeping the b8square under observation ... playing the computer the last few days has taught me a lot about the value of forks and double attacks.) 36...Qg8; It seems Black must defend the King, especially the g7square. 40.Bh6 Kh7; 41.Bg7 Kg6; 42.h4 Qd8; 43.Qg5+ Kh7; 44.Qh6+ Kg8; 45.Qh8#. ] I did not even look at the rest of the game. (Maybe later, after chess club.) 39...Qxh6; 40.Qxb8+ Kh7; 41.Re1 Qd2; 42.Qg3 Qh6; 43.Qd3+, 10 

Aug2111   Marmot PFL: <38.Qg3+ Kh8only move; This is also forced.
[ But not: </= 38...Kh7?; And now the simplest is: 39.Qd3+, and White wins back the Knight  with interest. " " ] 39.Bh6, This is not what I saw ...
[I THINK that better was: 39.Qe5!, >
39 Qe5 is a big improvement, on the other hand 38...Kh7 was playable, as 39 Qd3+ and Qxb5 is met with 40...Nc4 and 41...Nxd2. Now I see why this puzzle was so difficult. 

Aug2111   morfishine: <LIFE Master AJ> On <However, the main idea (for me  esp. on a Sunday) is to take as much time as you can, work as hard as you can ... and see what you can come up with.> I'm going to take that tack from now on. I've been on a 51020 (5min for Mon/Tues; 10min for Wed/Thurs; 20min for Fri/Sat/Sun) time limit. Your idea is more sensible as the benefit derived outweighs the stricture of an artificial timepressure: After all, these are not our games, but positions [to explore]. Oh, and thanks for the keen analysis as it completely supports all of the refutations of bad lines I had manage to selfbust! 

Aug2111   patzer2: <LIFEMaster AJ> <38.Qg3+ Kh8 only move; This is also forced. [ But not: </= 38...Kh7?; And now the simplest is: 39.Qd3+, and White wins back the Knight  with interest. " " ]> I think you're right about <36.Nf6+ gxf6 37.exf6 Qf8 38. Qg3+> 38...Kh7 39. Qd3+ (diagram below) leading to a White win, even if it's not clear to me that it's objectively worse than 38...Kh8, especially after your strong recommendation 38...Kh8 39. Qe5!! (stronger than 38...Kh8 39. Bh6! ). click for larger view(position after <36.Nf6+ gxf6 37.exf6 Qf8 38. Qg3+> 38...Kh7 39. Qd3+) In the above position, if Black now plays 39...Kg8 he sets a cheap trap. White can't capture the Knight immediately as 39...Kg8 40. Qxb5?? drops a piece to the discovered attack 40. Nc4! (diagram below). click for larger viewInstead, after 38...Kh7 39. Qd3+ Kg8, the in between move (i.e. zwischenzug) 40. Bf4! (diagram below) removes the Bishop to a safe and stronger position. click for larger view(position after <36.Nf6+ gxf6 37.exf6 Qf8 38. Qg3+> 38...Kh7 39. Qd3+ Kg8 40. Bf4!) In this position (identical to <stukkenjager>'s 38. Bf4! I might add), if Black defends with 40...Nc4, the "obvious 41. Bxb8 wins slowly. However, even stronger after 40. Bf4! Nc4 is 41. Qg3+ Kh7 (41...Kh8 42. Qh4! ) 42. Qg5! Nxd4 43. Qxh5+ Kg8 44. Bh6! . 

Aug2111
  agb2002: White has a bishop and a pawn for a knight.
Black threatens 36... hxg4.
White can further weaken the black castle with 36.Nf6+ gxf6 37.Bh6, menacing mate in two: A) 36... Kh8 37.exf6
A.1) 37... Nxd4 38.Qxd4 followed by Rc3 and Rg3.
A.2) 37... Nd6 38.Qg3 Nf5 39.Bg7+
A.2.a) 39... Kg8 40.Qg5 Nc4 41.Rc3 and 42.Rg3 with the threat 43.Bf8+ and 44.Qg8#. If 42... Nxg3 43.Bh6+ and 44.Qg7#. A.2.b) 39... Kh7 40.Qg5 Kg8 41.Rc3 as in A.2.a.
A.3) 37... Qg8 38.Bg7+ wins.
A.4) 37... Rc8 38.Bg7+ Kg8 39.Rxc8 Nxc8 40.Qd1 Ne7 41.Qxh5 Ng6 42.Qg4 followed by h4h5. B) 36... h4 37.exf6
B.1) 37... Nc4 38.Qe2 Kh7 39.Qh5 Qh8 40.Bf4+ Kg8 41.Qg5+ Kf8 42.Bxb8 + . B.2) 37... Nd6 38.Qd2 Nf5 39.Qg5+ Kh7 40.Bg7 followed by 41.Qh5+ winning. 

Aug2111   David2009: B McCamon vs B Carlier, 1993 White 36? Insane I am coming to this puzzle vary late. The sane continuation is 36.Ne3 threatening Rc5 and staying a solid Pawn up. The insane continuation is 36.Nf6+ gxf6 37.gxf6 Qf8 (if Kh8 38.Bh6 looks strong)
38.Qg3+ Kh8 (if 38...Kh7 39.Qd3+ picks up the Nb5) and now 39.Bh6 looks good since 39...Qxh6 allows 40.Qxb8+.
The alternative 37.Qg3+ seems less good because after 37...Kh8 38.exf6 Black can play 38...Qg8
I haven't time to analyse all the variations upfront today, so I'll
check:
====
36.Nf6+ gxf6 37.exf6 was indeed played, time to explore the variations with Crafty End Game Trainer (link http://www.chessvideos.tv/endgamet...)
starting from the puzzle position:
click for larger viewI see that <sevenseasman> has anticipated my analysis and also found the game line up till move 40.
The EGT defends differently: 38...Kh7 39.Qd3+ Kh8 40.Bf4 (if 40 Qxb5 Nc4! wins a piece) e5! 41.Bxe5 Nd6 42.Qg3 Rc8 and Black stays a piece ahead.
My alternative try 36.Ne3 runs into 36...Ra8 with the plan of Na3 and Nb5 and White gets stuck. After reading the kibitzes, I have tried out <patzer2> and <stukkenjager> line.
The EGT defends with 36.Nf6+ gxf6 37.exf6 Qf8 38.Bf4 Rc8 39.Rxc8 Qxc8 40.Qxb5 and now continues 40...Nd7. White is two Pawns up, so there ought to be a win somewhere. I continued 41.Be5 Nxe5 42.dxe5 Qc1+ 43.Qf1 Qb2 to reach click for larger view when I got stuck. The win is there and I leave it to the interested reader to find (I needed silicon help). Additional Crafty EGT link to this final position: http://www.chessvideos.tv/endgamet... 

Aug2111
  agb2002: 37.Bh6 is a blunder which loses practically all the advantage after 37... h4 38.exf6 Nd7 39.Qe3 Nf8. Better luck next week. 

Aug2111   Vladimir Zukhar: in game black make mistake.
should play 38 ... KR7 (not R8)
this prevent BR6
but white is still winning.
39 QB3 (QxP+) ... KN3
40 QK3 (KN5+) ... KR2
41 QK2 (QxP+) ... KN3
42 QK5 (QN5+)... QR1
43 RN1 ... NB6 (not good, but other moves are worse) (black is in zug no good moves, white play waiting game) 44 QN5+ ... KR2
45 BxN ... NQ2
46 QxP+ ... KN1
47 QN4+ ... KR2
(if ...KB1, PN5
48 BQ2 ... NxP
(nothing better for black?)
49 QR4+ ... KN2
50 QN3+ ... KR2
51 BB4 ... RN2
52 RQB1 ... NK5
53 QR4+ ... KN2
54 QN4+ ... KR2
55 PB3 ... NQ3
(upon ... NB3, then QR4+ and Black soon gets mated) 56 BxN ...
(and black can safely Resign) 

Aug2111   Vladimir Zukhar: from above:correction
should play 38 ... KR2 (not R1)
this prevent BR6 

Aug2111   jamesmaskell: The move is fairly easy but I had no chance with the continuation itself. 

Aug2111   stst: late to game, insane day with insane approach:
36.Nf6+ (a sac, try to fork and divert g7 P) Kh8 (else Qg3 onto g7 will mate)
37.exf6 to tighten the rope, also making the diag g3b8 thru, Qg8 (trying to defend Qg3)
38.Bg7+ QxB (only way)
39.fxg7 KxP
40.Qg3+ Kany
41.QxR and B has nothing to withstand W's Q&R material adv. 

Aug2111   sevenseaman: <Vladimir Zukhar> has given a sound looking continuation for dealing with 38...Kh7 response(in lieu of the tame 38...Kh8). To enable easier understanding by most posters here, I have taken the liberty of converting his moves(classical notation)to the more commonly used algebraic notation. (If any errors have crept in, I hope V. Zukhar will kindly correct them or you may verify from his original post above). After <36. Nf6+ gxf6 37. exf6 Qf8 38. Qg3+> V. Zukhar here gives;
<38...Kh7(not Kh8) but White wins with; 39. Qf3 (Qxh5+) ... Kg6
40. Qe3 (KN5+) ... Kh7
41. Qe2 (QxP+) ... Kg6
42. Qe5 (QN5+)... Qh8
43. Rb1 Nc3 (not good, but other moves are worse)>? <44. Qg5+ Kh7
45. Bxc3 Nd7
46. Qxh5 Kg8
47. Qg4+ Kh7
(if ...Kf8, Pb5
48. Bd2 Nxf6
(nothing better for Black?)
49. Qh4+ Kg7
50. Qg3+ Kh7
51. Bf4 Rb7
52. Rc1 Ne4
53. Qh4+ Kg7
54. Qg4+ Kh7
55. f3 Nd6
(upon ... Nf6, Qh4+ and Black soon gets mated)
56. Bxd6 10
(and black can safely resign)>
I hope this finally takes care of the tricky 38...Kh7 response in lieu of 38...Kh8. It has been one of the toughest of Sunday puzzles. 

Aug2111   sevenseaman: <David2009> This Sunday has been really 'insane'. Not all lines, including the ones vetted by the various analysis engines pass muster with Crafty. I am learning to be selective and not using the software in a 'catchall' manner. Nevertheless I need to thank you for the facility you offer every day and maintain consistent allegiance to this handy defensive tool. 

Aug2111
  Richard Taylor: Velly cleffer! 

Aug2211   patzer2: <David 2009> I found <36.Nf6+ gxf6 37.exf6 Qf8> 38.Bf4!, <stukkenjager>'s improvement, beats Craft after 38...Rc8 39. Rxc8 Qxc8 40. Qxb5 Nd7 (diagram below) click for larger view41. Qd3! Nxf6 42. Qg3+ Kh7 43. Qg5 Nd7 44. Qxh5+ Kg8 45. Bh6 Qc2 46. Qg5+ Qg6 47. Qd8+ Kh7 48. Be3 Nf6 49. b5 . 

Aug2411   Vladimir Zukhar: 38.Qg3+ Kh8only move; This is also forced.
[ But not: </= 38...Kh7?; And now the simplest is: 39.Qd3+, and White wins back the Knight  with interest. " " ] 
this move KR1 is not forced.
KR2 is still preferred here.
upon ...KR2
if W now 39. QQ3+ ...KN1
W can pick up N, but this not accomplish much here
after 40. QxN then ...NB5
and B is winning
for ex.
41. QQ7...NxB
42. QB7...NK5
etc. 

Sep1711   tacticalmonster: 36 Nf6+ gxf6 37 Bh6 Kh8 38 exf6:
a) 38 Nd7 39 Qg3 Qg8 40 Bg7+ Kh7 41 Qg5 Qe8 42 Qh6+ Kg8 43 Qh8# b) 38 Nd6 39 Qe2 Nf5 40 Qxh5 Kg8 41 Rc3 Nd7 42 Rg3+! Kh2 43 Bg7+ Nh6 44 Qxh6+ Kg8 45 Qh8# 

Sep1711   tacticalmonster: I was wrong in my analysis:
36 Nf6+ gxf6 37 Bh6 h4 ! 38 exf6 Nd6 39 Qf3 Nf5 40 Qg4+ Kh7 41 Bg7 Kg8 42 Rc3 Nc4 43 Rf3 N4d6 44 Bh6+ Kh7 45 Bg7 Kg8 = 

Sep1711   tacticalmonster: Just to add to my previous analysis: 43 N4d6 White can try for a win with 44 Rf4 Rxb4 a) 45 Qg5 Rb1+ 46 Kh2 Qb8! 47 Rxh4? Nxh4 48 Qh5 Ndf5+ 49 g3 Nxg7 b) 45 Qxh4 Nxg7! 46 fxg7 (46 Rg4 Nf5 47 Rxg7+ Nxg7 48 Qg5 Qf8 ) f5! 47 Qh8+ (47 Qh6 Nf7 ) Kf7 48 Rh4 Qg8 c) 45 Qh5 transpose to b) 



< Earlier Kibitzing · PAGE 4 OF 4 ·
Later Kibitzing> 