< Earlier Kibitzing · PAGE 2 OF 2 ·
Later Kibitzing> 
Aug0209
  OBIT: So, looking at the game continuation, White sacks two pieces, then plays a quiet move? Holy cow. Before deciding on 26. Rxh6 OTB, I think the only moves I'd have to analyze would be the captures of the knight, figuring if Black cannot capture the knight, then Rxh6 is tactically justified. Ruling out 26...Nxf4 27. Rxh7+ Kxh7 28. Qh4+ Kg6 29. Qh6+ Kf5 30. Qf6# isn't difficult, so that leaves the game continuation ending with that remarkable quite move 30. Re1. When I looked at this position for a few minutes, I decided on 30...Rxg5, but I see this loses to 31. g4+ Rxg4 32. fxg4+ Qxg4 33. Nd4+! Qxd4 34. Qh5+ Kh4 35. Qh4+, winning the queen. Definitely an amazing game... 

Aug0209   Summerfruit: Material is even.
White Nf4 is there to be taken, but black has no time for that after: 26.Rxh6 threatening 27.Rxh7+ Kg7 28.Qh4+ Kg7 29.Qh6#. Black can try the following defences:
a) 26...Nxf4 27.Rxh7+ Kxh7 28.Qh4+ Kg6 (Kg7 29.Qh6# or Nh5 29.Qxh5+ Kg7 30.Qh6#) 29.Qh6+ Kf5 30.Qf6# b) 26...Qxf4 27.Rxh7+ Kxh7 28.Qh3+ Kg6 (Kg7 28.Qh6#) 29.Qh6+ Kf5 30.Re1 Now the black king cannot escape via e5, e.g.:
b1) 30...Ne3 31.Qf6#
b2) 30..Qe3 31.Rxe3 Nxe3 32.Qf6#
b3) 30...Qxg5 31.Nd4+ Kf4 32.Re4+ Kg3 33.Qh3+ Kf2 34.g4 and white wins. b4) 30...Rxg5 31.Qh7+ Rg6 32.g4+ and white wins.
b5) 30...Rg6 31.g4+ and white wins.
c) 30...Kg7 31.Rxh7+ Kxh7 (Kf8 32.Rxc8+ winning) 32.Qh4+ Kg7 33.Qh6# d) 30...Rg7 31.Rc8+ winning.
e) 30.Qxe5 31.Ng6+ winning.
f) 30...Rd8/Re8/f6/f5 31.g6 and white has a strong attack. 

Aug0209   Samagonka: Was nowhere near the solution...crazy stuff this. I'm going back to my Guinness... 

Aug0209
  johnlspouge: Sunday (Insane):
B Ivanovic vs G Ginsburg, 2003 (26.?) White to play and win.
Material: N for B. The Black Kh8 has 1 legal move, g7, which Qg3 xrays through the White Pg5. Both the White Pg5 and Rh1 can capture Ph6. The White Rh1 attacks Bc8, burdening Rg1 and making 26.gxh6 an interesting candidate. The White Nf4 can play 26.Nf4g6+. Black threatens 26…Nxf4 or 26…Qxf4. The White Kb1 is vulnerable to 26…Qe4+ (stopped by Pf3), 26…Qd3+ (stopped by Nf4), and 26…Nc3+ (stopped by Pb2). Candidates (26.): gxh6, Rxc8, Qh4, Rxh6
[26.gxh6 Rxg3 27.Rxc8+ Rg8 covers the threat of 28.Qg7#] [26.Rxc8 Qxf4 27.Rxg8+ Kxg8 28.Qxf4 Nxf4 dissipates the attack] Although Black has a weak gfile and back rank, the two weaknesses even together are not fatal. The semiopen hfile now provides the most promising candidates: [26.Qh4 Qxf4 27.Qxh6 Qf5+ and Black holds again]
The position of Rg8 and Pf7 suggests a dovetail mate, however, so perhaps Rh1 should precede Qg3: 26.Rxh6 (threatening 27.Rxh7+ Kxh7 28.Qh3+ Kg7 29.Qh6#) Black cannot capture Nf4 to open a flight square at g6: (1) 26…Nxf4 27.Rxh7+ Kxh7 28.Qh4+ Kg6 [Kg7 29.Qh6#]
29.Qh6+ Kf5 30.Qf6#
(2) A similar initial sequence counters the alternative capture 26…Qxf4: 26…Qxf4 27.Rxh7+ Kxh7 28.Qh3+ Kg6 [Kg7 29.Qh6#]
29.<g4> (threatening 30.Qh6# and defending against 29…Qf5+) <[Here, I suffered failure of board vision on the obvious alternative 29…Qxg5, which wins easily for Black.]> (3) The only other feasible defense is to move Rg8, to avoid the dovetail mate: 26…Re8 [Rd8 or Rf8 are similar but worse]
27.<Qh4> (threatening 28.Rxh7+ Kg8 29.Rh8+ Kg7 30.Qh7#) <[While 27.Qh4 wins handily, Toga gives 27.g6 as better.]> 

Aug0209
  johnlspouge: < <An Englishman> wrote: [snip] I didn't and chose 26.g6!?, which almost works, but probably doesn't for reasons I can't see. > Hi, <An Englishman>. Toga gives 26...hxg6 as a draw, with best play as follows. [ply 14/58 time 00:45 value 0.00]
26. ... fxg6 27.Rxh6 Qxf4 28.Qxf4 Nxf4 29.Rc7 Rg7 30.Rxc8+ Rg8 31.Rc7 Rg7 32.Rc8+, etc. 

Aug0209
  An Englishman: Good Evening: Nice work, <johnispouge>. At least I would not have lost. 

Aug0209
  johnlspouge: < <An Englishman> wrote: Good Evening: Nice work, <johnispouge>. At least I would not have lost. > I'll take that as dry British humour and slink quietly away :) By the way, your greeting (and comment) reminded me of an apropos response I recently heard to the excessively American "Have a nice day." "Thank you, but I have other plans." 

Aug0209   Utopian2020: After finding the crushing moves 26  30, white makes a small mistake with 31. g4+ (mate in 13) instead of Nd4+ (mate in 5). 

Aug0209
  OBIT: <An Englishman>Oh, come now. Having come this far, do you really take the draw with an obvious and boring perpetual? Don't you play Nxg6+ and lose the rest of the half point? 

Aug0209
  An Englishman: Good Evening: With my tactical "skill," <OBIT>, I probably would. 

Aug0209   bengalcat47: Studying the game from move 23, if Black tries to defend with 23...Rh8
White has in reply 24 Rxc8! capturing Black's QB! Black has to defend with
24...Bh4, after which White comes out a piece ahead and his attack remains unabated. 

Aug0209
  Jimfromprovidence: I was baffled by why 25 Nf4, below, and not 25 Nf6?
click for larger view
I think it has to do with the black knight’s ability to successfully capture the f knight. So, if 25 Nf4 Nxf4?!, white has 26 Qxf4, unleashing a nasty threat.
click for larger view
White now threatens 27 Qf6+ Rg7, then Rxc8#. Black probably has to play 26…f5 to minimize material loss. If white plays 25 Nf6?, it looks like black can play 25 Nxf6 successfully.
click for larger viewHere, white cannot recapture with either pawn without giving up material. 

Aug0209   Athamas: 26.?
The glaring move is Rxh6; it looks setup, but pushing the pawn looks halfway interesting as well with black's rook on the back rank overworked. 26. Rxh6
I am blind, but I see nothing better for black than to recapture the knight and lose in a forced mate after Rxh7. The back rank rook can't move to d8 because of 27. Rxh7 Kg8 28. Rh8+ Kxh8 29. Qh4+ Kg8 30. Qh6 f6 31. g6 Rb7 32. Rh1 followed by a couple of useless checks by black and an unstoppable mate on h8. Or at least this is what I saw; engine analysis would be appreciated. Rg6 does not work with bishop hanging on the back rank, which is the same with Rg7. Looks like a completely forced mate after Rxh7 with the queen pushing the king back down to f5 and the rook guarding pawn... back 2 pawns come into play for mate and possibly the knight 

Aug0209
  patzer2: For today's Sunday puzzle solution, White springs a difficult but deep and decisive pursuit combination with the pseudo sacrifice 26. Rxh6!! 

Aug0209
  whiteshark: Very elegant combination. I especially like <30.Re1> as a non check move, when 3 moves before and after it are checks. 

Aug0209   wals: THE FOLLOWING MAY BE OF HELP TO THOSE WANTING HELP(OR ADDED TO THE COLLECTIVE UNCONSCIOUSNESS).
[Event "Metz op 21st"]
[Site "Metz"]
[Date "2003.04.19"]
[Round "9"]
[White "Bozidar Ivanovic"]
[Black "Gennadi Ginsburg"]
[Result "10"]
[ECO "B95"]
[Annotator "Rybka 3 1cpu (30m)"]
[PlyCount "65"]
♗95: Sicilian ♘ajdorf: 6 ♗g5 e6, unusual White 7th moves 1. e4.. c5 ( 2. Nf3 d6 Consolidates e5 3. d4... cxd4 4. Nxd4 {Black is behind in developement. Black's piece can't move:
f8 4... Nf6 ♗lack threatens to win material:
♘f6xe4 5. Nc3... a6 Covers b5 (5... g6 6.
Be2 ) 6. Bg5 (6. Bc4 g6 6... e6 ♗lack is behind in developement (6...
Nbd7 7. h4 ) 7. Qf3 (7. Be2 h6 8. Be3 b5 ) 7... Nbd7 (7... Bd7 8. OOO
Nc6 9. Kb1 ) 8. OOO (8. Be2 Qc7 ) 8... Qc7 9. Kb1 (9. a3 Be7 ) 9...
Be7 ♗lack has a cramped position. ♗lack's piece can't move: c8 (9... b5 10.
a3 ) 10. Qg3 (10. Be2 b5 11. a3 Rb8 ) 10... b5 11. Bd3 White has a very
active position (11. f3 b4 12. Nce2 Nc5 ) 11... b4 ♗lack threatens to win
material: b4xc3 12. Nce2 White has an active position Nc5 13. f3 Controls
g4 (13. Be3 OO (13... Ncxe4 14. Bxe4 Nxe4 15. Qxg7 ) 14. Bh6 Ne8 )
13... Rb8 (13... e5 14. Nf5 Bxf5 15. exf5 ) 14. Nb3 (14. Be3 OO ) 14...
Nxd3 (14... e5 15. Nxc5 Qxc5 16. Rdf1 ) 15. cxd3 ♗lack has the pair of
bishops. (15. Rxd3 e5 ) 15... OO (15... a5 16. Be3 ) 16. Nf4 (16. Rc1
Qd8 ) 16... Kh8 (16... Ne8 ) 17. d4 (17. Bxf6 gxf6 18. Rd2 Bd7 ) 17...
Rb5 (17... a5 18. e5 a4 19. Rc1 ) 18. h4 (18. d5 e5 19. Rc1 Qd8 20. Bxf6
Bxf6 ) 18... a5 19. Rc1 White threatens to win material: ♖c1xc7 (19. d5 e5
20. Bxf6 gxf6 ) 19... Qa7 (19... Qb8 $5 ) 20. e5 dxe5 21. dxe5 Nd5
♗lack lets it slip away (21... Ng8 and ♗lack can hope to survive)
22. Bh6 gxh6 23. Nh5 Bg5 24. hxg5 Rg8 (24... Qe3 25. Nf6 Nxf6 26. gxf6 (26.
exf6 Rxg5 27. Qb8 Rgg8 ) 26... Qd3+ 27. Ka1 ) 25. Nf4 Qe3
terrible, but what else could ♗lack do to save the game? (25... Ne7 26. Qh2
Nf5 ) 26. Rxh6 Qxf4 (26... Nxf4 27. Rxh7+ Decoy: f5 Kxh7 28. Qh4+ Kg6
29. Qh6+ Kf5 30. Qf6#) 27. Rxh7+ Kxh7 28. Qh3+ Kg6 29. Qh6+ Kf5 30. Re1 Qxg5 (
30... Qxe5 is the last straw 31. Qh7+ Rg6 32. g4+ Kxg5 33. Rxe5+ Kf6 34. Qh8+
Ke7 ) 31. g4+ (31. Nd4+ Kf4 32. Qh2+ Qg3 33. Re4+ Kg5 34. Qxg3+ Kh6 35. Rh4#
) 31... Kf4 (31... Qxg4 doesn't get the cat off the tree 32. fxg4+ Rxg4 33.
Qh5+ Kf4 34. Rf1+ Ke4 35. Qxg4+ Kxe5 36. Qg3+ Ke4 37. Re1+ Ne3 38. Rxe3+ Kf5
39. Nd4+ Kf6 40. Nxb5 Ke7 41. Qg5+ Kd7 42. Rd3+ Kc6 43. Rd6+ Kb7 44. Qg2+ Kb8
45. Rb6+ Bb7 46. Qxb7#) 32. Qh2+ Kxf3 33. Nd4+ (33. Nd4+ Kxg4 34. Rg1#) 10 

Aug0209   gofer: Nice puzzle...
26 Rxh6 ... ...threatening Rxh7+ Kxh7 Qh4+/Qh3+ mating 26 ... Qxe5 27 Ng6+ winning
26 ... Qxf4 27 Rxh7 Kxh7 28 Qh3+ mating
26 ... Nxf4 27 Rxh7 Kxh7 28 Qh4+ mating
26 ... Rg7 27 Rxc8+ Rg8 28 Rxh7 Kxh7 29 Qh4+ mating
26 ... Rg6 27 Rxc8+ Kg7 28 Rxh7 Kxh7 29 Qh4+ mating
26 ... Rxg5 27 Rxc8+ Kg7 28 Qxg5#
26 ... Rxg5 27 Rxc8+ Rg8 28 Qxg8#
26 ... Rf8 27 Rxh7 Kg8 28 Qh4 f6 29 Qh6 mating
26 ... Rf8 27 Rxh7 Kxh7 28 Qh4+ Kg8 (Kg7 29 Qh6+ Kg8) 29 Rh1 mating So the rook can't stay on the g file otherwise its curtains, and Rf8 isn't any better, so the rook must move to e8 or d8. There doesn't seem to be much difference. Once the rook has moved to e8 or d8 white can play g6 and "cry havoc and unleash the dogs of war"... The main line...
26 ... Re8/Rd8
27 g6! ...
27 ... Qxe5/Qxf4/Nxf4 28 Rxh7 Kg8 29 gxf7+ Kf8 (Kxh7 30 Rh1 mating/winning) 30 Qg8+ Ke7 31 f8=Q# (or if 26... Re8 then 31 ... Kd8 32 Qxe8#) 27 ... fxg6 28 Nxg6+ Kg7/Kg8 29 Nf4+ Kf8 (Kxh6 30 Rh1# or Kf7 30 Rxh7+ Kf8 31 Qg7#) 30 Rxh7 mating I may have missed the best defense, but I hope not... Time to check... 

Aug0209   gofer: Oh dear...
I missed that black could play 29 Kf5 in both these variations... "Null Points" this Sunday... :(
26 ... Qxf4 27 Rxh7 Kxh7 28 Qh3+ mating
26 ... Nxf4 27 Rxh7 Kxh7 28 Qh4+ mating 

Aug0209   CHESSTTCAMPS: Material is even, with black threatening Qxf4. If white plays Nxd5?, black plays exd5 freeing the Bc8 and winning white's g5pawn. White can't give up the advantage so easily; there is an initiative on the hfile with a chance to sacrifice material for a king hunt: 26.Rxh6!
I found this pretty fast, but had trouble nailing down the right continuation. Declines of the sacrifice (with attempts to escape from the corner) can be dismissed quickly: A) 26... Kg7? 27.Rxh7+ Kxh7 (Kf8 28.Rxc8+) 28.Qh3+ Kg7 29.Qh6# B) 26... Rd8 (or e8) 27.Qh4 Kg8 28.Rxh7 Kf8 29.g6 f6 (otherwise 30.Rh8+ followed by 31.Qh7#) 30.Nxd5+ Rxd5 31.Qxf6+ and mate follows quickly. B.1) 27... Q/Nxf4 28.Rxh7+ Kg8 29.Rh8+ Kg7 30.Qh6#
The acceptance Nxf4 also allows a quick finish:
C) 26... Nxf4 27.Rxh7+ Kxh7 28.Qh4+ Kg6 (Kg7 28.Qh6#) 29.Qh6+ Kf5 30.Qf6# However, the acceptance Qxf4 gives white chances to go wrong: D) 26... Qxf4 27.Rxh7+ Kxh7 28.Qh3+ Kg6 (Kg7 29.Qh6#) and now: D.1) 29.Rh1?? Qf5+ ends white's fun
D.2) 29.g4? Qxg5 30.Rh1 Kg7 and white has nothing
D.3) 29.Nd4? Qxg5! (not Qxd4? 30.Rh1) 30.Nxb5 Qf5+ and white is forced to trade queens a piece down Correct is:
D.4) 29.Qh6+ Kf5 30.Re1! (not Qh7+ Kxe5 31.Qxg8 Bd7) threatening g4+ winning the queen. This leaves no good options: D.4.1) 30... Qxg5 31.Nd4+ Kf4 32.Qh2+ Qg3 33.Re4+ Kg5 34.Qxg3+ Kh5/h6 35.Rh4# D.4.2) 30... Rxg5 31.Qxh7+ Rg6 32.Qh5+ Qg5 (Rg5 33.Qxf7+ Nf6 34.Qxf6#) 33.Nd4+ Kf4 34.Qh2+ Qg3 35.Re4+ etc. D.4.3) 30... Qxe5 31.Qxh7+ Rg6 (Kf4 32.Qh2+) 32.g4+ Kf4 33.Qh2+ wins D.4.4) 30... Qg3 31.Nd4+ Kf4 32.Re4#
D.4.5) 30... Qe3 31.Rxe3 Nxe3 32.Qf6#
D.4.6) 30... Qc4 31.g4+ Kf4 32.g6+ Kg3 (Kxf3 33.Nd2+ Kf2 34.Qh4+) 33.Rg1+ Kf2 (Kxf3 34.Nd2+ Kf2 35.Qh2+) 34.Qh2+ Ke3 35.Re1+ Kxf3 36.Nd2+ wins D.4.7) 30...Rg6 31.g4+ wins the Q
D.4.8) 30... Nc3+! 31.Ka1! and black can't defend the threats of 32.Qf6# and g4+ This seems to cover the major lines. Time to verify.... 

Aug0209
  DarthStapler: I got the first 4 moves but I chose g4+ instead of Re1 since I missed that black could play Kxe5. Still best I've ever done on a Sunday I think 

Aug0209   WhiteRook48: 26 Rxh6 is easy to find 

Aug0209
  patzer2: Although I didn't solve today's Sunday puzzle, for me the key to understanding the combination is finding the strong followup 30. Re1! and the mateinthree with 32.Qh2+! Kxf3 33. Nd4+ Kxg4 34. Rg1#. 

Aug0209   TheBish: B Ivanovic vs G Ginsburg, 2003 White to play (26.?) "Insane"
Material is even.
Candidate moves: Rxh6, g6
I discounted 26. Nxd5 (trading a key attacker) and 26. Rxc8 (both 26...Rxc8 and 26...Qxf4 break up any chance of attack). 26. g6?!? is interesting, with the idea 26...Qxf4? (or 26...Nxf4?) 27. g7+ Rxg7 28. Rxc8+ Rg8 29. Rxg8#, but simply 26...fxg6 and White has given away a key pawn which could be used as an attacker at the right moment, and Black defends after 27. Rxh6 Qxf4 28. Qxf4 Nxf4 29. Rc7 Rg7 30. Rxc8+ Rg8 31. Rc7 Rg7, with a likely draw. 26. Rxh6! is the right way. White threatens 27. Rxh7+ Kxh7 28. Qh4+ Kg7 29. Qh6#. Now: A) 26...Qxf4 27. Rxh7+! Kxh7 28. Qh3+ Kg6 (28...Kg7 29. Qh6#) 29. Qh6+ Kf5 30. Re1! Rxg5 (or 30...Qg3 31. Nd4+ Kf4 32. Re4#, or 30...Qxg5 31. Nd4+ Kf4 32. Qh2+ Qg3 33. Re4+ Kg5 34. Qxg3+ Kh6 35. Rh4#, or 30...Qc4 31. g4+ Kf4 32. Qh2+ Kxf3 33. Nd2+ Kxg4 34. Rg1+ Kf5 35. Qh7+ Kf4 36. Qxf7+ Ke3 37. Nxc4+) 31. Qh7+ Rg6 32. g4+ Qxg4 (or 32...Kg5 33. Qh5#) 33. Qxf7+ Kg5 34. fxg4 and wins. B) 26...Nxf4 27. Rxh7+ Kxh7 28. Qh4+ Kg6 (or 28...Nh5 29. Qxh5+ Kg7 30. Qh6#) 29. Qh6+ Kf5 30. Qf6 mate. C) 26...Kg7? 27. Rxh7+ is easy, since 27...Kxh7 28. Qh4+ is mating (original threat) and 27...Kf8 28. Rxc8+ Ke7 29. Rxg8 (not 29. Nxd5+? Rxd5 30. Rxg8?? Rd1+ 31. Kc2 Qd3#) Nxf4 30. Rh1! and White has a huge lead, a rook and two pawns. D) 26...Rd8 (or Re8) [best defense] 27. g6! and now:
D1) 27...Nxf4 28. Rxh7+ Kg8 29. gxf7+ Kf8 (or 29...Kxh7 30. Rh1+ Nh5 31. Rxh5+ Qh6 32. Rxh6+ Kxh6 33. Qh4+) 30. Qg8+ Ke7 31. f8=Q mate. D2) 27...Qxf4 28. Rxh7+ Kg8 29. gxf7+ Kxh7 (or 29...Kf8 30. Qg8+ Ke7 31. f8=Q#, as in D1) 30. Rh1+ Qh6 31. Rxh6+ Kxh6 32. Qh4+, as in D1. D3) 27...fxg6 28. Nxg6+ Kg7 (or Kg8) 29. Ne7+ Kf7 (or 29...Kh8 30. Rxh7+ Kxh7 31. Rh1+ Qh6 32. Rxh6+ Kxh6 33. Qg6#, or 29...Kf8 30. Qg8+ Kxe7 31. Rxh7#) 30. Qg6+ Kd7 31. Nxd5 mate. E) 26...f5 27. Rxh7+ Kxh7 28. Qh4+ Kg7 29. Qh6+ Kf7 30. Qh7+ Rg7 31. g6+ Ke8 32. Rxc8+ Kd7 33. Qh8 and mate is near. Probably some other variations I missed, but time to see the game! 

Aug0409   LIFE Master AJ: BTW, I have analyzed this game on my computer, it makes for some interesting variations. However, with all the good work already posted, I don't see much point in adding mine. 

Aug0409   LIFE Master AJ: POSTSCRIPT:
I just did a test with Fritz 11. (I had similar results with Rybka.) If you clear the hashtables and then engage the engine (750 MB of RAM) on the position after 25...Qe3; the first choice of the engine is 26.NxN/d5!, and this remains true for at least the full first minute. When I checked back ... several minutes later ... the metal monster had indeed found the winning line. (26.RxP/h6!!) So now I don't feel so bad about my move choice. If the almighty box does not find RxP/h6 right away, what chance does a mere human have? 

< Earlier Kibitzing · PAGE 2 OF 2 ·
Later Kibitzing> 