Mar0214
  bright1: I think after 24. Qh5 if Black plays Bxe5 then 25. Rd8+ will win 

Mar0214   charmarks: Does 24. Qg4 not work?
24. Qg4+ Kf8 (Kh8 25. Nxf7++, Bg5 25. Qxg5+ and see continuation)
25. Nd7+ Ke8 (Ke7 26. Bxf6+ Ke8 27. Qg8++)
26. Qg8+ Ke7
27. Bxf6++
If someone can get black out of 24. Qg4 please let me know. Thanks. CM 

Mar0214   chinokoli: 24. Qg4+ is met by 24... Bg7. After 25. Rd7 Qb1+ 26. Kh2 Be4... white is slightly better but black can definitely hold. 

Mar0214   PJs Studio: Very nice win against a super solid GM. I liked Qg4+ also but the game continuation seems flawless by white. 

Mar0214   charmarks: Ah, the simple move. Thanks, chinokoli. I am much better at visualing on a physical board than one on the monitor. After Bg7, I saw: 25. Nxf7 Qxb2
26. Qxe6 Qb1+
27. Kh2 Qxd3
28. Nh6+ Kh8
29. Qg8+ Rxg8
30. Nf7++
. . . but I just now saw 26. . . . Qf6. Does Qf6 hold? 

Mar0214   Rookiepawn: 25. Qg4+ Kh8
26. Nxf7+ Rxf7
27. Rd8+
I forgot 25. ... Bg7. Damn, I thought I got it... 

Mar0214   diagonalley: yep... QN4+ just looks so inviting... but it seems it doesn't quite work  i wonder what a machine would pick(?) 

Mar0214
  hoodrobin: Insane but interesting too. If you find the first move, just play the game. 

Mar0214   morfishine: I was unable to tell which move is strongest: 24.Qh5 or 24.Nxf7 

Mar0214   Cheapo by the Dozen: At the moment I'm quite stumped. 24 Qg4+ forces 24 ... Bg7, but: 25 Rd7 is answered by 25 ... Qxb2
25 Rg3 fxg3
26 Nd3 Qxb2
27 Nxb2 gf+
gives up too much for the queen (including, I think, the knight) 25 Rd2 is too slow
Meanwhile, 24 Rd7 runs afoul of interference via 24 ... f3, and 24 Qh5 Rf8 doesn't seem to help White at all. 

Mar0214
  OhioChessFan: White's attack is thematic in a French. Use the Black squares and make haste while Black's Queen is away from home. 24. Nxf7 also looks good but the best I can find is a perpetual. 24. Nxf7 Qxb2 25. Qxe6 Rf8 26. Rd7 Qb1+ 27. Kh2 Qe4 28. Qg4+ Qg6 29. Ng6+ Kh8 30. Qxg6 hxg6 31. Nf4+ Kg8 etc click for larger view 

Mar0214
  Penguincw: As soon as I saw it, I was thinking of some fork on d7. Turns out that would lose. 

Mar0214
  agb2002: White has a bishop and a knight for the bishop pair and a pawn. At the moment, Black does not threaten 24... Bxe5 25.Qxe5 Qb1+ 26.Kh2 Qxd3 because of 27.Qg7#. The open gfile and the pinned bishop suggest 24.Qg4+: A) 24... Kf8 25.Nd7+
A.1) 25... Ke7 26.Bxf6+ Ke8 27.Qg8#.
A.2) 25... Ke8 26.Nxf6+ Ke7 (26... Kf8 27.Qg8+ Ke7 28.Rd7#) 27.Rd7+ Kf8 28.Qg8#. B) 24... Kh8 25.Nxf7#.
C) 24... Bg7 25.Nxf7
C.1) 25... Kxf7 26.Qxg7+ Ke8 27.Qd7+ Bf8 28.Bg7+ Kg8 29.Qxe6+ Kxg7 30.Rd7+ and mate in two. C.2) 25... Qxb2 26.Nh6+ (26.Qxe6 Qf6) 26... Kh8 (26.... Kf8 27.Qxf4+ unclear) 27.Nf7+ with repetition. 
Another option is 24.Qh5, aiming at the weak square f7: A) 24... Bxe5 25.Qxe5 Qxb2 (25... Kf8 26.Qh8+ Ke7 27.Bf6#; 25... f6 26.Qxe6+ and mate soon) 26.Qxb2, etc. B) 24... Qxb2 25.Qxf7+ Kh8 26.Qxf6+ Kg8 27.Rd7 and mate soon. C) 24... Rf8 25.Rd8
C.1) 25... Bxd8 26.Qg4+ Kh8 (26... Bg5 delays the disaster just one move) 27.Nxf7#. C.2) 25... Rxd8 26.Qxf7+ Kh8 27.Qxf6+ Kg8 28.Qxd8+ Kg8 29.Qg5+ Kf8 (29... Kh8 30.Nf7#) 30.Qf6+ and mate in two. C.3) 25... Bxe5 26.Qg5+ and mate next.
C.4) 25... Qb1+ 26.Kh2 Qg6 27.Rxf8+ Kxf8 28.Nxg6+ is a massacre. C.5) 25... Qxb2 26.Qxf7+ Kh8 27.Q(R)xf8#.
D) 24... Bg7 25.Qxf7+ Kh8 26.Ng6+ hxg6 27.Qxg7#.

24.Qh5 looks far stronger than 24.Qg4+. 

Mar0214   GlennOliver: 24. Qg4+ seems a probable draw at best for White, e.g. 24. Qg4+ Bg7 25. Nxf7 Qxb2 26. Rd7 Bc6 +2.38/11 22 27. Qxe6 Qb1+ 28. Kh2 Qg6 29. Nh6+ Kh8 30. Nf7+ Kg8 31. Nh6+ Kh8 32. Nf7+ Kg8 1/21/2 

Mar0214   Cheapo by the Dozen: I totally missed the rook sacrifice. 

Mar0214   LAK: Good puzzle today!
I spent most of my time trying to make 24. Qg4+ work, because I misevaluated 25. Qh5. I even found some very nice lines under 24. Qg4+. Unfortunately, I did't even consider a candidate move (... Qb1+) by Black in the variation 25... Qxb2 26. Rg3 Qb1+ 27. Kh2 Qf5. What is inexcusable is that in all the rest of my analysis that particular move comes up again and again as a main defensive move for Black. If Black doesn't go for that move, there are some nice mates in this line. In fact, I started my analysis by working on 24. Qh5 but missed two key White moves that lead me to reject this line. In the variation 24. Qh5 Qb1+ 25. Kh2 Bxe5 26. Qg5+ Kf8 27. Rd8+ Rxd8 28. Qxd8+ Kg7 29. Bxe5+ Kg6 30. Qf6+ Kh5 31. Bxf4 Qg6, I missed that 32. g4+ simply mates: Qxg4 33. Qh6# I also missed that in the variation 24. Qh5 Qb1+ a possible, and the strongest, candidate move for White is 25. Rd1. Here's Stockfish's evaluation of this subvariation: 24. Qh5 Qb1+ 25. Rd1 Qxb2 26. Qxf7+ Kh8 27. Qxf6+ Kg8 28. Rd8+ Rxd8 29. Qxd8+ Kg7 30. Qg5+ Kf8 31. Qf6+ Ke8 32. Qf7+ Kd8 33. Qd7# Anyway, I had a lot of fun with this position today.
Here's what Stockfish thinks is the main line:
24. Qh5 Rf8 25. Rd8 Qb1+ 26. Kh2 Bxd8 27. Qg4+ Qg6 28. Nxg6 hxg6 29. Qxf4 f6 30. Qd6 Bc8 31. Qxc5 Bb7 32. Qxb4 Rf7 33. Qd2 Bc7+ 34. Kg1 Kg7 35. b4 Bc8 36. b5 Rd7 37. Qc2 Bb8 38. c5 Rd5 39. Kf1 e5 40. b6 Bf5 41. Qc3 Rd3 42. Qb4 Be6 43. Bc3 Rd8 44. c6 Kf7 45. Qb5 Bc7 46. bxc7 Rc8 47. Qb7 Bc4+ 48. Kg1 Be6 49. Qb8
+ (7.30++) Depth: 40/62 0:02:00 834 MN 

Mar0214   brazos: I decided on 24. Qg4+ Bg7 25. Qg5 but missed that after 25.. Qxb2 the plan of 26. Rd8+ Rxd8 27. Qxd8+ Bf8 28. Nd7 is thwarted by 28.. Qg7 Ah well, dismissed Qh5 after initially not feeling confident of Qg4, but thought 25. Qg5 was an improvement on other Nxf7 lines. 

Mar0214   Patriot: Black is up a pawn and the bishop pair.
There are several interesting candidates but my main focus is 24.Qh5. 24...Qxb2 25.Qxf7+ Kh8 26.Qxf6+ Kg8 27.Qxe6+ looks winning. 24...Bxe5 25.Qxe5 since there is no time for 25...Qb1+ 26.Kh2 Qxd3. 24...Rf8 25.Rd7 Bxe5 26.Qxe5 f6 27.Qxe6+
24...Rf8 25.Rd7 Qxb2 27.Rxf7 Qb1+ 27.Kh2 Rxf7 28.Qxf7+ Kh8 29.Qf8# 24...Rf8 25.Rd7 Qxb2 27.Rxf7 Qb1+ 27.Kh2 Bxe5 28.Rxb7 Qg6 29.Qxe5 There are a lot more variations after 24...Rf8, but I'll go with 24.Qh5 since white seems to have a lot of options. 

Mar0214   gofer: ~~~
Hmmm, I got nowhere today.
Option 1

24 Qh5+ ...
then after about 30 minutes I decide that its not that easy <as I don't find Rd8!> Option 2

24 Qg4+ ...
24 ... Kh8
25 Nxf7#
24 ... Kf8
25 Nd7+ Ke8 (Ke7 26 Bxf6+ Ke8 27 Qg8#)
26 Qg8+ Ke7
27 Bxf6#
24 ... Bg7
then after about 30 minutes I decide that its not that easy either! So after thinking LONG and HARD. I like a quiet move... Option 3

<24 Rd1!! ...>
This does nothing... ...well almost nothing.
a) It stops Qb1+
b) It threatens Ra1 trading rooks...
c) But mostly it allows the original plan of Qg4+, but with the winning
kicker that Rd3 has vacated d3 for Ne5!
<25 Qg4+ ...>
26 ... Kh8
27 Nxf7#
25 ... Kf8
26 Nd7+ Ke8 (Ke7 26 Bxf6+ Ke8 27 Qg8#)
27 Qg8+ Ke7
28 Bxf6#
<25 ... Bg7>
<26 Nd3 > winning the queen. But I still didn't know if this would work! Because black is really
unlikely to play Qxb3 losing or f3 losing and is far more likely to
play <24 ... Bg7>. So I still looked for something better... Option 4

24 Nxf7!
This too I looked at for ages and got nowhere. Lots of choices, but nothing concrete. A couple hours wasted. <But thoroughly enjoyed!> 

Mar0214   clement41: with 19...Bb7, black seems to lure white into playing Rd7, forking the two black bishops.
But after Rd7 black would reply ...Rd8, when Rxb7?? Qxb2 and black is crushing and if Rxe7the same ...Qxb2 is also strong (but less than on Rxb7??). Here Re8+ Kg7 wouldn't help white. however, the amazing Rd7 Rd8 Ng5! fg Rxe7 Qxb2 Re8+!! this time favours white. On Kg7 there is Qxb2+ now so ...Rxe8 Qxb2 

Mar0214   WhiteRook48: Like most Sunday puzzles, this one was over my head. I'd hoped that the more straightforward 24. Qg4+ would win. I did <suspect> that Rd1d8 would need to be played at some point in order to distract Black's darksquared bishop from being able to defend properly, I had all the wrong variations. 

Mar0214
  LIFE Master AJ: I got 24.Qh5 after just a minute or two, but missed Rd8!! 

Mar0314   patzer2: The followup 25. Rd8!! is the stunner I missed in rejecting 24. Qh5! (going instead for 24. Qg4+). The surprise followup 25. Rd8!! became obvious to me only after I looked at it for a minute or two. If 25...QxB, then 26. Qxf7+ Kh8 27. Qxf8#.
If 25...Bxd8, then 26. Qg4+ Bg5 27. Qxg5+ Kh8 28. Nxf7# (i.e. double check and mate). If 25...Rxd8, then 26. Qxf7+ Kh8 27. Ng6+ hxg6 28. Bxf6#. P.S.: However, after 25. Rd8!! I missed Black's inbetween move 25...Qb8+ and the game continuation which still wins for White. The only consolation for me is this is one of the few Sunday puzzles where I didn't need computer analysis to understand the followup moves. 

Mar0314   patzer2: Plugged 24. Qh5!Rf8 26. Rd8!! into the computer anyway. My variations (at least the ones I calculated) were correct, but I missed 25. Rd8!! Kg7 26. Rxf8! with mateinseven. P.S.: Fritz 12 gives 24. Qg4+ as leading to a draw by perpetual after 24. Qg4+ Bg7 25. Nxf7! Qxb2 26. Nh6+
(26. Qxe6 Qf6 27. Nh6+ Kh8 28. Nf7+ Kg8 29. Nh6+ Kh8 30. Nf7+ Kg8=) 26... Kh8 27. Nf7+ Kg8 28. Nh6+ Kh8 29. Nf7+ Kg8=. 

Mar0414   Moszkowski012273: Very instructive... 
