Now 37. Qxf8 allows Rg6+ and (I think) a draw by perpetual.

And after 37. Rxb6 Bc8+ 38. Kh4 Rh6+ and again black draws by rook checks.

But a winning position by 36. ... Bxf2 is clearly better than a perpetual, so no points for me.

All other moves also lose except one:

38.Kh3! Bf1 draws. I had thought that 38.Bf3?? won, but it loses to 38...g4! (I only saw how powerful this was when I found out 36.Bf2!! had been played.)

The f8 rook is not as out of the action as it initially appears...and when I saw I could activate it, this line became very attractive. I didn't get it as far the line below, but I saw that I had something.

Well on the bright side, tomorrow's problem is a chance to repair my ego.

One might consider 36...Bxf2!! to be a desperado, defensive move since both a Black Bishop and a Rook are under simultaneous attack as a result of White's last move. However, I believe it's no accident and that GM Kovalev enticed White into a false sense of security by setting him up for this position (36...?) -- where on the surface it looks like White is winning a piece with a decent game, but is actually losing after Black's astonishing 36th move.

I spent an embarassing amount of time on this thinking about the candidates 36 ...Kg6 and 36 ... Rh8. The idea of Rh8 is fairly straightforward - follow up with the move ...Kg6. The idea of playing ...Kg6 first is that after the queen captures the f8 rook, then playing ...Be2+ moves the white king over to the h-file and hopefully mate can be administered in some way.

Of course after spending time on all of this (and I won't bore you with my calculations), I realized that with either black move that the white response 37. Qe5 is tough for black to deal with.

So going back to the drawing board, what about:

36 ... Rxf2

Well for one thing it offers the possibility of following up with ...Be2, even with the white queen still on the e-file.

White can respond in a number of ways: 37. Qh4+ or Qxf8 or Rxb6 or Qe5

37. Qh4+ Kg8
Although this brings the K to g8 looks counter-productive (even if it is virtually forced), this might not be so bad. ...Bd8 might be possible. It might also be possible to go after the queen by bringing the rook back. Still I don't know.

There is also a problem with taking the f2 pawn because white (after taking the f8 rook) can move the queen over to d6, and this make it difficult for the black king to advance (or the queen may land on d6 with a check if the king has already advanced).

Alternatively, I don't think taking the f2 pawn with the b6 bishop does anything.

So now I have started to think about 36 ... Bd8. This looks promising.

37. Qxf8 Be2+ then

38. Kh4 Rh6# or
38. Kg5 Rxf2# or finally
38. Kh3 Bf3
And white has no defense. Taking the d8 bishop does not prevent Rh6#. White can only delay things by throwing away the queen with 39. Qg8 or h8

But what if white declines to take the f8 rook on move 37?

Well taking the d8 bishop with the queen is obviously not the answer. Moving the queen off the e-file spells the same sort of mating trouble as before by allowing ...Be2+.

So what about a move like 37. Qe5 (or somewhere else on the e-file)?

Well honestly I don't know. I have been looking briefly at ...Rd6, ...Rg6 and ...Bg2 and after spending so much time earlier on dead-ends I am calling it quits now.

But I am confident that 36 ... Bd8 is the key first move. After a goose-egg yesterday, I really wanted to find the first move, and I knew that if I gave up then I would be kicking myself with what was bound to be an obvious first move (obvious once looking at the solution page and playing through the game). At the very least, white cannot touch the f8 rook without losing and Bd8 gets the bishop out of harms way (from the white rook on b7). Moving ...Bd8 is bound to improve black's position, and there is probably a win or a good follow-up plan that could be found after a response like 37. Qe5. For a Friday and for my current skill level, I am satisfied.

I have to be honest and say that I was ready to give up before thinking about 36 ...Bd8. To me, this was not easy to see at all. I had thought about getting the light squared bishop to e2 with several of my incorrect initial ideas. I only really saw this one when I mentally connected the real threat of using the advantage of the two bishops to rake the white king over the diagonal coals.

Well, we're up material but not for long as both the f8 Rook and b6 Bishop are feeling some heat. Additionally, our king could enjoy better protection but white�s monarch is naked unto the world. Perhaps Metodi sees it as an attacking piece. Hey, it worked for Nigel Short! But I digress�

Candidate moves:

33. Rg6+. And we can beg for 34. Kh4 so we can play Bd8 with a wonderful pin but there is the off chance white actually sees this as a possibility; he can also move Kf4 inviting a perpetual check. Either way, I don't see Rg6+ leading anywhere but ending up with black still with two pieces facing execution. Next.

33. Ra8. Eyeing a possible Ra4+ but that falls to 34. Rxb6 and then 35. Rb4 with the block. White is still up material after the exchange. I'm beginning to look at the old golf clubs. Perhaps the utter futility of chasing a little white ball would be a relief right about now. Let's press on.

33. Bxf2. Screw it, I'm grabbing material. Pick up a pawn, make him bleed a little; let him know I mean business. It also puts the dark square Bishop in the same zip code as the naked white King. 34. Qxf8 Be2+. Hmmm. Let's see where this goes.

35. Kh3 Bf3 and mate in a move. Cool.
35. Kg5 Be3+ and mate in a move. Cool.

What if White doesn't take the rook? The white queen stays where she is, on the open e-file and puts an end to the sublime Be2+? Am I frustrated because I can't apply my pretty mate? Then black has taken a pawn, still has a mother of an attack and is up material to boot. I'll live with the frustration.

Er, could you tell me your zip code, so my ego can avoid moving there?

Er, could you tell me your zip code, so my ego can avoid moving there?

Good luck :)"

Thanks, but I didn't have to play him. I spent a few hours taking him through

Larsen vs Fischer, 1971

using both this web site and also Chessbase. Yesterday I played the English against him (which he hates) and I thought I would show him an example of how a first rate player who likes the attack confronts the English.

It turns out that he knows a good deal of chess, but very little of the history; he really didn't know any of the players that came up in the analysis other than Fischer himself. So for example he did not know Karpov or Kasparov, let alone Kramnick, Topalov, Anand, etc. So I took some time to tell him about Tal and some other guys he hadn't heard of. I also showed his dad (who was over this time) this site, because he apparently really liked "Guess the Move". In fact I had the kid play that with

Nezhmetdinov vs Tal, 1961

and he did pretty well, but not as well as he did the day before with

C Poulsen vs Najdorf, 1939

where he scored really well. (The Najdorf Sicilian is his favorite).

One of the terrifying things about a kid like this is that his memory is very strong. For example after seeing the Larsen-Fischer game played through at about two seconds per move, he was able to get almost all the moves for both sides about fifteen minutes later when I had him go through it on Chessbase in training mode (where only the previous move is displayed).

By the way we are not that far from White Plains, NY, where I am told there are a lot of strong kids.

<1. e4 c5 2. Nf3 Nc6 3. d4 cxd4 4. Nxd4 Qb6!?>

This is not the most frequently played move, but it does show up ever so often in strong Master play, with the Chinese GM Zhang Pengxiang being perhaps the strongest of the players currently employing it. Super GM Judit Polgar has also tried the move with fair success. Other Super GMs who have employed it on occasion are Tal, Kramnik and Morozevich.

If the Opening explorer results are any indicator, Black's results in this unbalanced position are about as good as White's, with Black winning about 36%, White winning 36% and 28% draws occuring in the OE database games. The moves more often played here are 4...Nf6, 4...e6, 4...g6 or 4...e5.

<5. Nb3 Nf6 6. Nc3 e6 7. g3?!>

This is an infrequently played alternative, which according to the CG.com Opening Explorer (OE) has scored poorly for White in practical play (18% White wins versus 47% Black wins). Most often seen here is 7. Bd3 (45% White wins to 24% Black wins) as in Lin Chen vs Z Pengxiang, 2008, Sedlak vs Z Pengxiang, 2006, Ponomariov vs Akopian, 2002, Anand vs Judit Polgar, 1992 and Leko vs Morozevich, 2000. Also more frequently played is 7. Be3 (25%% White wins versus 48% Black wins) as in F Caruana vs Z Pengxiang, 2008, R Gonzalez vs Tal, 1992 and R Gonzalez vs Tal, 1992. Another popular choice is 7. Qe2 (44% White wins versus 24% Black wins) as in D Jakovenko vs Z Pengxiang, 2007, Movsesian vs Z Kozul, 2002, Motylev vs Judit Polgar, 2002, Anand vs Kramnik, 2001 and Grischuk vs Smirin, 2000.

<7...Bb4 8. Bg2 d5 9. exd5?!>

Now Black has a small but clear advantage and the initiative.

<10. Bd2 Nxc3 11. bxc3 Be7 12. O-O O-O 13. Rb1 Qc7 14. Nd4 Rd8 15. Nxc6 bxc6 16. Qe2 Qa5 17. Rfd1 Ba6 18. Qe1 Qa4 19. Be4 Bb5 20. Qe3 Qxa2!?>

This move I believe anticipates and prepares Black's coming Queen sacrifice. It's the kind of pawn grab novices wouldn't think twice about , but which strong players and especially Masters have to consider deeply.

<21. Ra1 Qc4 22. Bd3 Qd5 23. c4 Bxc4 24. Ba5 Re8!>

Black I believe saw ahead to this possibility, and knew earlier he must now lose the Queen. However, in giving up the Majestic Lady, Black also knows that he gets more than enough compensation (superior pawn structure, Queen-side pawn majority, and the initiative)to give him the advantage after this "even trade" [i.e. Rook, Bishop and Pawn (9 points) for Queen (9 Points)].

<25. Bxh7+ Kxh7 26. Rxd5 exd5 27. Rb1?!>

This move drops a piece, and leaves White with a lost game after Black's next move. Instead, maybe white could try 27. Bc3 c5 28. Qe5 d4 29. Ra4 Be6 30. Bd2 g6 .

<27... Bb4!>

The discovered attack wins back another piece with decisive advantage.

<28. Qf4 Bxa5 29. Rb7 Re1+ 30. Kg2 Rf8 31. Rxa7 Bf1+ 32. Kf3 Re6 33. Qh4+ Rh6 34. Qe7 Rf6+ 35. Kg4 Bb6 36. Rb7>

On the surface, it looks like White has made a
reasonably good move and has his game under control, but in reality it leaves him with a lost game after Black's coming suprise.

White can and should put up more resistance with 36. Ra1, but with 36... Be2+! Black is back in control after 37. Qxe2 Rxf2 38. Qe7 f5+ 39. Kh3 Rf6 40. Ra4 d4 41. g4 Rg6 42. c4 dxc3 43. g5 Rf3+ 44. Kg2 Re3 45. Rh4+ Kg8 46. Qb4 Rxg5+ 47. Kf1 c2 48. Qc4+ Kf8 49. Qxc6 Rg1+ 50. Kxg1 Rc3+ 51. Kg2 Rxc6 .

<36... Bxf2!!>

With this brilliant continuation, Black sacrifices his Rook for a decisive attack on the helpless, exposed White King.

<37. Qxf8>

Black also wins after 37. Rb4 d4 (37... Kg8 38. Rf4 Re6 39. Qd7 Be2+ 40. Kh3 Be3 41. Kg2 Bxf4 42. gxf4 Bg4 43. Kg3 Bf5 $19) 38. Qxf8 Be2+ 39. Kh3 Bf3 40. Qe7 Rh6+ 41. Qh4 g5 42. Qxh6+ Kxh6 43. g4 f5 44. gxf5 g4#)

<37... Be2+! 38. Kh3>

Also winning for Black is 38. Kh4 Rh6+ 39. Kg5 Be3+ 40. Kf5 Rh5# and 38. Kg5 Be3+ 39. Kh4 Rh6#.

<38... Bf3 39. Rb4>

Winning even quicker for Black is 39. g4 Rh6# and 39. Qd8 Rh6+ 40. Qh4 Rxh4+ .

<39... d4! 40. Rxd4>

Black obviously wins after 40. Qe7 Rh6+ 41. Qh4 g5 (41... Rxh4+ 42. gxh4 f5 43. h5 Be3 44. Rb2 Be4 45. Kh4 f4 46. Ra2 f3 47. Ra1 f2 48. Kg3 Bxc2 49. Kg2 d3 50. Ra2 d2 51. Rxc2 d1=Q ) 42. Qxh6+ Kxh6 43. g4

[43. Rc4 f5 44. Rc5 g4+ 45. Kh4 d3 46. Rc3 Bd4 47. Rxd3 Bf6#; 43. Rb6 d3 44. Rb4 d2 45. Rb1 g4+ 46. Kh4 Bd4 47. Rb5 Bf6+ 48. Rg5 Bxg5#; 43. Rb8 d3 44. Rh8+ (44. cxd3 g4+ 45. Kh4 Bd4 46. Rg8 Bf6+ 47. Rg5 Bxg5#; 44... Kg7 45. Rd8 dxc2 46. Rd3 g4+ 47. Kh4 c1=Q ]

43... f5 44. gxf5 g4#.

<40... Bxd4 41. g4 Bf2 42. g5 Rf5 43. Qa3 Rf4 44. g6+ fxg6> 0-1

White resigns in a clearly lost poition. Otherwise, one of several possible winning continuations is 45. Qe7 Rf5 46. Qh4+ Rh5 47. Qxh5+ gxh5 48. c4 g5 49. c5 g4#.