Giri's Short King

Carlsen's reluctance to stay on board one in Doha brought there Kramnik, who on his turn freed it to the young and shiny Anish Giri. The Dutchman played a brilliant game to outplay another young diamond- Nils Grandelius from Sweden to secure the top position at least for the next round. In a certain moment Giri was brave enough to walk with his king in the middle of the board a la Short:
A game that I liked (ChessBase 13)

[Event "Qatar Masters Open"]
[Site "?"]
[Date "2015.12.21"]
[Round "2.2"]
[White "Grandelius, Nils"]
[Black "Giri, Anish"]
[Result "0-1"]
[WhiteElo "2632"]
[BlackElo "2784"]
[Annotator "Bojkov, Dejan"]
[PlyCount "80"]
[EventDate "2015.??.??"]
[WhiteClock "0:32:06"]
[BlackClock "0:04:36"]

1. e4 c5 2. Nf3 d6 3. d4 cxd4 4. Nxd4 Nf6 5. Nc3 a6 {Diagram [#] Giri
generally prefers the Taimanov Sicilian, but every once in a while he opts for
the Najdorf too.} 6. h3 e5 7. Nb3 Be6 8. f4 ({Another plan is} 8. Be3 Be7 9.
Qf3 O-O 10. g4 a5 11. O-O-O a4 12. Nc5 a3 {as in the recent masterpiece
Khismatullin,D (2654)-Sjugirov,S (2646) Khanty-Mansiysk 2015}) 8... Be7 {
Diagram [#] Giri revealed at the press-conference that he was not sure about
his memories of the line and mentioned two recent games, important for the
line:} (8... b5 9. f5 Bc8 10. Bg5 Bb7 11. Qf3 Nbd7 12. O-O-O Be7 13. Kb1 Rc8 {
Grischuk,A (2785)-Wojtaszek,R (2701) Riga 2013}) ({And} 8... Nbd7 9. f5 Bxb3
10. axb3 d5 11. exd5 Bb4 12. Bd2 O-O 13. Qe2 Bxc3 14. bxc3 e4 {Sutovsky,E
(2635)-Shirov,A (2712) Poikovsky 2015 Although White won both these games
Black did reasonably well out of the opening.}) 9. f5 Bd7 $146 {A novelty. The
bishop is heading for the c6 square.} ({A Najdorf specialist chose instead}
9... Bc8 {trying to fianchettoe the bishop asap and keeping the queenside
knight's options open-} 10. Qf3 b5 11. a3 Bb7 12. g4 d5 $1 {and Black was
already better in Sutovsky,E (2632)-Wojtaszek,R (2736) Bilbao 2014}) 10. Be3 b5
{The fight for the d5 square is fierce. Black dismissed the immediate} (10...
Bc6 11. Qf3 d5 {due to} 12. Nxd5 $1 ({Giri also calculated the cunning line}
12. exd5 Bxd5 $1 13. Qg3 $6 Ne4 $1 14. Qxe5 Nc6 15. Qxd5 Bh4+ {[%csl Rd5]
Diagram [#] and all of a sudden Black wins the queen} 16. Ke2 Nxc3+) 12... Bxd5
13. O-O-O Nxe4 14. Bc4 $1 {where Black's position falls apart.}) 11. Qf3 ({
Perhaps Grandelius should have gone for} 11. a3 Bc6 12. Qf3 {which seems more
stable.}) 11... b4 12. Nd5 Nxd5 13. exd5 {Now that the d5 outpost disappeared
Black should be fine.} Bg5 {[%csl Ye3,Yg5] Diagram [#] In order to trade the
"poor bishop". This however leaves the queenside somewhat poorly placed.} (
13... a5 14. Nd2 {would have been somewhat better for White.}) ({But} 13...
O-O {looks quite good for him. I suspect that Giri did not want to show where
is his king going yet. He often leaves it in the middle in the Taimanov and if
you remember his game again Saric in Wijk this year you might find a lot of
similarities to the one you are following right now.}) 14. Bxg5 ({White has
nothing after} 14. Bf2 O-O 15. Bd3 Bb5) 14... Qxg5 15. Bd3 Bb5 16. h4 {
Grandelius is a very strong attacking player and remains true to his style.
Objectively though a better approach would have been the one suggested by the
top commentator P.Svidler:} (16. Nd2 Nd7 17. Ne4 Qe7 18. O-O {followed by the
opening of the queenside with a2-a3} Nf6 19. a3 {[%csl Ya6,Yb4,Yb5][%cal Ra1a8]
Diagram [#] Giri believed this should not be too bad for Black but the
impression is that he is worse.}) 16... Qd8 17. g4 Nd7 18. g5 $6 {"A mistake
that gives the "h" file to Black." (Giri).} (18. O-O-O $11) 18... h6 $1 {
[%cal Rh8h1,Rh1h8] Diagram [#] Or else White's initiative on the kingside
might become annoying-} (18... Qb6 19. O-O-O a5 20. Nd2 Nc5 21. Bxb5+ Qxb5 22.
h5 {and White is ready to open files (Giri).}) 19. Rg1 {One more timid move.
The Dutch GM believed that White should do better if sacrifices a pawn:} (19.
f6 gxf6 20. g6 fxg6 21. Bxg6+ Ke7 22. O-O-O {intending Rh1-e1 and Nb3-d4-f5
with compensation for a pawn.}) 19... hxg5 20. hxg5 Bxd3 21. Qxd3 Rh5 $1 {
[%cal Gh5g5,Yd3e3] Diagram [#] "A strong move that forces the white queen to
be misplaced" (Giri). Things are not that simple though...} 22. Qg3 ({Black is
indeed doing great in the endgame after} 22. Qe3 Qb6 $1 $17 {as the white
pawns are overstretched.}) ({Also in the line} 22. Qd2 a5 $1 {when the knight
is deprived of the d2 square.}) ({However, there was a third possibility that
Giri did not mention:} 22. f6 $1 {Diagram [#] with the idea} gxf6 ({Safer is}
22... g6 23. O-O-O {but then the white queen would not have to leave.}) 23. g6
$1 {This looks like an improved version of the pawn sacrifice from above.}) ({
And even} 22. O-O-O $5 {was interesting with the idea to get the file back
after} Rxg5 23. Rh1 $132) 22... Qc7 23. O-O-O Rc8 {[%csl Rc2,Gc7,Gc8,Yd7,Gh5]
[%cal Gh8h1,Gc8c1] Diagram [#] Now Black has almost all his pieces wonderfully
placed. Only the knight can do better and once that the "a" pawn starts moving
White's position will become critical.} 24. Rd2 a5 25. Kb1 ({Giri was a bit
afraid of the line} 25. f6 g6 26. Qe1 a4 27. Nd4 {but it fail tactically due to
} Nb6 28. Nc6 Nxd5 {and Black nets a pawn.}) 25... a4 26. Nc1 Qc4 $1 {Better
than} (26... Nb6 27. Qg4 $1 {that forces the black rook back} Nc4 28. Re2 Rh8 {
(Giri)}) 27. Qf3 ({Here} 27. Qg4 Rh4 28. Qxc4 Rcxc4 {leads to perfect endgame
for Black.}) 27... Rh4 28. f6 g6 {Black controls all the files and does not
need to open more.} 29. Rh1 Nc5 {This brings the last piece into the action.} (
{Svidler mentioned} 29... e4 {but it is not better due to} 30. Re1 Ne5 31. Qd1
{Diagram [#]}) (29... Nb6 $5 {deserved attention too.}) ({But the preliminary}
29... a3 $1 30. bxa3 Nc5 {with huge attack seems best.}) 30. Rdh2 Ne4 {Diagram
[#] The critical moment of the game.} ({The computer believes} 30... Rxh2 31.
Rxh2 Kd8 {was best for Black, but I am not sure I understand a thing here.})
31. Qd1 ({Both players missed that in the line} 31. Qe3 a3 {White does not
have to go for} 32. Qa7 {which is indeed winning for Black after Giri's
recommendation} ({But can free himself of the mating threat with} 32. c3 $1 {
[%csl Gb1,Rc2,Re8][%cal Rh1h8] Diagram [#] which takes over the initiative for
him. For example} Rxh2 33. Rxh2 {is already losing for Black-} bxc3 34. Rh8+
Kd7 35. Qh3+) 32... Kf8 $1 $19 (32... Rc7 33. Qa4+ Kd8 34. Rxh4 (34. Qa8+ Rc8
35. Qa5+ Ke8 36. Qa7 {returns to the main line where Black can still find the
move} Kf8 $1) 34... Nd2+ 35. Ka1 axb2+ 36. Kxb2 Qc3# {Diagram [#]})) 31... Rxh2
32. Rxh2 Kd7 $1 ({Avoiding the cunning trap:} 32... Nxg5 33. Qg1 Nf3 34. Rh8+
Kd7 35. Qa7+ Rc7 36. Qxa4+ {and White wins.}) 33. Qg1 Qxd5 34. Qa7+ Ke6 $1 {
[%csl Ge6][%cal Ge6f5,Gf5g5] Diagram [#] Too many Short followers...} 35. Qg1 {
With the innocent threat of a check from g4.} ({The barce black king will eat
all the white kingside pawns after} 35. Qe7+ Kf5 36. Qd7+ Qe6 37. Qxa4 Kxg5 ({
Or} 37... Qc4 {first.})) 35... Kd7 36. Qa7+ Ke6 ({There was an alternative}
36... Rc7 37. Qxa4+ Ke6 38. Qe8+ Kf5 {with serious advantage for Black.}) 37.
Qg1 Qc4 $1 {Frees the d5 square for the king!} 38. Re2 $2 {This loses. For
good or bad Grandelius should have gone for the complications} (38. Qg4+ Kd5
39. Qd7 Rc7 ({Carlsen showed a sense of humor with his suggestion} 39... Kd4 {
[%csl Rd4] Diagram [#] but it is doubtful he would have used it in his own game
} 40. Ne2+ Ke3 41. Qh3+ $18) 40. Qxa4 {and here instead of the suggested by
Giri} Nxg5 {that leads to a mess after} ({Black has a safer try} 40... Ke6 $1 {
with good winning chances.}) 41. Rd2+ Ke6 42. Qe8+ Kf5 (42... Kxf6 43. Qd8+ Re7
44. Qxd6+ Qe6 (44... Re6 $2 45. Qd8+ Re7 46. Rf2+) 45. Qxb4 $13) 43. Rf2+ Kg4 {
Still messy, but objectively much better for Black.}) 38... b3 $1 {[%csl Rb1]
[%cal Rc8c1] Diagram [#]} 39. axb3 {Allows a nice finish of the game.} ({
We already know where is the king heading to after} 39. Qg4+ Kd5 $1) ({
Objectively best was} 39. cxb3 axb3 40. a3 Nc5 $19) 39... axb3 40. cxb3 Qxe2 $1
{A pretty final strike. White resigned due to the mate:} (40... Qxe2 41. Nxe2
Nd2+ 42. Ka2 Ra8+ {Diagram [#]} 43. Qa7 Rxa7#) 0-1


No comments: