[Event "?"]
[Site "?"]
[Date "2021.11.26"]
[Round "3"]
[White "Nepomniachtchi, Ian"]
[Black "Carlsen, Magnus"]
[Result "1/2-1/2"]
[ECO "C88"]
[WhiteElo "2782"]
[BlackElo "2855"]
[Annotator "samsh"]
[PlyCount "81"]
[EventDate "2021.??.??"]
{[%evp 0,81,19,38,25,7,30,29,18,-1,-1,-10,15,-24,6,2,6,1,-8,11,20,-25,-8,24,29,
27,65,47,65,45,42,48,38,38,35,25,80,70,70,81,81,81,81,75,57,61,37,41,48,13,33,
23,25,10,19,9,9,1,13,7,8,-10,-10,-13,-11,-3,3,0,0,-4,0,-17,0,0,0,0,0,0,0,0,0,0,
0,0] Game 3 saw another draw. I was curious who would be the first to deviate
from game one, where Magnus voluntarily went for a pawn-down endgame with some
compensation. It turned out to be Nepo—I suspect his team checked it closely
and decided Black should be holding on and decided to go a different direction.
} 1. e4 e5 2. Nf3 Nc6 3. Bb5 a6 4. Ba4 Nf6 5. O-O Be7 6. Re1 b5 7. Bb3 O-O 8.
a4 (8. h3 {This was seen in game one, when Magnus chose the very unusual
response} Na5 {which eventually reached a pawn-down ending. I felt White
should be better, but Magnus seemed very at-home and never had any trouble.})
8... Bb7 {Magnus has chosen this move several times before, so it can't have
come as a huge surprise.} 9. d3 d6 10. Nbd2 Re8 11. Nf1 h6 12. Bd2 {So far,
both players are following well-established theory. This position has seen
play at the top level as far back as the early 2000s.} Bf8 13. Ne3 Ne7 14. c4
bxc4 (14... c6 {This has generally been preferred in ICCF, though it looks a
bit passive to me. I like Magnus' choice better.}) 15. Nxc4 Nc6 $5 $146 {
[%CAl Gc6e5,Rd6d5] Finally, at long last, we are in fresh territory. In
general, White's position looks a little more pleasant to me, though I don't
love that the pawn on a4 looks a little overextended. Magnus was still playing
pretty fast and was very likely still in his preparation, so he must have been
more optimistic about his chances than the computers would indicate. When I
let them run for awhile, they definitely gravitated towards equality.} 16. Rc1
$5 {[%csl Gc7][%CAl Rc1c7] The machine gives this as best, but in light of how
the game played out, Nepo could have considered another option.} (16. Na5 $5 {
Hindsight is 20/20, but I think not letting Black advance a7-a5 had to be
considered. It means White will have fewer weaknesses on the b-file in the
long run, and the a6-pawn is fixed. After something like} Nxa5 17. Bxa5 Qd7 18.
h3 $14 {I'm not sure I love Black's position, but it is still very solid.})
16... a5 $1 {[%csl Gd3][%CAl Ga5b4,Rb4d3,Gc6b4] This was the first move Magnus
spent a fair amount of time on, which is a little surprising to me. His last
moves were played fast enough that I would think he was still in preparation,
but Rc1 is the first choice of the machine! It would be really weird if Magnus
had prepared to up to 15...Nc6 and then stopped. Perhaps he was just trying to
recall all the lines. In any case, Black takes the a5 square under his control
and is ready to re-route his bishop to e6 next. It has to be said that
Stockfish is giving White a nagging edge, but it never looked like much during
the game. Thus far, Magnus has clearly shown a willingness to enter lines the
computer doesn't entirely approve of. I like the fighting spirit!} 17. Bc3 $1 {
[%CAl Gf3d4,Gc3d4,Rd3d4,Gd1d3]} Bc8 $1 {[%CAl Gc8e6,Gc8g4,Rd6d5,Rg4d1] The
bishop had no good future on b7 and instead heads for e6 to head off the
monster on b3.} 18. d4 exd4 19. Nxd4 Nxd4 20. Qxd4 Be6 {Optically, the
position looks very nice for White, but there's nothing concrete for him to do.
Black is ready to push d7-d5 next, and then the structure will not be as bad
for him as it looks. The overextended a4 pawn weakens the b-file, particularly
the b2-pawn, and White will have just as many weaknesses as Black.} 21. h3 $6 {
[%CAl Gh3g4]} (21. Nxa5 $2 {White cannot get away with going pawn grabbing.
After} c5 $1 22. Qd2 Bxb3 $15 {[%CAl Ra8a4] White is losing his pawn back at
the very least, and he could end up in serious trouble as Black's pieces
spring to life.}) (21. Qd3 $5 {[%CAl Rc3g7] At high depths, this is my
machine's recommendation. Still, it does not look impressive. After} d5 22.
exd5 Nxd5 $1 {Black can aim for a pawn-down endgame with a lot of
activity—much like he did in game one.} 23. Bxa5 Nb4 {[%csl Gd3]} 24. Qxd8
Raxd8 {The threat of Nd3 more or less forces White to part with the bishop
pair.} 25. Bxb4 Bxb4 26. Re4 {[%CAl Re4b4]} Rb8 $5 $11 {The weaknesses on the
b-file make themselves felt, and Black is ready to regoup with Be7-f6 and Rb4.
He should be fine.}) (21. Bc2 {This was also possible immediately, but it
would have met a similar fate as the game continuation.} d5 $1 22. e5 $3 dxc4
23. Qxd8 Rexd8 24. exf6 $5 $11 {[%CAl Rc3h8,Ge1e3,Ge3g3] We would reach
approximately the same ending we saw in the game.}) ({Stockfish 14:} 21. Qd3 d5
22. Bxf6 $1 Qxf6 23. exd5 Bd7 24. Qc3 Rxe1+ 25. Rxe1 Rb8 26. Ne5 Bf5 27. Re3
Bb4 28. Qc6 Bd6 29. Nc4 g6 30. Nxa5 Qxb2 31. Qc3 Qxc3 32. Rxc3 Re8 33. Re3 Ra8
34. Nc6 Kg7 35. g3 g5 36. Kg2 Bd7 37. h4 $14) 21... c6 22. Bc2 {Now, if White
had another move, he could play Qd2-d3, threatening to take on f6 and play e5.
This would be a nightmare to deal with, but Magnus is not one to miss threats.}
d5 $1 (22... Qc7 {A slow move like this one would allow White to reveal his
plan. After} 23. Qd3 $1 $16 {Bxf6 is a very hard threat to deal with.}) 23. e5
(23. exd5 Nxd5 $11 {Black is ready to bring the knight to b4 next, taking
advantage of the weakened square. He certainly looks fine to me. He could also
consider just taking on c3 and c4 to reach an opposite-bishops position.})
23... dxc4 24. Qxd8 Rexd8 25. exf6 {Black has to be a little careful here due
to his pawn structure, but Magnus held without trouble.} Bb4 26. fxg7 Bxc3 27.
bxc3 Kxg7 28. Kf1 {The best White can hope for is to trade all the rooks, but
as we see in the game, this does not gaurantee victory.} Rab8 29. Rb1 Kf6 30.
Rxb8 Rxb8 31. Rb1 Rxb1+ $1 {Full credit to Magnus for calling the bluff. This
strikes me as a somewhat difficult move to make since it gets one step closer
to a lost pawn-ending, but White is unable to get the bishops off the board or
make any progress.} (31... Rd8 {I was expecting this in the game, which
probably also holds, but the text is surely cleaner.}) 32. Bxb1 Ke5 33. Ke2 f5
$1 {The last important move. Black needs to get his pawn to f4 to keep his
king secure in the center and box White's king out.} (33... Bd7 {Marking time
for a move is not to be recommended. After} 34. Ke3 {My computer claims this
is still a draw, but I would definitely worry about b4, g4, Kf3, h4, etc. when
White can make a passed pawn.}) 34. Bc2 f4 $1 {White can make no progress. g3
is never possible due to the hanging h3 pawn, the king cannot touch f3 without
allowing a check on d5, and there is nothing else to try.} 35. Bb1 c5 36. Bc2
Bd7 37. f3 Kf6 38. h4 Ke5 39. Kf2 Kf6 40. Ke2 Ke5 41. Kf2 {This was a pretty
dull game, more consistent with the kind of chess I would have expected in
games one and two, when Black normally draws without any trouble before White
knows what's coming. It's a bit surprising to me that Black did not manage to
equalize easily in games one and two, but then did in game three. In general,
I think Magnus can be very happy with how the match is going so far. Game one
looked a little shaky in the very early phases, but since then, his
preparation has been excellent, and he has shown a willingness to take some
strategic risks. Still, we are just getting started; only 3 games in and no
decisive ones just yet. The real fight is yet to come!} 1/2-1/2