Analysis
The game in detail
GM Javokhir Sindarov (UZB, 2778) vs GM Peter Leko (HUN, 2676) · 1-0
Board 1, Uzbekistan vs Hungary · English Opening, Symmetrical Variation, Four Knights Variation (A35)
By Professor Zugzwang AI Coach . This analysis was written by a Bogner Chess AI coach.
Board one of the round, Uzbekistan against Hungary, and two generations at one table: Javokhir Sindarov, 2778, against Peter Leko, who played for the world title a good twenty years ago. Sindarov had White, opened with , and after had a symmetrical English position in front of him, the kind where both sides can do everything right and neither gets anything.
That is how it went. Sindarov gave up a pawn on move ten, won it back three moves later with a small trick, held an advantage that was rarely bigger than half a pawn for forty moves, and decided the game in the ending with a single rook move. Fifty-eight moves of patience against an opponent who made patience his profession.
The pawn that is still defended: 11.Bg5
The obvious move loses a piece. After nothing recaptures: the queen on b3 does not see d4, neither does the bishop on c1, and the bishop on g2 sits behind its own knight on e4. The d4 pawn is defended by the knight on c6 and the bishop on g7, and attacked only by the knight on f3.
So Sindarov did not take, he kept developing for two more moves: and . Both are the only moves that hold the balance. costs a good half pawn, and a good pawn and a half each, and is precisely the move roughly one player in three reaches for at this level, because it looks so sensible.
The point is the knight on c6. After it stands on b4 and no longer covers d4, and only now does work. The counter attacks a bishop that nothing defends, and White comes out a good pawn better. The rule behind it holds in every opening: you do not take a defended pawn, you lure the defender away first.
GM Peter Leko (HUN, 2676)
GM Javokhir Sindarov (UZB, 2778)
Lure the defender away: 17.Nf6+
The same motif four moves later, with the colours reversed. The queen on a3 attacks the knight on b4, and exactly one piece defends it: the knight on d5. Count material before you count defenders here and you play and you have given the queen for a knight.
is the answer. The knight goes with check to a square only the knight on d5 and the bishop on g7 can take on, and either capture drags a defender away from b4. After the knights are traded and material is level again; after the same thing happens.
The order matters more than the move. looks identical and is a good third of a pawn worse, because Black keeps the bishop on f6 and the long diagonal belongs to him. Check the order last in every liquidation you calculate: usually only one of them also leaves standing the pieces you need afterwards.
GM Peter Leko (HUN, 2676)
GM Javokhir Sindarov (UZB, 2778)
The bishop that has to go: 20.Be5
The bishop on g7 is Leko’s best piece. It sits on the long diagonal, it shields the king, and while it stands there White has nothing on the dark squares. So Sindarov played : Leko propped the bishop up with , and White traded anyway.
What I like about these two moves is how little they gain. is about a fifth of a pawn better than anything else, rather less, and neither is a move players at this level pick. Everybody looks for , which takes the a2 pawn out of the d5 bishop’s diagonal and is not bad, just a move without a plan.
A fifth of a pawn is objectively nothing and practically half the game: g7 now holds a knight instead of a bishop, and the dark squares around Leko’s king belong to nobody. With a tiny advantage, do not ask which move increases it, ask which enemy piece is most in your way.
GM Peter Leko (HUN, 2676)
GM Javokhir Sindarov (UZB, 2778)
The rook takes the defended pawn: 51.Rxf7
Five pawns against five, rook and knight against rook and knight, and White stands not quite a pawn better without owning an extra one. The natural move is , and almost everyone plays it: it takes the undefended pawn, hands over the e4 pawn in return, and the game goes on with the same small edge.
takes the defended pawn instead, because the rook cannot actually be collected: wins the rook back with a fork and leaves White a pawn up. So the king has to stay where it is, and then the f7 pawn is simply gone.
Leko declined and played , and not quite a pawn turned into a good pawn and a half. is the only move that holds anything at all: after or Black has the pawn straight back and the position is level again. The lesson: before you take the free pawn, look at the defended one. A pawn a piece defends is often available for exactly that reason, because the piece leaves its post to capture.
GM Peter Leko (HUN, 2676)
GM Javokhir Sindarov (UZB, 2778)
Two pawns against one: 54.Nh4
The queenside majority decides the game, and the impatient also wins. It is simply worth a good pawn less than what Sindarov played, and the whole difference sits in one move inserted first.
Sindarov put a move in between: . From h4 the knight attacks the g6 pawn, Black has to defend it with , and a knight on e7 is one move further from the queenside than it was. Only then did the pawn set off, and now it runs.
The rest is arithmetic. gives the knight for Black’s last kingside pawn and turns the g5 pawn into a passed pawn, and the b7 pawn costs a rook: leaves White with rook against knight and a passed pawn on g5. Leko resigned. When you have a pawn majority, the usual question is not whether but when: the piece to its best square first, then the pawn.
GM Peter Leko (HUN, 2676)
GM Javokhir Sindarov (UZB, 2778)
Verdict
A game in which nothing happens for fifty moves and which is won anyway says more about a 2778 player than any sacrifice would. Sindarov’s advantage was rarely bigger than half a pawn up to move fifty, and he allowed himself no more than one imprecise queen move on move thirty-two. Then came , the one move nobody plays, Leko sidestepped it, and from move fifty-four the queenside majority was running. A point on board one for Uzbekistan against Hungary, and a good illustration that patience is not the absence of ambition but its most durable form.
Moves
1-0
Was this analysis useful?