The Bishop Move That Opened the Door: Svane Beats Vokhidov at the Samarkand Olympiad

Look at the bishop on d3. It looks like it is just sitting there, blocking its own queen's file. Vokhidov moved it away on move 16, and that single move lost the game.

8

7

6

5

4

3

2

1a

b

c

d

e

f

g

h

Position after 15...Nf4, with White to move. The knight on f4 attacks the bishop on d3. White has to do something about it.

Shamsiddin Vokhidov 0-1 Rasmus Svane 46th FIDE Chess Olympiad, Samarkand 2026 · Open Section, Round 8 · 24 September 2026 White: Shamsiddin Vokhidov, Uzbekistan (2654) · Black: Rasmus Svane, Germany (2610)

Germany won the match 2.5-1.5. Frederik Svane (Rasmus's brother) also won for Germany, beating Nodirbek Yakubboev. Vincent Keymer drew with Nodirbek Abdusattorov, and Javokhir Sindarov won the other game for Uzbekistan.

How the position happened

The opening was a Scotch Game, one of the oldest ways to meet 1.e4 e5. Both sides traded a pair of knights early, and then traded pawns in the center, which is normal in this opening and leaves an open, simple position.

1.e4 e5 2.Nf3 Nc6 3.d4 exd4 4.Nxd4 Nf6 5.Nxc6 bxc6 6.Bd3 d5 7.exd5 cxd5 8.O-O Be7 9.c4 O-O 10.Nc3 Be6 11.cxd5 Nxd5 12.Bd2 Bf6 13.Rc1 a5 14.b3 Bxc3 15.Bxc3 Nf4

Nothing here is sharp or forcing. Both kings were already castled and safe, both sides had traded down to a simple structure, and the position looked like the kind of game that ends in a quiet draw. That calm is exactly what makes the next move dangerous: it is easy to play a normal-looking move on move 16 without checking what changed underneath it.

A knight attacks a bishop, and something else too

Black's last move, 15...Nf4, does the obvious job of attacking the white bishop on d3. A knight on f4 also covers the squares e2, g2, and h3, right next to the white king, though that is not the point of this particular move. White needed to answer the immediate threat: move the bishop, defend it, or trade it off.

16.Be4??

This looks like a normal retreat. It saves the bishop, it stays on the long diagonal, and it even points at the black pawn on h7. But it is a mistake, and a serious one. Before this move, the bishop on d3 was doing a second job nobody was thinking about: it stood on the file between White's queen on d1 and Black's queen on d8, blocking it completely. Move the bishop off that file for any reason, and the two queens are suddenly looking straight at each other, with nothing in between.

8

7

6

5

4

3

2

1a

b

c

d

e

f

g

h

Position after 16.Be4, with Black to move. The d-file is now open all the way from White's queen on d1 to Black's queen on d8, with nothing standing between them.

Before you move a piece, ask what it is blocking, not just what it is attacking. The bishop on d3 was not doing anything dramatic. It was standing guard on a file, and nobody noticed until it left.

The safe move was 16.Qf3, defending the knight's attack on d3 indirectly by threatening the knight on f4 in return, rather than just retreating. Black can grab the bishop with 16...Qxd3, but after 17.Qxf4 the knight falls too, and the position stays level. The difference between 16.Be4 and 16.Qf3 is not really about which piece moves first. It is about which side is in control of the exchanges that follow.

The queen trade that was not a trade

16...Qxd1 17.Rfxd1

So far this just looks like queens coming off the board, an even trade, the kind of moment where a spectator might stop paying close attention. But Svane had already worked out what came next, three moves in advance.

17...Ne2+

A knight fork. From e2, the knight attacks the white king on g1 and the white rook on c1 at the same time. White has only two legal replies to the check, Kh1 or Kf1, and neither one has anything to do with the rook on c1. Wherever the king goes, the knight is still attacking c1 next move, and nothing else defends that square.

8

7

6

5

4

3

2

1a

b

c

d

e

f

g

h

Position after 17...Ne2+, with White to move and in check. The knight on e2 attacks both the king on g1 and the rook on c1. Kh1 or Kf1 answer the check, but neither one saves the rook.

18.Kf1 Nxc1 19.Rxc1

8

7

6

5

4

3

2

1a

b

c

d

e

f

g

h

Position after 19.Rxc1, with Black to move. Count the pieces: one rook and one bishop each side, but White is missing the rook that once stood on c1, replaced by nothing. Black is up the exchange.

Count the material. The queens are gone, even. The knight that forked twice is gone too, traded for White's rook. What is left is that Black turned a knight into a rook: a straight trade of knight for rook, called winning the exchange. A rook is normally worth about two pawns more than a knight, so this is a real, lasting material edge, not a knockout blow. It does not end the game by itself, but it gives the winning side an advantage that a good player can push for the rest of the game. That is exactly what Svane did.

Our own Stockfish run, depth 24, backs this up with numbers. The position right before 16.Be4 was close to even, only 0.13 of a pawn in White's favor. One move later, after 16.Be4, the same engine puts Black ahead by 1.66 pawns, roughly the value of the exchange Black was about to win. After the safer 16.Qf3, the engine's evaluation barely moves, 0.12 of a pawn for White. That gap, 0.13 to 1.66, all from one move, is the size of the mistake in numbers rather than words.

A long game, and a wild finish

Winning the exchange is not winning the game. Svane still had to convert a small, safe edge into a full point, and that took another 32 moves of careful play, not one more trick. Vokhidov defended for a long time by trading down and activating his own bishop and rook, and the position slowly turned into an endgame where both kings marched toward opposite corners, each chasing a set of passed pawns: White's on the queenside, Black's on the kingside.

40.b4?

By this point the position had simplified into a pure pawn race, rook and king against rook and king, with the outcome down to whose pawns ran faster and whose king arrived to help first. Our engine marks 40.b4 as a further mistake, one that let Black's kingside pawns run even faster than they needed to; the computer's own preference, 40.a6, would have kept the race closer. Positions like this reward counting moves, not calculating deep combinations: whoever's pawn queens with a safe king wins, and every tempo matters.

8

7

6

5

4

3

2

1a

b

c

d

e

f

g

h

Position after 40.b4, with Black to move. White's a- and b-pawns are already rolling, but Black's h-pawn is one square further along and Black's king is already close enough to help it.

By move 45, both sides had a pawn one square from the very back rank.

8

7

6

5

4

3

2

1a

b

c

d

e

f

g

h

Position after 45...h1=Q, with White to move. Black has already promoted a new queen on h1. White still has two pawns, on a7 and b7, one move from promoting too.

Black queened first, but arriving first is not the same as arriving safely. White's king was still out in the open in the center of the board, with no pawns of its own left nearby for shelter, while Black's rook and brand-new queen were already close enough to start giving checks. White pushed a second queen onto the board with 46.b8=Q, but a second queen does not help if the king cannot find safety. From here Black's queen and rook chased the white king with a series of checks it could never escape for long.

50.Kd2??

This is the final mistake, and our own engine confirms it turns a large advantage into a forced mate. After 50...Qd4+ 51.Ke1 Qg1+, White resigned rather than play on. The position at resignation shows exactly why: our engine finds a clean mate in three, 52.Kd2 Qf2+ 53.Kc1 Qf1+ 54.Kb2 Qa1#. Look closely at that finish. When the white king finally reaches b2, its own rook on c2 blocks one escape square, and Black's rook, sitting quietly on a3 since move 47, covers the rest of the board around it. The king runs out of squares because its own piece is in the way, the same idea behind a back-rank mate, just played out in the middle of the board instead of the last rank.

What this game teaches

1. Ask what a piece is blocking, not just what it is attacking. The bishop on d3 was not doing anything urgent on its own. It was standing between two queens, and moving it opened a door nobody meant to open.

2. Who chooses when the queens come off the board matters. 16.Qf3 and 16...Qxd1 both end with the queens traded, but only one side got to choose the moment. Offering a trade on your own terms is safer than allowing one that comes with a tactic attached.

3. A fork wins the piece the king cannot also save. The knight on e2 attacked the king and the rook together. The king has to move somewhere, and wherever it goes, the rook stays behind.

4. Winning material is the start of the job, not the end. Svane was up the exchange after move 19 and still needed 32 more moves, and a further mistake by White, to finish. A good position still has to be played out, move by move.

5. Whoever promotes first still has to survive. Black's new queen on h1 arrived before White's on a7 or b7, but the race was decided by which king was safer once the new queens were on the board, not by who promoted first.

6. Your own pieces can trap your own king. In the final mating net, White's own rook on c2 blocked an escape square just as much as Black's pieces did. A king needs room to run, and sometimes its own army is standing in the way.