Short's Kg3 King March vs Timman — Tilburg 1991

Tilburg 1991: The Setting

The Interpolis tournament at Tilburg in the Netherlands was one of the elite events of its era, a closed gathering of the world's best held each autumn. The 1991 edition brought together Garry Kasparov, Anatoly Karpov, Viswanathan Anand and the cream of the professional circuit. Nigel Short arrived as Britain's strongest player since the nineteenth century and a man on the rise: within two years he would defeat Karpov and Timman in succession to earn a title match against Kasparov, the first Englishman ever to play for the world championship. Those World Chess Championships battles were still ahead of him, but the fourth round at Tilburg, played in October 1991, gave an early signal of his imagination. His opponent, Jan Timman, was the leading Western player of the 1980s, a three-time Candidates finalist and the pride of Dutch chess. A game between these two carried genuine weight, and it produced one of the strangest finishes in the long record of chess history.

How the Position Arose

Timman chose the Alekhine's Defence, meeting 1.e4 with 1...Nf6 and inviting Short to overextend in the centre. Short declined the invitation. He developed soberly with 5.Bc4 and 6.Bb3, kept his space, and steered the game into a middlegame where his pieces simply stood better. The critical phase came between moves 20 and 30. Short's rook invaded on d8 and then settled on d7, the classic seventh-rank outpost. After 23.Bh6 and the exchange of dark-squared bishops on g7, his queen arrived on f6, eyeing the dark squares around Timman's king. By move 30 the picture was grim for Black: White's queen sat on f6, rooks were doubled with one buried on d7, and every black piece was tied to defence. Timman's queen on c6, rooks on the back ranks and bishop on c8 had no useful moves. Yet for all that domination, Short faced a real problem. His pieces controlled everything, but they could not actually deliver mate. The g7 square was covered, and no combination of queen and rooks could break through. He needed one more attacker, and the only unit left was his own king.

The March: Kh2, Kg3, Kf4, Kg5

Short began with 31.Kh2, a move that could pass for routine prophylaxis. Then came 32.Kg3, and the spectators understood. The king was not sidestepping a check, it was travelling. 33.Kf4 followed, the white monarch striding into the centre of a board that still held both queens, and after Timman's desperate 33...Bc8, hitting the rook on d7, Short played 34.Kg5. The plan was now naked: 35.Kh6 next, after which Qg7 would be checkmate, the queen defended by the king standing beside it. Timman resigned on the spot. Even 34...Bxd7, winning a whole rook, fails to 35.Kh6, when mate on g7 cannot be stopped. The march works because of a single calculated fact: Black has no checks anywhere on the route. Short had played 30.h4 and met 30...h5 first, fixing the pawns so that the path h2 to g3 to f4 to g5 was clean. It is one of the most precise pieces of calculation in tournament play, dressed up as the most flamboyant gesture imaginable. Students of chess tactics still use the finish as the canonical example of the king as an attacking piece.

Why the Move Is Extraordinary

Every beginner is taught the same law: in the middlegame, with queens on the board, the king hides. Centuries of chess strategy rest on that principle, and the exceptions are almost all found in chess endgames, where the queens have gone and the king becomes a fighting unit. Short's march violates the law so completely that the position looks like a composed problem rather than a game between two of the world's top ten. What makes it sound rather than merely showy is the logic underneath. Short had achieved total domination first. Black's pieces were not just passive, they were paralysed: the queen cut off, the rooks chained to the defence of g7 and f8, the bishop a bystander. Once no check was possible, the white king was as safe walking up the board as it would have been behind its pawns. The march was not a risk. It was the cheapest way to add a decisive attacker, and the only way to finish the game. Other king walks exist in the literature, but none so pure: queens on, a world-class defender, and a king strolling four squares up the board to administer mate personally.

Aftermath and Legacy

The game flashed around the chess world within days and has never left the anthologies. It appears in virtually every modern collection of famous chess games, and 34.Kg5 is routinely shortlisted whenever grandmasters are asked to name the greatest individual moves ever played. For Short it became a calling card, the proof of a fearless and original talent, and it foreshadowed his run to the 1993 world title match against Kasparov. Timman, sporting in defeat, later acknowledged the finish as a work of art; the two men would meet again with far more at stake in the 1993 Candidates final, which Short also won. The march has echoed through the decades since. Commentators reach for "Short against Timman" whenever a king ventures forward in a middlegame, and engines have confirmed what Short saw at the board: the walk wins cleanly. Among the famous chess players of the late twentieth century, many produced deeper games, but few produced a moment so instantly recognisable. One diagram, a white king on g5 with the queens still on, tells the whole story.

A Note on the Pieces

The game was played with standard tournament Staunton pieces, the pattern that has defined serious chess since the middle of the nineteenth century. The design takes its name from Howard Staunton, the Victorian champion who endorsed it, and it was created in 1849 by Jaques of London, working with Nathaniel Cook, who registered the design. The Staunton chess set solved a real problem: before 1849 competing pattern styles made pieces ambiguous, and a king that looked like a bishop was no small matter. There is a pleasing irony in that history meeting this game. The Staunton king, with its distinctive cross finial, was designed to be unmistakable at a glance, and at Tilburg in 1991 it was the king itself, marching in plain sight to g5, that decided a contest between two of the strongest players alive.

Frequently Asked Questions

Q: What exactly was Nigel Short's king march against Timman?

In the fourth round of the Interpolis tournament at Tilburg in 1991, Nigel Short reached a winning middlegame against Jan Timman with queens still on the board, but his pieces alone could not deliver mate. So he walked his own king up the kingside: 31.Kh2, 32.Kg3, 33.Kf4 and 34.Kg5, with the plan of playing Kh6 next move. Once the king reached h6, the threat of Qg7 mate, supported by the king itself, would be unstoppable. Timman resigned after 34.Kg5 because nothing could prevent the plan. The king, normally a piece to be hidden away in the middlegame, became the decisive attacker, which is why the game is celebrated as the most famous king march in modern chess.

Q: Why couldn't Timman simply check Short's king as it marched up the board?

Because Short had calculated, precisely, that Black had no checks at all. By move 31 Short's queen sat on f6, his rooks doubled on the d-file with one rook entrenched on d7, and Black's forces were completely tied down. Timman's queen on c6 had no path to the white king, his rooks on the back ranks were chained to defending against mating threats on g7 and f8, and his bishop on c8 could only watch. Every square on the king's route, h2, g3, f4 and g5, was calculated to be immune from attack. The march only looks reckless. In reality it was as safe as a stroll behind a wall, because Black's army had been paralysed first. That is the deeper lesson of the game: domination comes before the march.

Q: Did Short's king actually reach h6, and what was the final threat?

No, the king never arrived. Timman resigned immediately after 34.Kg5, with the white king standing on g5 in the middle of a queen middlegame. The point of the resignation was that 35.Kh6 could not be prevented, and once the king stood on h6 the move Qg7 would be checkmate, since the queen on g7 would be defended by the king itself. Black had tried 33...Bc8 to hit the rook on d7, but after 34.Kg5 even capturing the rook with 34...Bxd7 changes nothing: White simply continues 35.Kh6 and mates on g7. Resigning a move before the spectacle completes is common in master chess, but it slightly cheats the gallery. The diagram of the king on g5, queens on, is iconic anyway.

Q: What opening was played in Short vs Timman, Tilburg 1991?

The game began as an Alekhine's Defence: 1.e4 Nf6 2.e5 Nd5 3.d4 d6 4.Nf3 g6, a provocative choice by Timman that invites White to build a broad pawn centre and then tries to undermine it. Short met it with the calm 5.Bc4 and 6.Bb3 development, kept his space advantage, and gradually outplayed his opponent in the early middlegame. By move 24 his rook had landed on d8 and then d7, and after exchanges on g7 his queen penetrated to f6, fixing the bind that made the king march possible. The opening itself was respectable for Black; it was the middlegame handling that went wrong. The game is a good study in how an opening designed to provoke can backfire into passivity.