269 Moves and No Winner: The Longest Chess Game Ever Recorded

Twenty Hours and No Result

The game was played at the Yugoslav Chess Championship in Belgrade in 1989. Ivan Nikolić and Goran Arsović, two competent tournament players rather than world-class grandmasters, began a game that would eventually exhaust both players, the tournament officials, and the patience of everyone following it, before concluding after 269 moves as a draw.

The game itself is not remarkable for its quality — marathon chess games tend to degenerate into technical endgames where both players are simply trying to avoid losing rather than attempting to win. But it is remarkable for its duration: approximately 20 hours from first move to the final handshake, with the game resuming across multiple sessions as was standard in tournament practice of the era before computer adjudication.

To understand how a chess game can last 269 moves without a result, and why such a game led to a rule change that prevents it from happening again, requires understanding what was happening in the endgame — the phase of the game where fewer pieces remain and the path to a forced win, or the impossibility of one, becomes clearer.

The Fifty-Move Rule and Its Limitations

Chess has long had a rule called the fifty-move rule: if fifty moves pass without a pawn move or a capture, either player can claim a draw. This rule exists to prevent indefinite play in positions where one player has an advantage but cannot convert it to a win within a reasonable timeframe. The rule was originally designed with positions in mind where two or three pieces chase a lone king without being able to force checkmate.

The problem, known to chess theorists since at least the nineteenth century, is that certain endgame positions require more than fifty moves to force a win even with perfect play. Specifically, some rook-and-bishop versus rook endings require over fifty moves to deliver checkmate; certain queen versus rook endings require similar depths. Chess databases developed in the 1980s and 1990s — endgame tablebases that calculate the exact optimal moves from any position — proved definitively that some technically won positions required more than fifty moves to win with best play from both sides.

For a period in the 1980s, FIDE (the world chess federation) amended the fifty-move rule to allow extensions in certain specific endgame configurations — up to 100 moves in some cases — in recognition of these theoretical positions. This amendment was in effect in 1989 when the Nikolić-Arsović game was played, and it is precisely what allowed that game to continue to 269 moves: the players were cycling through one of the extended-rule endgame configurations without either side being able to force a win or a draw claim.

The Rule Change That Followed

The 1989 marathon, combined with the increasing use of computer analysis that made extended endgame play more calculable and therefore more exploitable, prompted FIDE to revert to the straight fifty-move rule in 1992, eliminating all exceptions. The logic was pragmatic: while theoretically some positions require more than fifty moves to win under optimal play, in practice these positions virtually never arise in over-the-board tournament games from normal play, and the exceptions created more problems than they solved.

The result is that the 269-move game between Nikolić and Arsović is a Guinness World Record that cannot be broken under current rules, because the rule that allowed it to occur no longer exists. Any game today that reached a similarly drawn technical endgame would be claimed as a draw long before 269 moves.

Endgame Tablebases and the Shrinking of Unknown Chess

The development of endgame tablebases — databases that store the exact optimal evaluation (win in N moves, draw, or loss) of every possible legal position with up to seven pieces on the board — has fundamentally changed how chess theorists understand the endgame. Positions that were previously considered drawn have been proven to be wins with dozens of moves of inhuman precision; positions thought to be wins have been proven to be draws.

The Lomonosov tablebases, completed around 2012, contain all seven-piece positions — approximately 549 trillion positions — and revealed several endgame positions requiring more than 500 moves to win under optimal play, a number so extreme that it cannot be exploited in any practical game under any reasonable time control or move-count rule. The gap between theoretical perfection and practical chess remains unbridgeable, even for endgames containing only seven pieces on a 64-square board.