Lomonosov tablebases

Also known as: Lomonosov endgame tablebases

Precomputed databases that provide exact results for endgame positions containing no more than seven pieces.

This page is temporarily shown in English. Complete localization will be added in a later stage.

Explanation

An endgame tablebase is a collection of positions solved in advance. Instead of calculating from scratch at query time, software looks up the position and retrieves its exact result within the covered set. The answer can show that the position is drawn or that one side can force mate, together with a mating distance when that information is provided.

The Lomonosov tablebases cover positions with no more than seven pieces in total, including both kings. They were calculated in 2012 at Moscow State University using the Lomonosov supercomputer. The amount of material, not how late the game appears, determines whether a position is covered.

The Lomonosov tables use DTM, meaning distance to mate. In a won position, this measure counts the winner's moves to checkmate when the winning side seeks the shortest route and the defender tries to delay it. DTM does not by itself enforce the rule allowing a draw claim after fifty moves by each side without a pawn move or capture. A theoretical tablebase win can therefore end as a draw in official play when that rule intervenes.

The queried position must exactly match the tablebase representation, including which side is to move. Piece placement alone is not enough: changing the turn can completely change the result.

This is not the same as asking a for an evaluation. An engine searches moves and estimates a position with limited resources; a tablebase returns a precomputed solution for a covered position. Outside the seven-piece coverage, the Lomonosov tables cannot provide that answer.

Common confusions

Tablebase versus engine

A tablebase retrieves an already solved answer for a covered position. An engine searches and evaluates during the query.

View term

© __2026__ __MindZug__