The reason for the 50-move rule, along with the triple-repetition rule is to make chess definitely finite. How did one arrive at the number 50? What would change if it was 40 or 60? Is the number 50 chosen carefully?

  • I've never seen the 50 move rule actually invoked, since maybe 6th grade.
    – Tony Ennis
    Commented Jun 10, 2012 at 16:15
  • 2
    @TonyEnnis, that's because you haven't seen this (utterly absurd) game. (Actually, the 50-move rule isn't invoked in this game, but Nakamura tricks the computer into making errors when it tries to avoid the 50-move rule.)
    – Kyralessa
    Commented Jul 6, 2012 at 21:35
  • @Kyralessa: Wow! Just... WOW Commented Jul 6, 2012 at 21:49
  • Don't make the mistake of showing that game to your seven-year-old, like I did. Apparently he decided that if that tedious back-and-forth stuff was good enough for Nakamura, it was good enough for him as well.
    – Kyralessa
    Commented Jul 6, 2012 at 23:54
  • @kyralessa that game is cool. You can see Nakamura figuring out how the computer is working, and then using this as a weapon.
    – Tony Ennis
    Commented Jul 7, 2012 at 1:20

1 Answer 1


At one time, it was believed that all winnable endgames could be won within fifty moves. However, in the early twentieth century, some exceptions were found, including A. A. Troitsky's (1866-1942) analysis of the two knights endgame as well as the endgame of a rook and bishop versus a rook. The rules of chess were revised several times to admit exceptions to the fifty-move rule for certain specific situations. Early on, the fifty-move rule applied to tournament games but not to match games (Troitzky 2006:197).

SOURCE: Fifty-move rule

  • 1
    This is interesting. There are some complex endgames that can not in fact be won in fifty moves - for example, multiple pieces vs. queen - but that can be won. Commented May 3, 2012 at 19:47
  • 1
    @Andrew Latham: This table of "pawnless endgames", under the "longest win" column, provides additional examples too; note that Stiller and Nunn both say 243 is the longest.
    – blunders
    Commented May 3, 2012 at 20:39
  • 2
    @Blunders: see my comment to Andrew's answer here: there's a setup in which it takes White 517 moves to force a winning endgame - not even mate!
    – Daniel
    Commented May 4, 2012 at 2:36
  • @Daniel I guess those extremely long endgames include captures (which are relevant for the 50 move rule), don't they? Commented Mar 24, 2017 at 22:41
  • 1
    @user1583209 No! It's 517 moves to get to a capture (which creates a "quickly" winnable end-game. (And the in the particular case, there are no pawns). Commented Nov 1, 2018 at 13:07

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.