Computer Chess Club Archives


Search

Terms

Messages

Subject: A simple note on KQK ending

Author: Guido

Date: 07:33:04 03/14/05


The EGTB of KQK shows that there is one only position on the board where white,
having the move, wins in 10 moves.
The position is the following: wKa1, wQb2, bKf5.

In this position the results of the possible legal moves are:

- 22 moves: win in 10 moves
- 2 moves: draw

The last two cases (wQb2-e5, wQb2-f6) are clearly related to the possibility of
bK capturing wQ.

Now the question is: is it only casual (or better pseudocasual) that all the
winning moves are perfectly equivalent, i.e. victories in 10 moves, and have we
to wonder at this?

My answer is no, as it is easy to demonstrate.
In fact, if it is evident by hypothesis that no win can be shorter than 10 moves
and that, if the wQ is captured, only a draw is possible, we can suppose by
absurd that a victory for white exists in 11 or more moves.
But after the white has done such move and the black has answered with another
move, the new reached position will never be a 10 or more moves victory position
for the white because such position will be different from the only one winning
position in 10 moves reported above. So what we suppose is absurd and the
demonstration is complete!

At a first sight it seems a strict reasoning and actually I can confirm that it
is correct for KQK but it is not right in two cases:

- when the move of the white and the answer of the black can change the
composition of the ending  by capture or promotion. In KQK this happens in the
two draw cases but without influencing the winning moves. In other ending it
could happen that some moves of the white are still victories but in more moves
than the maximum DTM, or are draws or also are losses.

- when after the move of the white and the black the original position has again
reached!

This second case would seem impossible but suppose that the KQK position of
victory in 10 moves had been wKa1, wQb1, bKf5  than the white move wQb1-a2 would
be a victory in 11 moves, because black would have answered with bKf5-e6
obtaining a position equivalent by symmetry to the initial. But this doesn't
happen with KQK, where the above mentioned position doesn't offer this
possibility, because wK and wQ are both on the main diagonal.

For association of ideas I finish noticing that the three repetition rule
doesn't keep into account positions equivalent by symmetry, even if I think that
a similar case would be extremely unlikely.

Guido







This page took 0 seconds to execute

Last modified: Thu, 15 Apr 21 08:11:13 -0700

Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.