Single Comment

First move advantage in Western Chess - why does it exist?[Subject Thread] [Add Response]

H. G. Muller wrote on Mon, Aug 13, 2012 12:09 PM UTC:

The low draw rates could be due to the very fast time control (10
moves/sec), which might lead to bungling drawn pawn endings because an
unstoppable promotion is beyond the horizon. For 10x8 Chess the normal draw
rate is 16%, though, so the numbers are not that far off.

However, the statistical significance is weak. They were all measured from
about 1000 games, which would have a standard deviation of ~1.6%-points in
the average result. But since many different positions were measured, it is
extremely likely some of them would be off by more than one standard
deviation, in either direction. So in an experiment like this (doing a
dozen series of 1000 coin flips with a perfectly fair coin), the difference
between the highest and the lowest percentage of heads between the series
will almost always be 6%.

And of course it cannot be excluded that one of the setups is not
tactically quiet (e.g. because of unprotected Pawns in the array). In a
tactical situation the advantage of having the move can of course be
enormous. (E.g. 2 Queens, when unprotected Queens are attacking each other,
and 4 Queens when in addition each side has a passer that can only be
stopped by the opponent's Queen.)

But in principle this is the correct way to measure the first-move
advantage. If it is in doubt whether variants with only short-range pieces
have a first-move advantage, just let a computer play a few thousand games,
and count the number of black and white wins. (E.g. using Fairy-Max and
letting it play Great Shatranj, which is a supported variant with onlty
short-range pieces.)