Mike Murray wrote:
"J.D. Walker" wrote:
How would the following work out?

White win = +1
Black win = +1.11
White draw = .45
Black Draw = .50

Looks good. As a side benefit, I'd guess that it would cut back the
number of times tie-breakers would need to be applied.

Hmm. Now I think about it, wouldn't it actually be better to use a
tie-break, at least for matches? If the match is drawn under conven-
tional 1-1/2-0 scoring, declare the match to be won by the player who
won the greater number of games with black or, if that's still a tie,
by the player who did X or, ... And then award the match to the
champion if it's still tied after some reasonable list of tie-break
criteria.

My reasoning is that any system of differentially scoring wins and
draws for black and white implicitly defines some sort of tie-break
system for `tied' matches but one has to sit down and work out what
the system is. (For example, in the original proposal of 1 for a win,
0.45 for a white draw and 0.55 for a black draw, the system was the
rather counter-intuitive and probably unintended `The player with the
most wins as white wins the match'.) Rather than try to come up with
a scoring system that implicitly defines some tie-break criteria,
wouldn't it be easier to just explicitly define the tie-break?


Dave.

--
David Richerby Crystal Atlas (TM): it's like a map
www.chiark.greenend.org.uk/~davidr/ of the world but it's completely
transparent!