"Peter Clinch" <p.j.clinch@[EMAIL PROTECTED]
> wrote in message
news:64injaF2ce0ajU1@[EMAIL PROTECTED]
>
> But since losing gets as much worse as winning gets better, and
combative play
> is often riskier, you haven't really changed anything in the net
risk/reward
> equation. It still averages out.
This is incorrect. In anti-draw scoring, drawing and
losing becomes closer. So it pays to take risks that
give you winning chances, even if you might lose.
Earlier I gave an example: A player can chose
between two moves. In the first, he can simplify the
position to an ending and he estimates his
chances as: W=0, D=0.9, L=0.1 In the second move,
launching a counterattack, he estimates his chances
to be W=0.3, D=0.3, L=0.4.
With 1867-scoring (Wins = 1 point, draws = 1/2 point
and losses =0), the expectation of either move is
identical (0.45). Now imagine that we used "soccer"
scoring, for example (Wins = 3 point, draws = 1 point
and losses =0) Now the simplifying move is worth 0.9,
and the counterattacking move is worth 1.2
Without changing the underlying rules of chess, we've
altered the meta-game to provoke more interesting
play. Just because something seemed reasonable to
try 141 years ago, doesn't mean it makes sense for
today's game.
>
> Pete.
> --
> Peter Clinch Medical Physics IT Officer
> Tel 44 1382 660111 ext. 33637 Univ. of Dundee, Ninewells Hospital
> Fax 44 1382 640177 Dundee DD1 9SY Scotland UK
> net p.j.clinch@[EMAIL PROTECTED]
http://www.dundee.ac.uk/~pjclinch/
|
