Re: XieXie Master -- a recommendation

On Feb 23, 4:01=A0am, mark <mschr...@[EMAIL PROTECTED]
> wrote:
> On Feb 22, 2:37=A0am, mikeh...@[EMAIL PROTECTED]
 wrote:
>
> You say even in 2028 that a 24 game match with long time controls Xu
> Yin-chuan will draw the match with the computer. I think in 2013 that
> Xu Yin-chuan will lose such a match. We will wait and see.

if he practices for say, a month's time, using the strategy of
"playing for a draw" for every game (even when he's red), then yes,
that's exactly what i'm saying.  he may have to systematically study
his opening strategy in order to find the most drawish lines, but
those are just details.

if, OTOH, he plays using the strategy of trying to win every game
(i.e. maximizing his EV), then i think he'll lose simply because
humans inevitably blunder while computers do not (programming bugs
notwithstanding).

think: tinsley (if he was still around and in top form) vs chinook.  i
think xiangqi is getting there (w/ respect to Xu).

> Can you read the results to the 2008 human computer match
athttp://chess.c=
c.ntu.edu.tw/?

no.