[computer-go] an idea... computer go program's rank vs time

[Sanghyeon Seo]

2007/1/22, Darren Cook <darren at dcook.org>:
> A couple of related comments. First, in the 2-day games the pros spend
> almost all their thinking time in the opening, i.e. considering
> different joseki and how they work together.  By the time they get into
> the endgame they are playing almost all moves in under a minute (and
> most of that time is spent checking their counting to see if they should
> play the quiet move or if they need to complicate things).
>
> Second, the very top pros all seem to say their endgame is the strongest
> and most important part of their game. I.e. not the moves they spend
> most time on. That always strikes me as strange :-).

Note that professionals do not play perfect endgame, actually they are
far from it. Bill Spight soundly proved this with his "Go Thermography"
paper analyzing endgame of Jiang 9p and Rui 9p. He found 5 blunders
in the late endgame sequence of ~50 moves, all with value less than
4 points.

Also, post-mortem analysis of pro games published in go magazines
routinely finds some game-result changing improvements in the endgame.
And sometimes, even amateur readers can understand them, once
better sequence is found.

-- 
Seo Sanghyeon