________________________________
From: Nick Wedd <[email protected]>

In message <[email protected]>, Ingo Althöfer 
<[email protected]> writes

>Question (pun intended):
>Does there exist a collection of go positions
>where humans tend to evaluate wrongly whereas
>computers have no or less problems with?

Nick Wedd:

> Ishi-no-shita problems.

I had to look up ishi no shita; these are "under the stones" problems where one 
plays "under" captured stones.

They are indeed difficult for humans.

Are they, however, readily solvable by computer programs during the course of a 
game? 


We've only recently seen computers understanding nakade, and the game under 
discussion appears to involve a top program ( pachi ) failing to recognize that 
its group is dead due to nakade. In fact, Zen, which won the game, may not have 
disputed the declaration by Pachi that the group was alive.

Hence, I wonder if there are examples of pachi or zen (or any other 
highly-ranked program ) exploiting ishi no shita to win a game. 
_______________________________________________
Computer-go mailing list
[email protected]
http://dvandva.org/cgi-bin/mailman/listinfo/computer-go

Reply via email to