[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [gnugo-devel] Bad testcases
From: |
Gunnar Farneback |
Subject: |
Re: [gnugo-devel] Bad testcases |
Date: |
Wed, 02 Apr 2003 17:26:35 +0200 |
User-agent: |
EMH/1.14.1 SEMI/1.14.3 (Ushinoya) FLIM/1.14.2 (Yagi-Nishiguchi) APEL/10.3 Emacs/20.7 (sparc-sun-solaris2.7) (with unibyte mode) |
I wrote:
> The following test cases have at least one occupied vertex among the
> correct (or incorrect in the negated case) answers:
Proposed changes:
> strategy2:98
N6 -> N7
> trevorc:740
This test seems to have been intended for move 22 (by white) rather
than move 23.
> trevorc:880
Not sure what was intended with J4 but it seems best to replace [!J4|H4]
with [J10].
> 13x13:42
M3->M4
> 13x13:50
Possibly M10 should be M8, the actual game move. Arend?
> handtalk:6
F16 was probably intended to be the game move at E16. It doesn't look
all that horrible, although G13 is locally better. I propose changing
the test to
6 restricted_genmove black E16 G13
#? [G13]
> handtalk:10
No idea what was intended instead of D8. We have two other answers and
a comment that more moves are possible. For now I propose just
deleting D8.
> nngs3:310
C5 should be C4?
> nngs4:470
R9->R8
> ninestones:560
> ninestones:570
C11 -> D11
> ninestones:615
K16 -> J16
> tactics1:106
No idea what F9 should be.
/Gunnar