Re: [gnugo-devel] OWL tuning: blocking escape with a knight's move

[Gunnar Farnebäck]

Martin Holters wrote:
>>  > However, it breaks strategy5:225 and ninestones:40. In the latter,
>>  > the semeai reading looks bogus with and without the patch and I have
>>  > not really figured out why the patch results in preferring D2 over
>>  > C3.
>>
>> I agree about ninestones:40. D2 is extremely ugly but C2 just isn't
>> working either.
>
> Maybe we should revise the regression test accordingly, also forbidding
> C2?

Yes, that would be an improvement. I think the pattern below fixes the
worst error in the semeai reading there, but I haven't checked yet how
it behaves on the rest of the tests. It may be somewhat expensive.

/Gunnar

diff --git a/patterns/owl_attackpats.db b/patterns/owl_attackpats.db
index ee75579..495e34f 100644
--- a/patterns/owl_attackpats.db
+++ b/patterns/owl_attackpats.db
@@ -3939,6 +3939,22 @@ x*.A
 ;owl_strong_dragon(B) && !xplay_disconnect(*,A,B)


+Pattern A1025
+# gf New pattern. (3.9.1)
+
+YO.X       stop escape along edge
+OX*x
+----
+
+:8,-,value(75)
+
+AO.B
+OX*x
+----
+
+;owl_escape_value(B)>0 && !xplay_disconnect(*,A,B) && !oplay_connect(*,A,B)
+
+
 #########################################################
 #                                                       #
 #              Reinforce the perimeter                  #