9x9:360

360 reg_genmove black
#? [F4|E6|D5]*

Looking at the trace file at http://trac.gnugo.org/regression/regress.plx?tstfile=9x9&num=360&trace=1 it seems that the engine would like to play D5

Top moves:
1. E2  75.01
2. D5  50.74

but is prevented to do so by

Fuseki database move at E2 with relative weight 2, pattern F-H2-332+6
Fuseki Player suggests E2 with value 75

So, who is wrong here ? The testcase submitter or the fuseki database maintainer ? At any rate, it's not the engine's fault.

-- nando