[gnugo-devel] GnuGo and Free Handicap

[Allan Crossman]

[Sorry Dan - You'll get this twice - I accidentally sent it to you instead 
of the whole list]

On Saturday, October 13, 2001, at 07:36 PM, Daniel Bump wrote:

Could you submit some SGF files to illustrate these statements?

Sure. Here's two games. In both cases, I was Black, but passing a lot to 
let GnuGo have a free handicap. In the second game, I played one stone 
then repeatedly passed.

GnuGo's play in the second game seems markedly superior, except for what 
happens near my one stone. And this isn't a one off result - it's always 
this way.

First (bad) game:

(;FF[4]GM[1]
SZ[19]
GN[GNU Go 3.1.9 Random Seed 1002998502 level 10]
DT[2001-10-13]
KM[0.0]AP[gnugo:3.1.9]RU[Japanese]
 ;B[]
 ;W[dd]
 ;B[]
 ;W[dc]
 ;B[]
 ;W[pp]
 ;B[]
 ;W[qp]
 ;B[]
 ;W[dp]
 ;B[]
 ;W[cp]
 ;B[]
 ;W[oc]
 ;B[]
 ;W[od]
 ;B[]
 ;W[pl]
 ;B[]
 ;W[ql]
 ;B[]
 ;W[dl]
 ;B[]
 ;W[cl]
 ;B[]
 ;W[hd]
 ;B[]
 ;W[hc]
 ;B[]
 ;W[kd]
 ;B[]
 ;W[kc]
 ;B[]
 ;W[ch]
 ;B[]
 ;W[dh])

Second (good) game:

(;FF[4]GM[1]
SZ[19]
GN[GNU Go 3.1.9 Random Seed 1002998650 level 10]
DT[2001-10-13]
KM[0.0]AP[gnugo:3.1.9]RU[Japanese]
 ;B[as]
 ;W[br]
 ;B[]
 ;W[qd]
 ;B[]
 ;W[dd]
 ;B[]
 ;W[qo]
 ;B[]
 ;W[oc]
 ;B[]
 ;W[pq]
 ;B[]
 ;W[qj]
 ;B[]
 ;W[jq]
 ;B[]
 ;W[kd]
 ;B[]
 ;W[fp]
 ;B[]
 ;W[dj]
 ;B[]
 ;W[oj]
 ;B[]
 ;W[qg])

Allan Crossman