gnugo-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[gnugo-devel] Bug in critical status ?


From: Stéphane Nicolet
Subject: [gnugo-devel] Bug in critical status ?
Date: Sat, 19 Jul 2003 10:19:56 +0200



Hello,

Here is a strange case involving a change in the critical
status of a group in a real game, as opposed to static analysis.


When feeded with the attached game, the following command
shows that the N4 white dragon is considered critical at
move 68, and gnugo accordingly plays 06 to help it:

    gnugo -l test_match_5_012.sgf -T --until 68


However, during the game the engine plays something completely
different because the N4 dragon is no longer considered critical
at move 68, as confirmed by the following command :

    gnugo -l test_match_5_012.sgf -T --replay white



Does anybody know why the N4 white dragon is critical  when the
engine is started at move 68, but not when GNU Go plays the game?
Or, maybe more importantly, where I should look in the
code to try to figure it out?

Stephane



(;GM[1]FF[4]RU[Japanese]SZ[19]HA[0]KM[6.5]RE[W+1.5]
PW[GNU Go 3.3.23 (random seed 104077214)]PB[GNU Go 3.3.23 (random seed 104077214)] ;B[pd];W[dp];B[qp];W[dd];B[fq];W[op];B[oo];W[pg];B[nd];W[lc];B[cf];W[qk] ;B[ci];W[fe];B[iq];W[dn];B[cl];W[qe];B[qd];W[ce];B[df];W[nb];B[re];W[no] ;B[po];W[mp];B[oq];W[nq];B[or];W[cn];B[dr];W[be];B[cq];W[le];B[nf];W[rf] ;B[rc];W[se];B[rd];W[fg];B[lg];W[jg];B[nh];W[pi];B[kf];W[jd];B[jf];W[he] ;B[bf];W[af];B[ag];W[ae];B[bh];W[pb];B[qf];W[qg];B[pf];W[rg];B[og];W[qm] ;B[kp];W[go];B[rn];W[fp];B[gq];W[fj];B[io];W[qn];B[nn];W[pq];B[pp];W[ro] ;B[np];W[lq];B[mo];W[nk];B[rp];W[rm];B[if];W[sp];B[sq];W[so];B[qq];W[hm] ;B[ig];W[ij];B[gp];W[ep];B[kn];W[dl];B[dk];W[el];B[jh];W[cm];B[bl];W[ek] ;B[dj];W[kk];B[bo];W[bn];B[co];W[jm];B[ho];W[hn];B[qb];W[ao];B[ap];W[an] ;B[cp];W[al];B[ak];W[am];B[bj];W[mi];B[oc];W[ob];B[oi];W[oj];B[mh];W[ni] ;B[oh];W[ph];B[mc];W[mb];B[li];W[lj];B[ki];W[hh];B[ie];W[id];B[hd];W[hc] ;B[do];W[eo];B[md];W[ld];B[jj];W[jk];B[kj];W[mj];B[ii];W[hi];B[hf];W[gd] ;B[qo];W[sn];B[jn];W[eh];B[gg];W[gh];B[gf];W[dg];B[ei];W[cg];B[bg];W[ge] ;B[ih];W[sd];B[sc];W[sf];B[ol];W[nl];B[nm];W[pl];B[om];W[ml];B[lm];W[ll] ;B[mm];W[qa];B[ra];W[pa];B[ee];W[ed];B[ef];W[fi];B[ok];W[ej];B[di];W[pj] ;B[eq];W[km];B[dq];W[fn];B[me];W[in];B[lf];W[ke];B[pn];W[pc];B[qc];W[nc] ;B[od];W[ff];B[eg];W[dh];B[ch];W[fh];B[pm];W[pk];B[je];W[de];B[bm];W[hg] ;B[]
;W[]
)

Attachment: test_match_5_012.sgf
Description: application/text


reply via email to

[Prev in Thread] Current Thread [Next in Thread]