(;GM[1]FF[3]
US[Brought to you by No Name Go Server]
CP[
  This game was played on the No Name Go Server
  Permission to reproduce this game is given.]
GN[Computer go ladder game]
EV[None]
RE[B+Resign]
PW[GottaGo]
WR[NR ]
PB[Gnu Go 2.7.210]
BR[6k*]PC[NNGS Worldwide: nngs.cosmic.org 9696]
DT[2001-01-15]
SZ[9]TM[480]KM[0.5]
C[This game was played on the computer go ladder between GottaGo and Gnu Go.
Gnu Go won by resignation after a relatively peaceful game, but still not without
chances to upset things.
]
;B[gd]
C[Unusual opening for GNU Go. The most common opening is in the middle at E5.
]
;W[eb];B[fg];W[cg];B[gb]LB[dd:A]
C[Steady, but slow. A better approach would have been to split white by a move
somewhere around D6 (A), and then attack both groups.  White could possibly live
with both groups, but not get very much territory in the process. However, GNU
Go doesn't know about splitting attacks.
]
;W[gg]
C[This move could lead to anything. It is a probe to see how black answers it.
]
;B[gf]
C[GNU Go liked G2 almost as much as G4 because it thinks it is a joseki move (which
of course should not be applicable at 9x9 play). But clearly G4 is much better since
it loosely connects F3 and G6. Another possibility here would be to play F4, let white
live in the corner and then use the overwhelming influence to crush either C3 or E8
or both if white makes a few bad moves.
]
;W[ff]
C[This is tricky. An easy trap to fall into here is to atari at G2 and after white E3 be
completely surrounded with almost no chance in the game. See variation.
]
(;B[eg]
C[If cross-cut, extend.  So GNU Go makes the right move in this situation, but for the
wrong reasons. It thinks that it attacks both F4 and G3, but if white really wanted, F4
could of course be saved. Still, this is the correct move here since it avoids being
shut in. White can still live here starting at H4 which leads to a standard shape.
]
(;W[hf];B[ge];W[cc]
C[White abandons the corner for some security on the upper side. Probably a wise
decision, but the sequence starting at G3 has lead to nothing.
]
;B[hg]
C[This kills the corner definitely, but is slow. C5 (A) would have been much better. If white plays C5 or D5 now, the game is close.
]
LB[ce:A];W[gh]C[White can't do anything here.
];B[hh];W[ig];B[ih]
C[F2 (A) would have been better because it connects all black stones firmly and leaves
fewer forcing moves on the outside.
]
LB[fh:A];W[hc];B[gc];W[ib]C[What?
];B[ef]LB[gg:A]
C[This was of course the time to play C5 and wrap up the game. But C5 isn't even 
considered because of missing patterns. 
]
;W[fh];B[eh]
C[This exchange was bad for white since it strenghtens blacks influence to the left.
]
;W[hi];B[if];W[ce]C[At last! Now there is only the endgame left.
];
B[bi];W[dh];B[di];W[ie];B[he];W[ed]
C[Takes a few points, but the correct move is B2.
];B[bg]
C[This move looks big, but is actually wrong because it can be captured (see variation).
The correct move was C2.
]
(;W[bf];B[bh];W[ei];B[ci];W[dg];B[fe]LB[de:A]
C[Better to play the forcing move at D5 (A) first. And if black wants to capture, E5 is a 
better way to do it.
]
;W[df];B[ee];W[ag];B[ah]C[D6 is bigger.
];W[ch]
C[This move is a big mistake. If black captures at A4 (which actually happened), black has
both a deep incursion into whites territory at A6 and the threat to capture the D4 group
at D5. White can't defend from both. Besides, this move is pure dame.
]
;B[af];W[de];B[ad];W[bd]C[
];B[ae]C[And here white resigned.
])

(;W[bh];B[ah];W[ch]
(;B[ci];W[bf];B[ag])

(;B[cf];W[bf];B[ag];W[af])
)
)

(;W[hf];B[ge];W[gh])
)

(;B[gh];W[eg]
(;B[fh];W[ge];B[hg];W[fe];B[he];W[eh]C[White has a big lead.
])

(;B[hg];W[ef])
)

)