Promoting Go in the UK and beyond

British Go Journal No. 60. September 1983. Page 22.

Black: Jim Bates, 4d
White: Richard Granville, 3d
The game-file in Ishi and SGF format.

This game from the Challenger's league is typical of many in that the mistakes by which it was lost and won were at a far lower level than most of the moves. It is cause for much regret among those of us, who love Go for the amazing range of interesting ideas it contains, that simple blunders cause so many of our resignations.

Richard Granville is White, Jim Bates Black, comments are by Matthew Macfadyen.

Figure 1 (1-68)

• 1-24: Reasonable for both - Although Black gets two shimaris, white 22 is an excellent point (Black could consider playing 21 one point to the left of 22).
• White 28 is a bit strange - the usual idea is to play hane at A instead. 28 would make more sense if white 6 were one line closer, but the shape in the game leaves Black with good chances to invade the side later.
• Black 35: Jim grabs his fourth corner - this play may look small but it is essential. Once the corner is secure it becomes relatively easy to live in the centre.
• Black 49: Bad - he obviously feels that he is strong enough in this area to kill White completely, but the game sequence leaves many weaknesses.
  Dia 1 shows the obvious sequence - White gets a favourable ko (if he wins it with A, he can continue at B). Black can do better than this by playing 4' at 5, but the result is still ko (exercise for the reader).
  If Black had played 49 correctly at X in the figure then there would be no chance at all for White to live in the corner, and Black's stones 23 and 47 would be nicely placed to trap him if he tried to run away.
• White 54 threatens to surround the centre on a large scale. Black 55 is a good, light invasion, but 57 is an overplay - he must settle himself in the centre before playing here.

Dia 1

Figure 2a (69-100) BGJ had Fig 2a and Fig 2b as one diagram, Fig2.

• Black 69 is reasonable, but 71 is not. It is neither necessary, sufficient, nor indeed possible to live so deep in the white moyo.
  Even after 73 Black can cut his losses and abandon these stones for the moment. Jm Bates is quite strong enough to know this, but he fell into a common psychological trap, letting himself continue with a bad sequence even after he knew it was bad. It is extraordinarily easy to do this - the sensation is rather like a bad dream - you just sit and watch your hands playing a string of ridiculous moves and are quite powerless to stop them.
• White 82: Now it was Richard's turn to err despite himself. He had spent several minutes thinking about 76, but had clearly not thought about the consequences of black 81 while he was doing so. Nevertheless, when 81 was played, he replied at 82 after about half a second's thought. Had he devoted an extra few seconds to the problem, he should have had no trouble in spotting that 83 would be a much better play than 82. this makes miai of 82 and 121 and kills the black stones cleanly. Again the problem was not in knowing what to do, but in actually doing it.
• Up to 97, Jim has cause to rue his premature plays 57 - 65.

Figure 2b (101-135) BGJ had Fig 2a and Fig 2b as one diagram, Fig2.

• 98 - 114: Black's helplessness is becoming progressively more apparent.
• Black 115 - 121: Jim succeeds in cutting off the side group and connecting the ko, but he seems to have too few liberties to be able to do anything.
• White 124: However, Richard was in byoyomi by now, and this is the blunder Jim was waiting for. It gives Black time to connect his group together. Now the capturing race is looking close.
• White 128: Richard felt that this was necessary to keep him ahead in the capturing race, but it fails simply to the sequence to 135. (after 135 Black kills three stones below or five stones above). White 134 should be to the right of 133, but that gives Black his extra liberty anyway.

Jim Bates (Black) wins, presumably by resignation.
BGJ did not have this paragraph. Details from table on a previous page and from comments on this page.

If White had played 128' at 133, then a naive count of liberties seems to show that he would be one liberty ahead. Actually things are not so simple. One possbility is discussed overleaf.

Dia 2

Dia 2 shows how the semeai might develop. Black 2 there is essential to stopWhite making eyes, white 3 is also essential - otherwise Black can throw in at 3 and gain an extra liberty (the one at 14, which would be atari and therefore not cost a move). If Black continues to fill in liberties then he will lose the race. Black 4 is a very odd looking play, but it seems to be the only way to win - after 16 both sides have two liberties, but White cannot play either A or B.

Both players were in byoyomi and more or less panicing, so it is unlikely that this sequence would actually have followed, but our readers may derive some amusement from seeing what might have happened (actually, all of the moves in this diagram can be found quite easily by counting liberties, and applying the Holmesian technique of eliminating the impossible systematically and then seeing what is left).