Concept notes

Amashi アマシ

Amashi and amaru are two words in normal Japanese which have special meanings in go. Segoe Kensaku said the go usage was common among pros but was very difficult for amateurs to understand. It was also very hard for pros to explain the meanings to amateurs. He did, however, make a decent stab at this in an article in Kido in 1954. We will use this to define the concept here.

Normally, amaru 剰る means 'be left over'. Amasu is the transitive form: 'to leave over'. It is probably safest for the non-Japanese speaker to ignore that, though, as any attempt to find those meanings in the go usages is problematic.

A further difficulty for the western player is that the Japanese usages are grammatically wider ranging. Western usage seems to focus, however hazily, on amashi as a noun, and there may also - though very rarely, it seems - be a mention of amarigatachi, another noun. In Japanese, however, the commonest forms are amarigatachi アマリ形 and various forms of the verb amasu, in particular the two forms amashite utsu アマして打つ and amasareta アマされた. The latter (a passive form) can be regarded as meaning 'to suffer amarigatachi' and so we can ignore it, so long as we know what amarigatachi means.

Segoe chose not so much to define it, but to give a familiar analogy. He equated it with the shogi term sashikiri 指切り. This is used when a player's attack runs out of steam, usually because he has no more useful pieces in hand and all his previous attacking pieces have been captured. You can see the same idea in chess. You sacrifice your queen and maybe a couple of other pieces only to discover that the combination doesn't actually work. Although the term amarigatachi is not usually used in shogi, the attacker could be said to have fallen into an amarigatachi position (the -gatachi portion means 'shape' or 'position'), and the defending side (assuming he was actually skilful enough to have anything to do with this) could be said to have played an amashi strategy (which is the locution we shall use for amashite utsu).

This shogi analogy illustrates the problem with trying to rely the base meaning of the words. Amaru means 'to be left over' yet the term is used of a player who has nothing left!

Segoe then concedes that an example would be easier to follow. He focuses first on the concrete noun amarigatachi.

The diagram below shows a White tactic in a handicap game. Black has fallen into White's trap and will now have a tough game on the left side of the board. Black 8 is a mistake.

The next diagram shows what Black ought to have played. White has been brought to a state of collapse. This is an example of amarigatachi (White's situation) but do be careful not to assume it refers specifically to White's disastrous shape in the corner (despite the common use of -gatachi to mean a shape).

In practice, having seen that Black cannot be gulled, White would be forced to find a way to live, as in the next diagram.

The point to note is that, despite living, White is still in amarigatachi. As Segoe explains it, it denotes the result of White having made an overplay and having been punished for it. The punishment doesn't even have to be this dire.

In the examples above, White can be excused for making an overplay, perhaps, as he is in a handicap game. But it is important to realise that you can overplay and fall into amarigatachi even when you dominate the position. In the example below (also from Segoe), Black starts in the left quadrant with three stones to White's one, and has sente to boot. Yet, in spite of the 4-to-1 superiority, Black 1 is an overplay. The diagram shows why. Now it is Black who suffers amarigatachi.

The specific reason is that Black has to break off from his attack to defend at 7. He has lost a tempo. He can't be in two places at once, so now White gets the initiative on the outside. Black, as Segoe puts it, has forgotten the maxim to watch one's back when attacking.

White 8 threatens to play at 9, so Black has to defend in the corner yet again and to allow White 10. Whether or not Black chooses to cut at 11 he ends up with his attack achieving nothing while White emerges unscathed. That is the typical result of amarigatachi. The correct strategy for Black is to play Black 1 at 7 (and then threaten to attack with 1). White can then move his stone out easily, but in gote and without any sort of useful shape, and Black has lost nothing on the outside.

Below is one more example, from a different author, of falling into amarigatachi. White 3 is the overplay. In fact it is a well known trick play (hamete) which hopes that Black will play 8 at 9, and then White plays A. But Black 8 is more than good enough. The resulting White position may look strong, but it has cutting points at B and C, and White has used an extra stone anyway. This is amarigatachi for White - a highly unsatisfactory position as the result of overplaying.

This is all rather tactical and probably easy enough to see when the various mistakes are pointed out. But this doesn't add much to the sum of go knowledge. Amarigatachi is just a descriptive term, really. Where things get interesting is with playing amashi strategy. The ways to get the opponent to fall into amarigatachi are very subtle. Of course, White will often play this way in handicap games, but top pros use it against each other, and successful examples end up in anthologies. One of the most famous cases is shown in these pages (see the index page), and is actually the one given by Segoe, so that can be considered a continuation of this page.

There are many examples in no-komi games where it became an essential strategy for White in overcoming Black's advantage of first move. This was not limited to Japan. Go Seigen has claimed that Huang Longshi, a contemporary of Honinbo Dosaku, was the first to demonstrate mastery of the technique, in 17th century China.

But amashi is also alive and well in the modern game, and is popular (for both Black and White) with masters such as Cho Chikun. Even if you find it hard to use, it is a joy to see it used.


© John Fairbairn & T Mark Hall (GoGoD), London 2009.