Forcing Move Misunderstandings

  Difficulty: Advanced   Keywords: Strategy

The kikashi concept is probably reasonably clear to most amateur dan players - but that understanding may depend on substantial intellectual work clearing away misconceptions. This page is inspired by the kikashi sente discussion.

A: There are forcing moves in go

No, there aren't.[1][3] There are certainly many plays that create what are objectively urgent local situations. If you play one of those your opponent may answer unexpectedly, in a perfectly valid way or in a way that can be refuted by good play on your part: for example

  1. by using a sacrifice or mutual damage plan;
  2. by finding some counter-moves which are even more urgent for you to answer, before returning to the situation;
  3. or by choosing some alternative local answer to the obvious one.
[Diagram]
Resisting a peep  

This is an interesting but dangerous idea, combining (1) and (3): B1 doesn't connect after the peep but allows White some chances to cut.

All one can say in general is that some plays do turn out to be sente; and strong players tend to agree on which those should be, anyway in many cases.


B: Forcing the opponent is good

No, in general it won't be. If you are keeping sente then that may be good. But playing moves that look forcing, hopefully rather than in a coherent way and because you can't think what else to do, is quite likely to be bad.

It depends on a number of factors: you should play real kikashi rather than thank you moves, and you should plan ahead to the point at which you will accept gote, or launch an attack, or in some other way move from improving your position to realising some gains.

So: you need at least to be able to tell which plays and exchanges do improve your position rather than just use up aji you might need later; and you need some kind of 'exit strategy' (Tokimoto Hajime makes the point that it is extremely amateurish to play a forcing move that gets an obvious answer and then think). But simply using up ko threats, for example, is just a bad way to play. You are possibly also throwing away liberties that will be needed later.

The following kinds of forcing moves are good, quite generally:

What is discussed at kikashi and influence is something different: a type of play that is good if your opponent treats it as forcing, but where that shouldn't happen. So, also

  • kikasare is (by definition) advantageous to the forcing player.
Bill: And typically a kikasare accepts a force when it should not. Although sometimes there is nothing better, and the mistake came earlier.

C: Interleaved leaning moves

Explanation: a leaning attack is conventionally set up by playing forcing moves on one side ('the left') before attacking from another direction ('the right'). Some of the forcing moves may be loss-making rather than kikashi, but are then played because a successful attack requires them, perhaps as influence towards which a weak group can be driven, perhaps to strengthen a group of the attacker's that otherwise would suffer too much.

Since losses may be taken in such a procedure, why not wait to play the 'bad' forcing moves? That is, one can imagine the 'just-in-time' attack for which each forcing move required is played at the correct moment, when one can see it is really necessary. The attack proceeds in interleaved fashion: attacking sequence/forcing plays/renewed attack/forcing plays ... et cetera.

This may look like good business, but it rarely works in practice. It might possibly do so when the forcing plays were absolute sente - threatened a group larger than the fight going on outside - but that isn't the typical middle game situation.

That is one basic flaw in the idea: intermediate plays are a problematic concept in go. The other major problem is that timing questions are often not to be resolved so simply, without deep reading. A single tempo near a thin but influential group can make a huge difference, for example between a group dying surrounded by thickness and the same group cutting the encircling forces into small pieces.

This theoretical idea has one very familiar case: the ladder-breaker.

So - it is very hard to make this concept work. Pro play is very patient, and to avoid amarigatachi attackers may wait until the oyose stage.

D: Forcing exchanges can be evaluated

Certainly some of the difficulties would go away, if there was a clear criterion distinguishing between kikashi (good) and aji keshi or thank you moves (bad).

It seems there isn't. Strong players, it appears, are much more likely to be in agreement on whether a move is really forcing, than on when it should be played because it is kikashi. Commentaries on pro games mostly seem to avoid discussing whether the pro played a correct kikashi. This may be out of respect for the difficulty of the question, and the player's right to take an 'executive decision' on the timing - perhaps a reasonable description of the type of decision involved. There are some examples where the pro gets it wrong, provably, and this is pointed out.[2]

Charles Matthews


[1] Bill: This reminds me of a joke.

One night a burglar broke into the house of a Quaker (member of a pacifist Christian sect, the Society of Friends). He was surprised when the lights came on and the Quaker was standing on the landing at the bottom of the stairs with a shotgun in his hand. The Quaker said, "Friend, I would not harm thee for the world, but thou art standing where I am about to shoot."

Now, was the burglar forced to avoid getting shot?

[3] sigs?: Actually, there are, or is. If the board is filled so that there are no other valid moves besides passing (given a ruleset that forbids suicide).

[2] Bill: When to play kikashi is to some extent a matter of style. Two extremes among top pros are Sakata and Takagawa. Sakata played kikashi early and Takagawa played them late. Their games are good to study for learning to judge when to play kikashi. If you see a kikashi that Sakata does not take, that's a good indication that it would have been a bad idea. On the other hand, if you see Takagawa taking one, then it was probably urgent.


Forcing Move Misunderstandings last edited by 82.181.222.187 on January 4, 2012 - 13:44
RecentChanges · StartingPoints · About
Edit page ·Search · Related · Page info · Latest diff
[Welcome to Sensei's Library!]
RecentChanges
StartingPoints
About
RandomPage
Search position
Page history
Latest page diff
Partner sites:
Go Teaching Ladder
Goproblems.com
Login / Prefs
Tools
Sensei's Library