|Table of contents||Table of diagrams
Black and White are involved in a ko fight at ↔ a. Black to play can end the ko and win the fight by capturing seven white stones, possibly ignoring a ko threat to do so. White, on the other hand, can not end it immediately. First she will have to fill a liberty at b (or c, but b is better), possibly ignoring a ko threat. The ko is now a regular, direct ko. Thus, White may have to ignore two threats to capture the four black stones.
The example above is a one-move approach ko, simply, an approach ko. Going on, if Black had even more liberties we could also get a two-move approach ko, a three-move approach ko, and so on.
It is important to distinguish this type of ko from a ten-thousand-year ko, which also involves a kind of approach move.
The exact phrases for the terms are not agreed upon. The following occur: "indirect ko with one approach move", "one-move approach ko", "one-step approach ko", "one-move approach move ko", "one-step approach move ko". Also see the links for yet further variation. (Note: In SL the "step" terminology is deprecated: see Two-step Ko - Two-stage Ko - Discussion.)
Having to make extra moves means one must ignore two or more ko threats in order to win the fight, making it roughly twice as expensive; the details depend on the value of the threats and on whether finishing the ko is sente or gote. It also means that one must find more threats while one’s opponent gets to retake the ko without a threat.
One’s chances of winning a ko of course diminish the more opposing threats one must ignore and the more threats one must oneself find. A two-move approach ko can be considered a reasonable fight, but as the saying goes,
A three-move approach ko is no ko.
- Multi-stage ko (page aliases: Two-stage ko, three-stage ko) — a ko where fighting can move back and forth between adjacent ko mouths.
- Rogue ko — a hyperactive ko which can be played in ways placing it in multiple categories. (Definition provisional, see article!)
 Robert Pauli: Instead of talking about a two-move approach ko, which does not say who has an advantage, I prefer to talk about a ko where, say, White owes two approaches. E.g. rather than "White will have to play at b" I would say "White owes an approach at b".
Contemplating about how to convince Robert Jasiek
not to glue qualifiers to nouns (X-ko),
I developed a new terminology:
If they fight a needs-one ko she owes the move, and he runs out of ko threats while she possesses exactly one, he'd be glad to find a needs-one ko threat.
However, then I discovered that "needs-one ko" is ambiguous, it could (at least) mean
- an indirect ko that needs one move to become direct, or
- a potential ko that needs one move to become real.
 Robert Pauli: Well, you could as well say that Black has a ko threat for free . . . wrong. A free move enabled by a to-be-done approach gains points, a free threat does not, it only allows the retake.