Responding to the 3-3 Invasion

When playing 4-4 and your opponent responds with the 3-3 invasion, seems like there are three good ways to respond. I exclude the old single hane since consensus seems to be that’s too weak. I’m curious to hear people’s opinion on when to choose from these.

  1. Double hane. I believe this is good when the corner is more valuable than the outside, especially if the wall you would otherwise build on the outside doesn’t have supporting stones or especially if it’d be threatened by opposing stones. Con is you lose sente.
  2. Knight’s move. Made popular by AlphaGo. Builds the outside and you have the benefit of keeping sente. Especially good when the outside has support for building a base.
  3. Simple extension. Same benefits as #2, but it gives up more territory. So I’m not sure why you’d want to play this, but I see it’s popular.

Do you agree or disagree with my assessment of #1? Under what circumstances is #3 better than #2?


Some good resources in video form, Yeonwoo:

And recently Michael Redmond has been posting several videos on this:

I’ll leave the actual discussion to the stronger players :wink: .


To begin with, the differences among those four moves are smaller than you think. On a board with only the corners occupied, KataGo rates them within 1 point of one another—even the single hane. The differences matter even less the less skilled you are at Go, since unpredictability increases. So my advice is to play what you want to play. Play what you like, what makes sense to you, what you think your opponent is ill prepared for, all that jazz.

If you know all the ins and outs of those josekis, the high-level difference between #2 and #3 is what color stone occupies the opposite corner. #2 is susceptible to the flying knife joseki. One of the many perils in that joseki is this ladder. As a result, the invading player can play the flying knife only if the opponent does not occupy the far corner. And if that is the case, then the opponent should avoid it by playing B or C here, which some players consider inferior to #3.


Best thing to do (I think) is to take your hypotheses into a game and apply them. If things don’t work out like you thought, then you can modify your ideas.

1 Like