The 3-3 Invasion 2-2 - The 3-4 Invasion

This is the third article in the series “The 3-3 Invasion,” and the second part of a 2-part article on the double hane line.

Previous article, Next article

Direction of play in the double hane line

In the previous article in this series we discussed the double hane line of the early 3-3 invasion. An important theme of the discussion there was that in the double hane line, the defender has complete freedom to decide the direction of play. Thus in the diagram above, Black has the choice to block at either A or B, essentially deciding where White’s ponnuki is placed, and can adapt this choice to the status of the game overall. White can try to resist with a hane in the opposite direction, but as we discussed in the previous article, this leads to a locally unfavorable result.

This is not specific to the double hane line: all lines of the early 3-3 invasion give the defender the freedom to choose the direction of play. The importance of choosing the correct direction of play can never be overstated. It is generally preferable to choose a locally suboptimal jōseki with the correct direction of play over a locally optimal jōseki with an incorrect direction of play. Thus it is actually quite common that the 3-3 invasion is not one of the better ways to play a corner because it cedes control over the direction of play to the opponent.

Fortunately, there is a way around this conundrum: the 3-4 invasion.

The 3-4 invasion: Basics

The idea behind the 3-4 point is to introduce a directional asymmetry into the corner which allows the invader to decide the direction of play. Here it is White that has the freedom to choose: depending on the situation, White could choose to attach at Q3 instead of R4, as we have done here.

After the attachment it is Black’s turn to choose, and the choice is between the hane on the outside (A) or on the inside (B). (A) leads to one of two outcomes: the jōseki either reverts back to the double hane line but with Black having blocked on the left meaning White gets a ponnuki on the right; or to a line of a 3-3, 4-4 approach jōseki in which Black takes the right side but White takes the corner. (B) leads to a number of similar outcomes in which Black takes the corner and bottom side, while White establishes a position on the right side.

As we can see, this jōseki (with the R4 attachment) has a bias toward letting White get a position on the right side. Therefore White can use it when White prefers to have a position on the right and is not interested in a position on the bottom. If White wants to have a position on the bottom and not on the right, White can instead attach at Q3 to swap the directions.

The essence of the jōseki is to offer Black the choice of two unpleasant options:

• give up the freedom to choose the direction of play and let White settle in the direction of White’s choosing, or;
• insist on the direction of choice, but give up a fairly sizeable corner.

Let us now see each of these lines in detail. In what follows, we will assume that White is not interested in having a position on the bottom side.

Outside hane

If Black blocks from the outside, then White extends. Then Black can choose between two options.

If Black extends to (A), then this transposes to the double hane jōseki discussed in the previous article. But this time, Black has been forced to block on the left instead of on top, and consequently White’s ponnuki goes on the right side. Since White was not interested in having a position on the bottom side, this is a successful outcome for White.

Black can still choose the direction of play here if Black connects at (B), which prevents White from getting a position on the right. Then this transposes to a 3-3, 4-4 approach jōseki with White defending. White gets the corner, while Black gets the outside and sente, though Black occasionally spends its sente to extend. In this scenario, White fails to get a position on the right side, but in return comes out with a decently sized corner. This corner is larger than White can normally expect from a 3-3 invasion, which will become evident in later articles in this series. In particular, in most lines of the 3-3 invasion where White keeps the corner, White is substantially more sealed in on the bottom side. Here, White has a stone poking out toward the bottom side, and Black cannot keep White sealed in completely. The tradeoff of a larger-than-usual corner for letting Black choose the direction of play is one that White can accept.

Inside hane

If Black blocks from the inside, then White again extends. The options (A) and (B) are fairly similar in nature. In both cases White will extend along the side. These variations are by no means exhaustive: these are simple positions and both sides have a lot of flexibility on how to conclude the sequence. For instance, in the sequence after (B) White is not specifically limited to the two-space extension: depending on the situation, wider extensions could be possible. Similarly, Black can choose to spend an extra move here to extend from its corner, but could also spend sente in other ways.

Again, the point to note here is that White ends up with groups on the right side with decent shape, which is a successful outcome given that White is not interested in a position on the bottom side. Black is not too badly off either: in both cases the corner is fairly sizeable.

The 3-4 invasion in action

Let us now see how the 3-4 invasion works in practice.

Here Black has opted for a Kobayashi-style opening. Of note is that in the top right corner, White opted to descend to R18, allowing Black to push to Q14. Overall, it is a very even game, with KataGo assigning a Black winrate of ~50%. Obviously, on a board this open, there are lots of viable ways for White to proceed. But if White were specifically to choose among the possible corner invasions (A) - (C), which should White choose?

The key to answering this question lies in the fact that Black has the push to Q14. This implies that Black values the right side and wants to develop in this direction. In this position Black must also always mind the peep at S13, which can lead to a fight for Black’s group in the top right. Black must therefore be careful to keep its top right properly supported.

With these considerations in mind, (A) turns out to be suboptimal. This is because this leaves Black free to block on the top. Black plays out the knight’s move variation of the 3-3 invasion in order to take sente, and then plays the center hoshi on the right, leaving Black with a winrate of ~55% - a 5% gain from before. All of Black’s stones on the right are synergizing extremely well. Besides just the sheer amount of territory Black is claiming, Black can play more comfortably knowing that the peep at S13 is now much less of a threat. White’s stones on the bottom, on the other hand, are not synergizing at all. They are all pressed low to the second or third line, and thus claim comparatively little territory. Moreover, the squared White stones were not under threat. Having the triangled White stones nearby is therefore inefficient, as they are reinforcing stones that do not need reinforcement.

From this we can also see that the correct direction of play for White here is the right side. Indeed, the bottom side is a no-go zone for both sides. Black cannot expect much development here due to the fairly strong White group on the bottom left, while White has no interest in reinforcing stones that need no such support. Thus the correct choice of invasion in this case is (B), which we examine now.

If Black blocks on the inside, then White extends and makes a group on the right while Black takes the corner. White can then either settle peacefully, or try to get some value out of the S13 peep, though it cannot be too aggressive here so long as its lower right stones remain somewhat unsettled. This leads to a very even result, with essentially no change in winrate or score.

If Black blocks on the outside, then the recommended result is for Black to take the outside while White takes the corner. Then Black has sente and its outside stones support the top right, alleviating the weakness at S13, but White in turn has the corner, resulting in another fairly even game, perhaps slightly favoring White.

Worth noting is when Black blocks on the outside and opts for the double hane line. This outcome is quite bad for Black, even though on the surface it seems just like the result of blocking on the inside - Black gets the corner, while White gets a group on the right. However, the difference is that when Black blocks on the inside, the resulting White group on the right is still unsettled. Therefore White cannot exploit the peep at S13 too heavily, lest its lower right stones be attacked in turn. But when Black takes the double hane line, White instead gets a ponnuki on the side - a much stronger shape - as well as sente. This makes the peep at S13 much more dangerous, as Black cannot threaten the ponnuki in any meaningful way to get sente moves to help it settle. Consequently the double hane results in a dangerous fight for Black. From here we can perhaps draw a secondary lesson in the form of a greater appreciation for the power of the ponnuki in the double hane line.

Lastly, for the sake of illustration let us examine what happens if White really mistakes the direction of play and attaches at (C) instead. KataGo rates this as the worst of the three corner invasions, and the reason is the same as before: this 3-4 attachment is biased toward White getting a group on the bottom side, but the bottom side is not the correct direction of play for White right now. The lower left White group still needs no reinforcement. Black will happily let White strengthen an already strong group while keeping the corner and indirectly supporting its top right group against the peep at S13. Note that in (A) and (B), Black could only choose one option between keeping the corner and supporting the top right.

Conclusion and takeaways

In this article we have discussed the meaning of the 3-4 invasion and how it can be used to shape the direction of play. As we have seen in the example, being able to control the direction of play is a crucial skill, and not having control of the direction of play can quickly lead to giving away significant advantages. Therefore the 3-4 invasion is a crucial weapon to have in your arsenal if you intend to play with the 3-3 invasion often. We hope that this article has not only given you a new tool to experiment with, but also a renewed or newfound appreciation for the importance of the direction of play.

Future articles in the series will be on the non-double-hane lines of the 3-3 invasion.

Return to the 4-4 Jōseki page.