Shikiri - the concept of fall back scenarios

[JohnnieDarko]

Today I visited Matsuyama Castle in Matsuyama Japan. It was gorgeous and it has many parts accessible, and there was a great exhibition inside.

The defensive design is interesting. The core castle tower is defended by its design of multiple “fall-back” options.

At first, attacking troops march to the lowest gates, and from peepholes they get shot by the defending troops inside the castle.

If the attacker somehow breaches the gates, then defending soldiers will fall back. The attackers now have to walk another walkway to the next gates, but on that walk, they will again get shot from peepholes along that walkway, from inside the castle.

Repeat this classic tower defense idea a couple of times, making this a very well defended castle. This idea was called Shikiri, which literally means compartmentalising / divide into rooms.

This got me thinking about Go (…of course…). In Go you don’t build a defensive castle, though you might enclose an area in a way that you don’t want it invaded.

Is there a similar concept in Go, or has there been in historic games, where players design a clear “fall back” scenario? Where you aim for goal A, but play in a way that if that fails, that due to a safety net, a clear goal B is nearly certain to happen?

Vaguely this applies for every situation, for example I might want to kill invading stones, but if my kill fails, at least the opponent is weakly on the run. But that is not explicitly designed. I’m wondering if deliberately designed fall back scenarios are or were a part of Go strategy. Many thanks for entertaining this vague question.

[Uberdude]

If you talk about “fall back” in go I think of the situation where you have an emergency back-up plan like a way to connect a group in danger to safety, but in a submissive and bad for points way. So you don’t want to use it and would rather fight and be more aggressive and attack them or live independently, but you can do that safe in the knowledge that if you get in trouble, you can always come back and play that pathetic gote 1st line connection.

[Sadaharu]

@JohnnieDarko

Is there a similar concept in Go, or has there been in historic games, where players design a clear “fall back” scenario? Where you aim for goal A, but play in a way that if that fails, that due to a safety net, a clear goal B is nearly certain to happen?

I think all ideal attacks are that way?
You don’t see a lot of top pros going all-in during an attack.
Instead, they attack such that even if they can’t kill the dragon, they can still gain something along the way and continue the game.

[Groin]

Two cases which contains the idea of shikiri:

position which use the miai concept.

Play on the 4th line: if your opponent invades then you get influence in return.

[JohnnieDarko]

Thanks for the responses, it helps me to sharpen my vague idea. Below I’ll respond to each why I think it doesn’t quite fit my idea. Please don’t see that as criticism but as a discourse to sharpen the idea, or perhaps prove that the idea simply doesn’t fit Go.

@Uberdude

If you talk about “fall back” in go I think of the situation where you have an emergency back-up plan like a way to connect a group in danger to safety, but in a submissive and bad for points way.

This is physically the concept of shikiri (falling back to a lower line or different compartment), but not conceptually, because it incurs heavy loss.

On the contrary, a strategic fall back in a castle is used to move troops back before they are overrun. There they try again, at close to full strength, while the opponent keeps incurring losses.

This isn’t what I originally asked. So you do answer my question, but it made me realise my actual idea/question is a little different from what I asked.

@Sadaharu

Instead, they attack such that even if they can’t kill the dragon, they can still gain something along the way and continue the game.

This is conceptually it, use of an alternative fall back strategy that leads to a desired outcome, but doesn’t physically use the idea of compartments.

It might though, a pro might have certain “sector lines” (as described in EZ-Go by Wilcox) as compartments in mind, as they chase a dragon.

@Groin

Two cases which contains the idea of shikiri:

position which use the miai concept.

Play on the 4th line: if your opponent invades then you get influence in return.

Both are great examples, it’s both a physical concept and conceptually and equal alternative. However, it is also the trivial answer as all of Go is kind of like: you do A so I do B, or vice versa, and both outcomes are equal (Assuming perfect play).

—-

In summary: I am looking for something that’s recognisable on the board as a compartmentalised fall back, as if looking at a map or a floor plan, yet the fall back leading to a close to equal outcome.

Considering that stones don’t move in Go, yet the whole concept of falling back is about stones (troops) moving, it might be far fetched to find a well fitting game situation.

[Counting_Zenist]

Go was really like planning for cities and towns (city building), or alliance forming that neutral territory can switch side if they got “cut-off” from trade. Military troop analogy is really not that fitting.

[Groin]

@JohnnieDarko

However, it is also the trivial answer as all of Go is kind of like: you do A so I do B, or vice versa, and both outcomes are equal (Assuming perfect play).

I don’t think so. There are sometimes no choices at all. Or more as 2 choices. The concept of miai has its usefulness.

[shamisen]

@JohnnieDarko

In summary: I am looking for something that’s recognisable on the board as a compartmentalised fall back, as if looking at a map or a floor plan, yet the fall back leading to a close to equal outcome.

Considering that stones don’t move in Go, yet the whole concept of falling back is about stones (troops) moving, it might be far fetched to find a well fitting game situation.

My tengen style is probably exactly what you’re looking for as “shikiri” or as close to it as you’ll get in the game of Go while playing optimally.

example from a recent game: shamisen vs. GoGameSeries

Act 1: The Castle

From the first few moves, I’ve established a base in the center that I can fall back to from ANY fight on the board at any point in time, a room in the lower right corner, and 2 wings created by my two corner approaches.

Act 2: First Invasion

White invades my castle at the 4-4 point in the bottom right and succeeds in gaining some footing of his own. AI suggest that I can harass the group a bit more, but I think it’s not possible to kill White completely here so I’m going to fortify my left wing first and come back later.

Act 3: Outwards Expansion

After a failed invasion where I died for no reason in the upper left corner, I continue using my strength in the lower left to expand on the left side (which is like the castle’s courtyard?), which I’m starting to win over as well. Notice that my central base is still looming over every battlefield, making it unlikely for White to escape with both of his two stones on the left side.

Act 4: Courtyard Concession

I play a bit too passively/slowly in this area, so White seizes the chance to jump in with a large knight move to offer a rescuing hand to his two stranded soldiers on the left side.

I fall back and allow White to connect and rescue both stones in exchange for the 4 stones near the lower left and happily expand my left wing – it now has a strong influence. White could have saved the 4 stones if he gave up on rescuing one of the stones on the left, but he wanted both.

Final Act: No Escape

Finally, we return to where the first invasion happened. Black is behind in estimated territory, but the Black castle has been fully enclosed, with all the escape routes and entrances sealed off except for cracks between the 3 stones in the center. If White is not careful, something bad can happen inside. If I reclaim my castle, I win.

White makes the mistake of not staying connected and allows Black to divide the battle into two halves, with each half being surrounded on all sides by Black’s fortified castle borders – impossible to survive both assaults. Final result: White surrenders

[Counting_Zenist]

What if we expand the center compartment to the place where the Shite fuseki stone is located (one space away from the center), and change the sequence a bit

Even if the opponent played all star-points at move 6, the “framework” for comportments is still there

And it looks surprisingly natural as a framework somehow.

[shamisen]

It’s okay if the opponent doesn’t separate the formation on move 4 or 6. When I play moves 1 and 3 in a large-knight formation, there’s less space. Even if they jump in, it’s more controlled and something you can prepare for, something like this:

@shamisen

You’re right, this looks much better for White:

but I can see White has a small weakness here too:

and this:

@shamisen

Okay, I’ve come up with countermeasures.

Method 1: Change target from 3x3 symmetrical center square to 4x4 asymmetrical square:

Method 2: If they’re foolish enough to challenge even the 2nd square, it’s their loss, White will have a superior position:

[Counting_Zenist]

The more solid shape act more like static defense and blockade, like in your game, the box never got finished, but just “one side” act like a separator for isolating the opponent’s groups. And if your opponent is better at directional play, they could very well just run pass the sides instead of running head in.

My concept is more like laying traps, or building “watch towers” in strategic locations, which can be quickly fortified or redirected on all directions when needed and on a larger scale, have more strategic options.

[shamisen]

Yes, it’s a similar, related strategy with different tactics. One difference is that you would have to watch your watch towers more carefully to use them as ladder breakers – I’m lazy on reading ladders so I prefer to put 3 stones in the middle so every ladder automatically works

I have tried your idea before, I think: 友誼對局

The game of Go is very free if you’re willing to step away from AI josekis and lose a few points in the opening, and I began to enjoy it 2x more after learning about these strategies. Before this, I only played blitz Go because it was a fun and quick game I could play with one hand while eating food.