A small base is worse than no base

dfan · Post by **dfan** » Mon Feb 05, 2024 10:18 am

This is the sort of thing about which stronger players will say "That's obvious! You got to 3k without knowing that??" (either that, or they will say "That's wrong!"), but it's pretty basic and I somehow got to this point without realizing it.

We get told as early as beginner level that it's important to make a base. Probably "base" is more well defined than I've noticed (although I looked in a couple of John Fairbairn's Go Wisdom appendices and it's mentioned but not defined, so maybe the definition is too obvious to give!), but I always thought of it as "a bit of space on the side that can make an eye". My new realization is that, ceteris paribus, a single eye on the side is worse than no eyes on the side.

The first event in this journey was at a strong-kyu game review by Mateusz Surma at the US Go Congress last year, where someone's running group was making eye shape near the side, and Mateusz said (paraphrased) "if you're running to connect, just run, don't make eyes". That was new to me, but in retrospect it made total sense; you're going to end up alive anyway when you connect, so why make extra eyes you're not going to end up needing? So I filed it away as a special rule.

Then I recently watched an introductory lecture by Guo Juan about that most basic (sorry) example of a base, a two-space extension on the third line. One of her points was that if this group is hemmed in by opponent's stones on either side (say a one-space jump away), it's not only weak, but extremely inefficient. When your opponent makes all of their endgame moves, you're going to end up with a ton of your own stones surrounding only a couple of points.

Of course I am familiar with the topic of overconcentration, but to me it always meant a bunch of very strong stones enclosing only a moderate amount of territory. It didn't really occur to me that using six stones to take one point of territory (for example) is an even worse form of overconcentration!

So here's my new way of thinking. When getting a foothold on the side, unless you have the possibility of getting two eyes locally (or other factors are at play, such as a nearby weak group of your opponent, or you really think you need exactly one more eye), you're better off making sure that you are well equipped to develop into the center than making a "small" (one eye max) base. Your stick poking through your opponent's territory has already done enough damage by making your opponent place border stones on either side of it; in the case of a straight stick perpendicular to the side, you've taken away 3 points per row plus your opponent had to play twice as many stones as you locally. If you widen that stick, you're placing twice as many stones for a very small amount of extra points (a tiny bit of territory for you, a small reduction of your opponent's) without even ensuring life.

Here are a couple of examples. The first is a 3d Fox game I played recently:

Click Here To Show Diagram Code: [go]$$c $$ +---------------------------------------+ $$ | . . . . . . . . . T . T . . . . . . . | $$ | . . . . . . . . . T . T . . . O . . . | $$ | . . . O . . . W . X . a . O . . O . . | $$ | . . . , . O . . . , . . . . . X . . . | $$ | . . X . . . . . . . . . . . . . . O . | $$ | . . . . . . . . . . . . . . . . X O . | $$ | . . . . . . . . . . . . . . . . X . . | $$ | . . X . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . X . . | $$ | . . . , . . . . . , . . . . X , . . . | $$ | . . . . . . . . . . . . . . . . O . . | $$ | . . O . . . . . . . . . . O X X X . . | $$ | . X O . . . . . . . . . . . O . O X . | $$ | . X O . . . . . . . . . . O . . O X . | $$ | . . X O . . . . . . . . . . . . X . . | $$ | . . X O . . . . . , . . . . . X . . . | $$ | . . X O . . . . . O . . . X . . . . . | $$ | . X O O . . . . . . . . . . . . . . . | $$ | . X . . . . . . . . . . . . . . . . . | $$ +---------------------------------------+[/go]

As Black, I instinctively "made a base" with A. But KataGo thinks this move is pointless. White's groups on either side are strong, my single stone can easily run, so this is just points, and it's not very many points! I don't really want to play on all the triangled spots (eventually), but in that case why did I even start claiming this area? It's immensely more important to play on the lower side.

The second example comes from the game of another player who's a little stronger than me:

Click Here To Show Diagram Code: [go]$$Wc $$ +---------------------------------------+ $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . X X X . . . . . . . . . . . | $$ | . . X X . O O . . . . . . . . X . . . | $$ | . . O O O . . . . , . . . . . , . . . | $$ | . . . . . . . . . . . . . . . . X . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . 2 . 3 . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . X . a . . . . . . . . . . . . . . | $$ | . . . , . . . . . , . . . . . , O . . | $$ | . . 1 . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . O . . . . . . . . . . . . O O . . | $$ | . . . . . . . . . . . . . . . O X . . | $$ | . . . O . . . . . , . . . . . , X . . | $$ | . . . . . . . . . . . . . X . X . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ +---------------------------------------+[/go]

After the 1-2-3 sequence, he was wondering how best to move out with his Black group. But my new principles got me wondering whether

was actually a good move in the first place. White's group at the top is plenty strong (things might be a little different if there were a Black stone at E17), so the move doesn't have much effect on it, and Black's not threatening to live on the side. Better to immediately jump out with something like A. KataGo agreed with me (although it only thought the 2-3 exchange lost a point or so).

I look forward to hearing whether this is actually as good a principle as it seems to me. I used to think that a stick poking into my opponent's territory is fine, a fully living group is great, and a narrow foothold on the side is somewhere in between. But it now feels like it is the worst of both worlds, a large commitment of stones without much profit either in points or in safety.

John Fairbairn · Post by **John Fairbairn** » Mon Feb 05, 2024 11:34 am

I think you've had a genuine Eureka moment. First, I don't think stronger players should say it's obvious, because it's not (and part of the reason is that 'base' is a misleading translation). If they do, then their smug factor is far too high.

Second, you are right that it's not normally defined as a technical term.

I have always found that, for myself, the most useful, and certainly the deepest, examples of knowledge acquisition are these Eureka moments. They are not discoveries really, but rather moments of realisation of snippets of knowledge that other more experienced people take for granted - which is not the same as saying they are obvious. The experienced players just haven't shared them, and I think there are several reasons why they haven't - none of them nasty.

In my own experience, these snippets most often arise when I am playing over a game (or, more often actually transcribing a game, because that takes more concentration and thus gives greater focus. My reaction is always of the type: "Eh??!!! I didn't know you could do that!"

I had such a reaction to the topic you are adumbrating and came to similar-sounding conclusions by a completely different route. In my case, it was while going over old Chinese games. Because it was for a book, again I was concentrating more than usual. I notice that the old masters did not make bases (they made what they called extensions, but when a side group was attacked, they jumped out into the centre rather than "make a base" - just as you said Surma explained. He refers to "running away" into the centre but the Chinese masters seemed, to me, more concerned with connecting groups because of group tax. I'm sure that's true enough, but I think your point about overconcentration is a better way of putting it.

But, having noticed this, I became more aware of what we call "bases" and then started noticing that they were played mainly as boundary plays and in order to settle existing groups. That was when it dawned on me that the existing groups did have a name: "ukiishi" or floating groups. And to have that name they must have existed BEFORE the base was made. This, I think, is counterintuitive to us, because we think the base has to be built first. And from then on I was on something of a roll, and realised that the main word in the Japanese word for 'base' (根拠 or konkyo - also used in Korean; Chinese normally just uses 根) really refers to putting down roots. Konkyo is used for a basis or foundation, but the underlying feeling is rather different to our 'base'. I'm not sure how we all see that, but I think I'd associate that with stone slabs, or concrete, or flattened earth. The CJK associations are, however, more organic. And this seems confirmed in go usage because a phrase that you will see in Japanese go term texts is 'ne wo hayasu' 根を生やす　(also in Chinese as 生根) and, as a go term, it is defined as "creating eye shape by burrowing into the corner or side from a floating group" (to repeat, the base comes second).

There is a related phrase 根拠地　(common in the ordinary language but also in go) which is a place where you can make or let roots grow. I mentioned in another thread recently that I have noticed that AI plays at these points rather early (as did the old Chinese masters and Go Seigen) and so rarely ends up getting clobbered by moves such as monkey jumps.

I hope this recounting of my experience gives you confidence in your own discovery - which has given me confidence! Thank you for sharing.

Maybe as a group we could try and tighten this concept up. My contribution would be to argue for trying to find a better word for 'base' as I think it has misled all of us. I haven't got a suggestion yet, but I'll sleep on it, and look forward to further contribution from the VAST number of lurkers here!

dfan · Post by **dfan** » Mon Feb 05, 2024 12:10 pm

Thanks John, that helps to reassure me that I'm on the right track, and the vocabulary information is fascinating as always. It's interesting how something so simple could have eluded my grasp for so long. As you imply, it's easy for people to bring their own baggage to the simple English word "base", and once you have a misleading definition in your head, all your knowledge that uses the term as shorthand is shaky as well, because it's shorthand for the wrong thing. And if a stronger player is trying to explain related concepts to you, they might not even realize that you're not talking about the same thing!

It's nice to occasionally make my own connections instead of having to always be told them, and of course concepts that you figure out yourself always stick better. I feel like every lesson I've learned lately boils down to "maximize efficiency". Of course that's true - the whole point of Go is to maximize your point differential with the N moves you are given - but it's easy to forget just how fundamental it is to everything you do.

jlt · Post by **jlt** » Tue Feb 06, 2024 3:10 am

So what would be a less misleading terminology instead of "base"? "Eyeshape on the side"? Or would that still be misleading?

John Fairbairn · Post by **John Fairbairn** » Tue Feb 06, 2024 4:05 am

I slept on it and didn't come up with anything entirely convincing.

My first thought was 'platform', which to me had the nuance of correctly emphasising activity ON the platform, i.e. in the centre (base seems to emphasise active WITHIN the base). But platform still seems to exclude the edge (or corner) to some extent.

The image I eventually decided that best encompassed the necessary associations for me (though still imperfectly) was of trapeze artist working in the air with a 'safety net' below him. The point about that is that it makes you think, when making a centre-oriented group, about allowing for the konkyochi - the place where the roots can be planted an grow organically, or where a safety net can be allocated. But it's imperfect in that it may still make you think about erecting the safety net first.

Another idea, taking its cue from a Matthew Macfadyen seminar in which he described running battles as jungle warfare, was the call the konkyochi a 'base camp'. That has the merit of retaining the familiar word 'base' but without the associations of foundations, concrete or other fixed structures. It may have the disadvantage of making people think you need to build a base camp first, but a base camp can (at first) be a fairly rudimentary affair, and that could perhaps equate to an initial extension - but not all extensions are equal.

Taking a completely different view of the imagery, the 'base' area could be called the foothills of a mountain. You will certainly build round the foothills, but you obviously don't create the base rock yourself (that would equate to putative territory). At the same time, your main aspiration might be to build a monastery (or tourist cafe) on TOP of the mountain.

There may be some merit in trying to follow the old Chinese tack and not talk about bases at all, but instead try to assets the relative points on the side. They debated a lot over, say, the 8-3 points and the 6-3 points, but eventually seemed to settle on the 9-3 points (points, note, not bases) as the best ones to occupy. If you get the best point to occupy - the sweet spot - maybe considerations about bases, extensions, or access to the centre all follow in a natural way that hardly needs to be expressed. I assume modern play (minus group tax) would bring fourth-line points into the mix. In other words, when deciding which point to choose on the side, rather than thinking about side-oriented things like "I can make a two-space extension on either side of it", you think first and foremost about "what effect will this have on the more important centre? And, if that all goes pear-shaped, can I retrieve the situation by making a pear shape on the side?"

So what would be a less misleading terminology instead of "base"? "Eyeshape on the side"? Or would that still be misleading?

I have already asked the first question, and shown that I am unsure of a good answer. But the eyeshape one proposed here is something I personally also shoot down, not so much because it may be misleading in the way 'base' is, but because, like 'base' itself, it follows the old, bad, western habit of focusing on static shape, when the main focus should be on the dynamic - running into the centre as per Mateusz Surma or fighting in the jungle as per Matthew, or the more sophisticated fulcrum points of old Chinese theory.

xela · Post by **xela** » Tue Feb 06, 2024 6:55 pm

dfan wrote:We get told as early as beginner level that it's important to make a base. Probably "base" is more well defined than I've noticed (although I looked in a couple of John Fairbairn's Go Wisdom appendices and it's mentioned but not defined, so maybe the definition is too obvious to give!), but I always thought of it as "a bit of space on the side that can make an eye". My new realization is that, ceteris paribus, a single eye on the side is worse than no eyes on the side.

Funny you should mention it now. I've also been questioning this recently, in my case influenced by KataGo reviews of my games. I can't remember which beginner book drilled into me the importance of making bases. Many times I've dutifully made the two-space extension along the third line, only to find that my stones come under severe attack anyway.

My current working theory is that in a fight, your stones don't need to be alive, they just need to be no deader than the opposing stones next to them. I'm calling it "dynamic life", as opposed to the "static" life you get from making two eyes.

Here's an example from a game where I was giving two stones (and maybe shouldn't have been).

Click Here To Show Diagram Code: [go]$$Wc19 $$ +---------------------------------------+ $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . X X X X X . | $$ | . . . . O . . . O O . O O X O O X O . | $$ | . . X . . . . . . X . . X O O X O X . | $$ | . . X O . . . . . . . . X . . . O . . | $$ | . . . O . . . . . X . . . O . . O . . | $$ | . X . . . . . . . . . X . O . . . . . | $$ | . . . . . . . . . O . . . . . . . . . | $$ | . . . . . . . . . . . X . O . . . . . | $$ | . . X . . . . . . . . . . . . O O . . | $$ | . . . . . . . . . . . . . . d X . . . | $$ | . . . . . . . . . . . . . X . . X . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . X . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . O . . . | $$ | . . . X . . . . . . . . . X . . a . . | $$ | . . . . . O . . . O . . . . c X O . . | $$ | . . . . . . . . . . . . . . b . B . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ +---------------------------------------+[/go]

Black has just played the marked stone. I thought if white a next, then black b. and I don't have room for a base on the side. I might be able to scrape out a small life, or run out into the centre, but it looked pretty uncomfortable. So I tried to create a "clever" (not so clever) sabaki sequence starting with c. (Luckily black helped me out with some mistakes.)

In the review, KataGo said: hey, black has a weakness at d! So you don't need a base. Instead, you can start a fight on equal terms. White a is a fine move!

Here's one of many possible variations.

Click Here To Show Diagram Code: [go]$$Wc19 $$ +---------------------------------------+ $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . X X X X X . | $$ | . . . . O . . . O O . O O X O O X O . | $$ | . . X . . . . . . X . . X O O X O X . | $$ | . . X O . . . . . . . . X . . . O . . | $$ | . . . O . . . . . X . . . O . . O . . | $$ | . X . . . . . . . . . X . O . . . . . | $$ | . . . . . . . . . O . . . . . . . . . | $$ | . . . . . . . . . . . X . O . . . . . | $$ | . . X . . . . . . . . . . . . O O . . | $$ | . . . . . . . . . . . . . . 4 B . . . | $$ | . . . . . . . . . . . . . B . . B . . | $$ | . . . . . . . . . . . a 9 0 . 3 c . . | $$ | . . X . . . . . . . . . 8 7 . . d . . | $$ | . . . . . . . . . . . . 6 5 . O . . . | $$ | . . . X . . . . . . . b e X . . 1 . . | $$ | . . . . . O . . . O . . . . . X O . . | $$ | . . . . . . . . . . . . . . 2 . X . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ +---------------------------------------+[/go]

and next, white a, black b, white c. (Of course, KataGo says black shouldn't play

but should do something trickier, but that's another story. You can get lost in the variations. I don't think that changes the overall message.) White's looking pretty unhealthy to my eyes, but the point is that black is no better off.

Instead of

, what if black goes for the kill by playing d? Then white can cut at e. The lack of a base by white is offset by the weaknesses in black's position. It's not balanced in the sense of a sturdy table with solid legs. It's balanced in the sense of a spinning top which won't fall over as long as it keeps moving.

dfan · Post by **dfan** » Tue Feb 06, 2024 8:02 pm

xela wrote: My current working theory is that in a fight, your stones don't need to be alive, they just need to be no deader than the opposing stones next to them. I'm calling it "dynamic life", as opposed to the "static" life you get from making two eyes.

Yes! I am by nature a safety-first player, so I have to keep reminding myself of this. As the joke goes, "I don't have to outrun the bear, I just have to outrun you." You don't have to make yourself immune to any possible attack if you can make sure that an adjacent opponent group will die before you do.

Harleqin · Post by **Harleqin** » Wed Feb 07, 2024 6:48 am

Thanks to xela for the »spinning top« metaphor!

One thing that I thought should be kept in mind: there are often considerations that are orthogonal to whatever principle you are thinking about at the moment. Your first example seems like that:

dfan wrote:
$$c
$$ +---------------------------------------+
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . O . . . |
$$ | . . . O . . . W . X . a . O c . O . . |
$$ | . . . , . O . . . , . . . d . X . . . |
$$ | . . X . . . . . . . . . . b . . . O . |
$$ | . . . . . . . . . . . . . . . . X O . |
$$ | . . . . . . . . . . . . . . . . X . . |
$$ | . . X . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . X . . |
$$ | . . . , . . . . . , . . . . X , . . . |
$$ | . . . . . . . . . . . . . . . . O . . |
$$ | . . O . . . . . . . . . . O X X X . . |
$$ | . X O . . . . . . . . . . . O . O X . |
$$ | . X O . . . . . . . . . . O . . O X . |
$$ | . . X O . . . . . . . . . . . . X . . |
$$ | . . X O . . . . . , . . . . . X . . . |
$$ | . . X O . . . . . O . . . X . . . . . |
$$ | . X O O . . . . . . . . . . . . . . . |
$$ | . X . . . . . . . . . . . . . . . . . |
$$ +---------------------------------------+
Click Here To Show Diagram Code
[go]$$c $$ +---------------------------------------+ $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . O . . . | $$ | . . . O . . . W . X . a . O c . O . . | $$ | . . . , . O . . . , . . . d . X . . . | $$ | . . X . . . . . . . . . . b . . . O . | $$ | . . . . . . . . . . . . . . . . X O . | $$ | . . . . . . . . . . . . . . . . X . . | $$ | . . X . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . X . . | $$ | . . . , . . . . . , . . . . X , . . . | $$ | . . . . . . . . . . . . . . . . O . . | $$ | . . O . . . . . . . . . . O X X X . . | $$ | . X O . . . . . . . . . . . O . O X . | $$ | . X O . . . . . . . . . . O . . O X . | $$ | . . X O . . . . . . . . . . . . X . . | $$ | . . X O . . . . . , . . . . . X . . . | $$ | . . X O . . . . . O . . . X . . . . . | $$ | . X O O . . . . . . . . . . . . . . . | $$ | . X . . . . . . . . . . . . . . . . . | $$ +---------------------------------------+[/go]

I think that 'a' is just a bad move because it lets White jump to 'b', where it erases all potential Black had there while still keeping the black group weakish and separated, and even reaching out a bit to the centre where it might help O8, should that become separated from the bottom. I think that locally (yes, the bottom might be more urgent globally) the move is to enclose with 'c', 'd' etc.

Kirby · Post by **Kirby** » Wed Feb 07, 2024 8:27 am

At the end of the day, you win the game if you have more points. A caveat to this is that, if your group is weak, you cannot efficiently make points - your opponent can attack and get better exchanges. Note that you can also get points in the center. So if you aren't making a lot of points on the side, playing toward the center may very well be getting more points.

Click Here To Show Diagram Code: [go]$$c $$ +---------------------------------------+ $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . O . . . | $$ | . . . O . . . O . B . . . O . . O . . | $$ | . . . , . O . . . , . . . . . X . . . | $$ | . . X . . . . . . . . . . . . . . O . | $$ | . . . . . . . . . . . . . . . . X O . | $$ | . . . . . . . . . . . . . . . . X . . | $$ | . . X . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . X . . | $$ | . . . , . . . . . , . . . . X , . . . | $$ | . . . . . . . . . . . . . . . . O . . | $$ | . . O . . . . . . . . . . O X X X . . | $$ | . X O . . . . . . . . . . . O . O X . | $$ | . X O . . . . . . . . . . O . . O X . | $$ | . . X O . . . . . . . . . . . . X . . | $$ | . . X O . . . . . , . . . . . X . . . | $$ | . . X O . . . . . O . . . X . . . . . | $$ | . X O O . . . . . . . . . . . . . . . | $$ | . X . . . . . . . . . . . . . . . . . | $$ +---------------------------------------+[/go]

Above, the marked black stone is weak. If black doesn't do anything, white can get lots of points by attacking it.

If black jumps to the side, the group is still weak, and can still be easily attacked later. For example, even if white just splits black, the black stones are still weak:

Click Here To Show Diagram Code: [go]$$c $$ +---------------------------------------+ $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . O . . . | $$ | . . . O . . . O . B . B . O . . O . . | $$ | . . . , . O . . . , . . . . . X . . . | $$ | . . X . . . . . . . . . . W . . . O . | $$ | . . . . . . . . . . . . . . . . X O . | $$ | . . . . . . . . . . . . . . . . X . . | $$ | . . X . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . X . . | $$ | . . . , . . . . . , . . . . X , . . . | $$ | . . . . . . . . . . . . . . . . O . . | $$ | . . O . . . . . . . . . . O X X X . . | $$ | . X O . . . . . . . . . . . O . O X . | $$ | . X O . . . . . . . . . . O . . O X . | $$ | . . X O . . . . . . . . . . . . X . . | $$ | . . X O . . . . . , . . . . . X . . . | $$ | . . X O . . . . . O . . . X . . . . . | $$ | . X O O . . . . . . . . . . . . . . . | $$ | . X . . . . . . . . . . . . . . . . . | $$ +---------------------------------------+[/go]

However black jumps out, white will be more efficient - black is just jumping out without getting points, in order to save his group. White continues to get points in the center by attacking.

Stepping backwards, if black just jumps out:

Click Here To Show Diagram Code: [go]$$c $$ +---------------------------------------+ $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . O . . . | $$ | . . . O . . . O . X . . . O . . O . . | $$ | . . . , . O . . . , . . . . . X . . . | $$ | . . X . . . . a . B . c . . . . . O . | $$ | . . . . . . . . . . . . . . . . X O . | $$ | . . . . . . . . . b . . . . . . X . . | $$ | . . X . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . X . . | $$ | . . . , . . . . . , . . . . X , . . . | $$ | . . . . . . . . . . . . . . . . O . . | $$ | . . O . . . . . . . . . . O X X X . . | $$ | . X O . . . . . . . . . . . O . O X . | $$ | . X O . . . . . . . . . . O . . O X . | $$ | . . X O . . . . . . . . . . . . X . . | $$ | . . X O . . . . . , . . . . . X . . . | $$ | . . X O . . . . . O . . . X . . . . . | $$ | . X O O . . . . . . . . . . . . . . . | $$ | . X . . . . . . . . . . . . . . . . . | $$ +---------------------------------------+[/go]

There are many directions for black to escape to after this. So it's hard (for me) to think of an effective attack on black. For example, even if I play a stone in the a, b, or c area, black may not even answer. There are so many areas for black to move in the center.

I'll probably tenuki now and see if there's a better time later to revisit attacking. So black effectively avoided a strong attack with just one move. Extending to the side induces white to split black, which is something white wanted to do anyway.

---

But, this is not a DEFINITE rule. Sometimes jumping out may not have a lot of benefit.

Let's say the board is like this:

Click Here To Show Diagram Code: [go]$$c $$ +---------------------------------------+ $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . O . . . | $$ | . . . O . . . O . X . . . O . . O . . | $$ | . . . , . O . . . , . . . . . X . . . | $$ | . . X . . . W . . . . . . W . . . O . | $$ | . . . . . . . . . . . . . . . . X O . | $$ | . . . . . . W . W . W . W . . . X . . | $$ | . . X . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . X . . |[/go]

In this case, black doesn't really avoid an attack by jumping out. Just jumping almost wastes a move. Since escaping doesn't help here, making a base is probably better - it's a tough situation for black since there are many white stones in the area, but trying to use forcing moves to make a base could be effective. I'm not sure if it's definitely possible to live in an area like in this example, but supposing it's necessary, this may be a case where trying to get two eyes by whatever means possible may be pragmatic.

Maybe I'd start by attaching.

Click Here To Show Diagram Code: [go]$$c $$ +---------------------------------------+ $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . O . . . | $$ | . . . O . . . O . X . . B O . . O . . | $$ | . . . , . O . . . , . . . . . X . . . | $$ | . . X . . . W . . . . . . W . . . O . | $$ | . . . . . . . . . . . . . . . . X O . | $$ | . . . . . . W . W . W . W . . . X . . | $$ | . . X . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . X . . |[/go]

Seems tough.

But anyway, in this kind of scenario, jumping out may not be helpful even though the top part of the board is the same. The example may not be the best if black ends up dying anyway, but my point is, sometimes living inside is still a better option than jumping, when life is necessary.

Kirby · Post by **Kirby** » Wed Feb 07, 2024 8:34 am

By the way... I have been playing a little bit more, lately, and I don't think it's super useful to make a lot of heuristics. I mainly keep in mind: I need more points. If a group is too weak, the opponent can get forcing moves, which may lead to more points/efficiency.

Outside of that, I'm not sure how much is that helpful.

I was playing a game the other day, and I made the silliest mistake because of a heuristic. When I reviewed with katago, the win % was above 99% in my favor, and I was leading by over 10 points. The middle game had nearly finished, but my opponent made an attack on my group. I had a lot of space, so it was not something hard to handle - just live, and I should win the game. But instead of jumping out, I read that I could kill my opponent's group in the corner (thereby living), and I went for it. I read to a certain sequence and was like, "yeah, that's 'rabbity six'", and I decided to go for it.

But it wasn't a "rabbity six" shape - there were seven spaces, not six there

If I had just read it more clearly rather than remembering that rabbit heuristic, I would have just jumped out and finished the game.

Some heuristics are happen naturally, I guess. You recognize certain shapes (or fail to properly recognize in my case here). But I think that flexible thinking with the primary goal of making more points is ideal when possible.

Gérard TAILLE · Post by **Gérard TAILLE** » Wed Feb 07, 2024 9:07 am

Click Here To Show Diagram Code: [go]$$c $$ +---------------------------------------+ $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . O . . . | $$ | . . . O . . . W . X . a b O . . O . . | $$ | . . . , . O . . . , . . . . . X . . . | $$ | . . X . . . . . . . . . . . . . . O . | $$ | . . . . . . . . . . . . . . . . X O . | $$ | . . . . . . . . . . . . . . . . X . . | $$ | . . X . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . X . . | $$ | . . . , . . . . . , . . . . X , . . . | $$ | . . . . . . . . . . . . . . . . O . . | $$ | . . O . . . . . . . . . . O X X X . . | $$ | . X O . . . . . . . . . . . O . O X . | $$ | . X O . . . . . . . . . . O . . O X . | $$ | . . X O . . . . . . . . . . . . X . . | $$ | . . X O . . . . . , . . . . . X . . . | $$ | . . X O . . . . . O . . . X . . . . . | $$ | . X O O . . . . . . . . . . . . . . . | $$ | . X . . . . . . . . . . . . . . . . . | $$ +---------------------------------------+[/go]

I this position I understand that looking at a base by playing at a or b could be bad move.

But what is you view if the black stone is stronger like in the following poisition:

Click Here To Show Diagram Code: [go]$$c $$ +---------------------------------------+ $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . O X X . . . . . . O . . . | $$ | . . . O . . O O . X . a b O . . O . . | $$ | . . . , . O . . . , . . . . . X . . . | $$ | . . X . . . . . . . . . . . . . . O . | $$ | . . . . . . . . . . . . . . . . X O . | $$ | . . . . . . . . . . . . . . . . X . . | $$ | . . X . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . X . . | $$ | . . . , . . . . . , . . . . X , . . . | $$ | . . . . . . . . . . . . . . . . O . . | $$ | . . O . . . . . . . . . . O X X X . . | $$ | . X O . . . . . . . . . . . O . O X . | $$ | . X O . . . . . . . . . . O . . O X . | $$ | . . X O . . . . . . . . . . . . X . . | $$ | . . X O . . . . . , . . . . . X . . . | $$ | . . X O . . . . . O . . . X . . . . . | $$ | . X O O . . . . . . . . . . . . . . . | $$ | . X . . . . . . . . . . . . . . . . . | $$ +---------------------------------------+[/go]

dust · Post by **dust** » Wed Feb 07, 2024 9:13 am

dfan wrote: [snip]

I look forward to hearing whether this is actually as good a principle as it seems to me.

I'm not sure about the certainty of the principle that "a small base is worse than no base".

I wonder if the main lesson is really "If you want/need to make a base, make the best base you can. If you want/need to escape to the centre, make a good move that does that (and even better if the move does more than just escaping). If none of those work well, then consider tenuki-ing or sacrificing." Whether a base or escaping move etc is best will depend on context (as Kirby demonstrates).

Kirby · Post by **Kirby** » Wed Feb 07, 2024 9:50 am

Gérard TAILLE wrote: But what is you view if the black stone is stronger like in the following poisition:
$$c
$$ +---------------------------------------+
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . O X X . . . . . . O . . . |
$$ | . . . O . . O O . X . a b O . . O . . |
$$ | . . . , . O . . . , . . . . . X . . . |
$$ | . . X . . . . . . . . . . . . . . O . |
$$ | . . . . . . . . . . . . . . . . X O . |
$$ | . . . . . . . . . . . . . . . . X . . |
$$ | . . X . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . X . . |
$$ | . . . , . . . . . , . . . . X , . . . |
$$ | . . . . . . . . . . . . . . . . O . . |
$$ | . . O . . . . . . . . . . O X X X . . |
$$ | . X O . . . . . . . . . . . O . O X . |
$$ | . X O . . . . . . . . . . O . . O X . |
$$ | . . X O . . . . . . . . . . . . X . . |
$$ | . . X O . . . . . , . . . . . X . . . |
$$ | . . X O . . . . . O . . . X . . . . . |
$$ | . X O O . . . . . . . . . . . . . . . |
$$ | . X . . . . . . . . . . . . . . . . . |
$$ +---------------------------------------+
Click Here To Show Diagram Code
[go]$$c $$ +---------------------------------------+ $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . O X X . . . . . . O . . . | $$ | . . . O . . O O . X . a b O . . O . . | $$ | . . . , . O . . . , . . . . . X . . . | $$ | . . X . . . . . . . . . . . . . . O . | $$ | . . . . . . . . . . . . . . . . X O . | $$ | . . . . . . . . . . . . . . . . X . . | $$ | . . X . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . X . . | $$ | . . . , . . . . . , . . . . X , . . . | $$ | . . . . . . . . . . . . . . . . O . . | $$ | . . O . . . . . . . . . . O X X X . . | $$ | . X O . . . . . . . . . . . O . O X . | $$ | . X O . . . . . . . . . . O . . O X . | $$ | . . X O . . . . . . . . . . . . X . . | $$ | . . X O . . . . . , . . . . . X . . . | $$ | . . X O . . . . . O . . . X . . . . . | $$ | . X O O . . . . . . . . . . . . . . . | $$ | . X . . . . . . . . . . . . . . . . . | $$ +---------------------------------------+[/go]

IF we assume that this sequence is white's best play, locally (I'm not 100% sure):

Click Here To Show Diagram Code: [go]$$c $$ +---------------------------------------+ $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . 2 1 3 . . . . . . O . . . | $$ | . . . O . . 4 O . X . a b O . . O . . | $$ | . . . , . O . . . , . . . . . X . . . | $$ | . . X . . . . . . . . . . . . . . O . | $$ | . . . . . . . . . . . . . . . . X O . | $$ | . . . . . . . . . . . . . . . . X . . | $$ | . . X . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . X . . | $$ | . . . , . . . . . , . . . . X , . . . | $$ | . . . . . . . . . . . . . . . . O . . | $$ | . . O . . . . . . . . . . O X X X . . | $$ | . X O . . . . . . . . . . . O . O X . | $$ | . X O . . . . . . . . . . O . . O X . | $$ | . . X O . . . . . . . . . . . . X . . | $$ | . . X O . . . . . , . . . . . X . . . | $$ | . . X O . . . . . O . . . X . . . . . | $$ | . X O O . . . . . . . . . . . . . . . | $$ | . X . . . . . . . . . . . . . . . . . | $$ +---------------------------------------+[/go]

then it's the same situation.

That is, if this is bad:

Click Here To Show Diagram Code: [go]$$c $$ +---------------------------------------+ $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . O . . . | $$ | . . . O . . . O . X . 1 . O . . O . . | $$ | . . . , . O . . . , . . . . . X . . . | $$ | . . X . . . . . . . . . . 2 . . . O . | $$ | . . . . . . . . . . . . . . . . X O . | $$ | . . . . . . . . . . . . . . . . X . . | $$ | . . X . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . X . . | $$ | . . . , . . . . . , . . . . X , . . . | $$ | . . . . . . . . . . . . . . . . O . . | $$ | . . O . . . . . . . . . . O X X X . . | $$ | . X O . . . . . . . . . . . O . O X . | $$ | . X O . . . . . . . . . . O . . O X . | $$ | . . X O . . . . . . . . . . . . X . . | $$ | . . X O . . . . . , . . . . . X . . . | $$ | . . X O . . . . . O . . . X . . . . . | $$ | . X O O . . . . . . . . . . . . . . . | $$ | . X . . . . . . . . . . . . . . . . . | $$ +---------------------------------------+[/go]

then this is also bad:

Click Here To Show Diagram Code: [go]$$c $$ +---------------------------------------+ $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . O X X . . . . . . O . . . | $$ | . . . O . . O O . X . 1 . O . . O . . | $$ | . . . , . O . . . , . . . . . X . . . | $$ | . . X . . . . . . . . . . 2 . . . O . | $$ | . . . . . . . . . . . . . . . . X O . | $$ | . . . . . . . . . . . . . . . . X . . | $$ | . . X . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . X . . | $$ | . . . , . . . . . , . . . . X , . . . | $$ | . . . . . . . . . . . . . . . . O . . | $$ | . . O . . . . . . . . . . O X X X . . | $$ | . X O . . . . . . . . . . . O . O X . | $$ | . X O . . . . . . . . . . O . . O X . | $$ | . . X O . . . . . . . . . . . . X . . | $$ | . . X O . . . . . , . . . . . X . . . | $$ | . . X O . . . . . O . . . X . . . . . | $$ | . X O O . . . . . . . . . . . . . . . | $$ | . X . . . . . . . . . . . . . . . . . | $$ +---------------------------------------+[/go]

That's because, after the first diagram, you can always play out that local sequence:

Click Here To Show Diagram Code: [go]$$c $$ +---------------------------------------+ $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . 4 3 5 . . . . . . O . . . | $$ | . . . O . . 6 O . X . 1 . O . . O . . | $$ | . . . , . O . . . , . . . . . X . . . | $$ | . . X . . . . . . . . . . 2 . . . O . | $$ | . . . . . . . . . . . . . . . . X O . | $$ | . . . . . . . . . . . . . . . . X . . | $$ | . . X . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . X . . | $$ | . . . , . . . . . , . . . . X , . . . | $$ | . . . . . . . . . . . . . . . . O . . | $$ | . . O . . . . . . . . . . O X X X . . | $$ | . X O . . . . . . . . . . . O . O X . | $$ | . X O . . . . . . . . . . O . . O X . | $$ | . . X O . . . . . . . . . . . . X . . | $$ | . . X O . . . . . , . . . . . X . . . | $$ | . . X O . . . . . O . . . X . . . . . | $$ | . X O O . . . . . . . . . . . . . . . | $$ | . X . . . . . . . . . . . . . . . . . | $$ +---------------------------------------+[/go]

So there's not a way to make it better just because of the initial exchanges with the group on the top left (it's just a different order).

That being said, it's not 100% clear that white's response to the black attachment and follow up is optimal. So there may be some possibility of tenuki...

My feeling is still toward jumping. I like jumping because it is hard for me to find a good way to follow up and attack as white. I'd regret the exchanges if they were played first, though.

jlt · Post by **jlt** » Wed Feb 07, 2024 10:02 am

I picked a random pro game.

Click Here To Show Diagram Code: [go]$$W $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . , . . . . . , . . . . . , . . . | $$ | . . . . . . . . . . . . . . . X . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . O . . . . . . . . . . . . X . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . O . . . . . X . . . . O O X . . | $$ | . . . . . X . . . . . 1 . . O X X . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ +---------------------------------------+[/go]

Click Here To Show Diagram Code: [go]$$W $$ +---------------------------------------+ $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . O . . . . . . . . . . . . . | $$ | . . . O . . . . . , . . . . . X X . . | $$ | . . . . . . . . . . . . . . . . . O . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . 1 . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . , . . . . . , . . . . . , . . . | $$ | . . . . . . . . . . . . . X . X . . . | $$ | . . . . . . . . . . . . . . . . . . . |[/go]

Click Here To Show Diagram Code: [go]$$W $$ +---------------------------------------+ $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . X . . | $$ | . . . . . O . . . . . . . . . . . O . | $$ | . . . O . . . . . , . . . . . X X X O | $$ | . . . . . . . . . . . . . . . . . O . | $$ | . . . . . . . . . . . . . . . . 1 . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . O . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . , . . . . . , . . . . . , X . . | $$ | . . . . . . . . . . . . . X . X . . . | $$ | . . . . . . . . . . . . . . . . . . . |[/go]

Are these small bases?

Gérard TAILLE · Post by **Gérard TAILLE** » Wed Feb 07, 2024 11:50 am

Kirby wrote: Above, the marked black stone is weak. If black doesn't do anything, white can get lots of points by attacking it.

If black jumps to the side, the group is still weak, and can still be easily attacked later. For example, even if white just splits black, the black stones are still weak:

$$c
$$ +---------------------------------------+
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . O . . . |
$$ | . . . O . . . O . B . B . O . . O . . |
$$ | . . . , . O . . . , . . . . . X . . . |
$$ | . . X . . . . . . . . . . W . . . O . |
$$ | . . . . . . . . . . . . . . . . X O . |
$$ | . . . . . . . . . . . . . . . . X . . |
$$ | . . X . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . X . . |
$$ | . . . , . . . . . , . . . . X , . . . |
$$ | . . . . . . . . . . . . . . . . O . . |
$$ | . . O . . . . . . . . . . O X X X . . |
$$ | . X O . . . . . . . . . . . O . O X . |
$$ | . X O . . . . . . . . . . O . . O X . |
$$ | . . X O . . . . . . . . . . . . X . . |
$$ | . . X O . . . . . , . . . . . X . . . |
$$ | . . X O . . . . . O . . . X . . . . . |
$$ | . X O O . . . . . . . . . . . . . . . |
$$ | . X . . . . . . . . . . . . . . . . . |
$$ +---------------------------------------+
Click Here To Show Diagram Code
[go]$$c $$ +---------------------------------------+ $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . O . . . | $$ | . . . O . . . O . B . B . O . . O . . | $$ | . . . , . O . . . , . . . . . X . . . | $$ | . . X . . . . . . . . . . W . . . O . | $$ | . . . . . . . . . . . . . . . . X O . | $$ | . . . . . . . . . . . . . . . . X . . | $$ | . . X . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . X . . | $$ | . . . , . . . . . , . . . . X , . . . | $$ | . . . . . . . . . . . . . . . . O . . | $$ | . . O . . . . . . . . . . O X X X . . | $$ | . X O . . . . . . . . . . . O . O X . | $$ | . X O . . . . . . . . . . O . . O X . | $$ | . . X O . . . . . . . . . . . . X . . | $$ | . . X O . . . . . , . . . . . X . . . | $$ | . . X O . . . . . O . . . X . . . . . | $$ | . X O O . . . . . . . . . . . . . . . | $$ | . X . . . . . . . . . . . . . . . . . | $$ +---------------------------------------+[/go]

Click Here To Show Diagram Code: [go]$$c $$ +---------------------------------------+ $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . a . O . . . | $$ | . . . O . . . O . X . B . O . . O . . | $$ | . . . , . O . . . , . . . . . X . . . | $$ | . . X . . . . . . . . . . W . . . O . | $$ | . . . . . . . . . . . . . . . . X O . | $$ | . . . . . . . . . . . . . . . . X . . | $$ | . . X . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . X . . | $$ | . . . , . . . . . , . . . . X , . . . | $$ | . . . . . . . . . . . . . . . . O . . | $$ | . . O . . . . . . . . . . O X X X . . | $$ | . X O . . . . . . . . . . . O . O X . | $$ | . X O . . . . . . . . . . O . . O X . | $$ | . . X O . . . . . . . . . . . . X . . | $$ | . . X O . . . . . , . . . . . X . . . | $$ | . . X O . . . . . O . . . X . . . . . | $$ | . X O O . . . . . . . . . . . . . . . | $$ | . X . . . . . . . . . . . . . . . . . | $$ +---------------------------------------+[/go]

I do not understand why the exchange of the two marked stones is good white. After this exchange the black group has been really reinforced. It is now difficult to attack these stones because black can also play at a.

Click Here To Show Diagram Code: [go]$$c $$ +---------------------------------------+ $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . a . O . . . | $$ | . . . O . . . O . Y . Y . O . . O . . | $$ | . . . , . O . . . , . . . . . X . . . | $$ | . . X . . . . . . . . . . O . . . O . | $$ | . . . . . . . . . . . . . . . . B O . | $$ | . . . . . . . . . . . . . . . . B . . | $$ | . . X . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . B . . | $$ | . . . , . . . . . , . . . . B , . . . | $$ | . . . . . . . . . . . . . . . . O . . | $$ | . . O . . . . . . . . . . O B B B . . | $$ | . X O . . . . . . . . . . . O . O X . | $$ | . X O . . . . . . . . . . O . . O X . | $$ | . . X O . . . . . . . . . . . . X . . | $$ | . . X O . . . . . , . . . . . X . . . | $$ | . . X O . . . . . O . . . X . . . . . | $$ | . X O O . . . . . . . . . . . . . . . | $$ | . X . . . . . . . . . . . . . . . . . | $$ +---------------------------------------+[/go]

With a strategic point of view is it a bad idea to use the strong black stones marked with a circled to create a strong group based on the stone marked with a triangle and considering the area between the triangle group and the circle group is dame for white.

BTW in the original position I think the lower side is more important than the upper side but it is another point.

Life In 19x19

A small base is worse than no base

A small base is worse than no base

Re: A small base is worse than no base

Re: A small base is worse than no base

Re: A small base is worse than no base

Re: A small base is worse than no base

Re: A small base is worse than no base

Re: A small base is worse than no base

Re: A small base is worse than no base

Re: A small base is worse than no base

Re: A small base is worse than no base

Re: A small base is worse than no base

Re: A small base is worse than no base

Re: A small base is worse than no base

Re: A small base is worse than no base

Re: A small base is worse than no base