Number of stones

[John Fairbairn]

I once bought a Japanese go book in LA simply because it looked so utterly different from anything else I'd ever seen. Also, it was written by the Igo Club team, which is normally a sign of very high quality. But it infuriated me when reading it on the flight home. It was so, so basic (very like RJ's book on fundamentals) that it had nothing for me, I concluded. It wasn't that which infuriated me directly, but rather that, as it was an import, it cost me about three times what it would have cost in Japan. It was a sheer waste of $30.

But I accidentally knocked it off a shelf today when looking for something to read while waiting somewhere, and I thought I'd give it another, fairer go. I confess it still wasn't my cup of tea, but a few intriguing things did emerge, which I pass on for consideration.

1. The bit on the true fundamentals was very well done. This ranged from things like the characteristics of single stones, types of connection, shimaris, extensions, what is territory, standard shapes, etc. All truly basic, but I did start to wonder whether these things are truly internalised in stronger amateurs. Specifically, I can't really remember being aware of the number of liberties in a contact fight, which is a pretty important consideration. I think my subconscious takes care of this (on the basis that I rarely get caught out in a liberty trap), but I'm not at all sure that my subconscious kicks in with this detail as early as it should (on the basis that "rarely" is not the same as "never" and, in particular, I'm sometimes prone to overlook liberty shortage in tsumego). Does this ring bells with other people? If so, how do we move the liberty alert higher up the scale in our subconscious - simply more tsumego?

2. There was an interesting section on the difference between go terms that relate to stones of the same colour and terms that are used when opposing stones enter the frame. I won't say more here but I think it's a point worth mulling over the next time you have a boring bus ride.

3. The most interesting point was a section on "the number of stones". I hadn't latched on to it before, maybe because it sounds boring or maybe because it had the word "number" in it, but the basic premise is that when you decide to play in an area your choice depends on how many stones you already have there. Obvious enough, but less obvious is what the book calls the "latent number of stones" in a position. If you have a forcing move but have not yet played it, the Japanese term kiki is used for that. Once you have played it is a kikashi (i.e. a kikashi is not really a forcing move but a having-forced move). Anyway, the number of kiki you have should be added to your count as latent stones. Furthermore, positions where you can count kiki are much more numerous than you might expect at first sight . None of this felt new to me, but again I had another feeling that I maybe hadn't internalised it deeply. Should I have? The fact this topic was deemed fit for a very basic book suggests yes (and I do trust the Igo Club team). It also occurs to me that some western kyu players may not have internalised this at all, simply because kiki is not normally part of their vocabulary. Yay or nay on these various points?

[Joaz Banbeck]

I'm a bit hesitant to count kiki as extra stones when my opponent also gets to play one stone of his for every one of mine. It is nice to have the additional flexibility of kiki, but too many turn out to be aji kikeshi.

[EdLee]

Specifically, I can't really remember being aware of the number of liberties in a contact fight, which is a pretty important consideration.

Hi John, I had no idea about this until I bought Bruce Wilcox's lessons on Contact Fights,
and very early on -- maybe on page 4 or 5 ? I don't recall now --
the lessons explicitly talk about the number of liberties in a contact fight.
I found it tremendously helpful at that time (around 10k ?).

I think my subconscious takes care of this...

I'm also pretty sure better players like yourself do take care of it mostly subconsciously,
and when you are in a cap race, etc., then you consciously and accurately read until the bitter end.

[Bill Spight]

I'm sometimes prone to overlook liberty shortage in tsumego). Does this ring bells with other people? If so, how do we move the liberty alert higher up the scale in our subconscious - simply more tsumego?

Here is something I ran across recently.

$$ Half eye
$$ . . . . . . . . .
$$ . . . . . . . . .
$$ . . X X X X X . .
$$ . . X O O . X . .
$$ . . O X . . X . .
$$ . . O O O X X . .
$$ . . . . . . . . .
$$ . . . . . . . . .

Click Here To Show Diagram Code

[go]$$ Half eye
$$ . . . . . . . . .
$$ . . . . . . . . .
$$ . . X X X X X . .
$$ . . X O O . X . .
$$ . . O X . . X . .
$$ . . O O O X X . .
$$ . . . . . . . . .
$$ . . . . . . . . .[/go]

I did not at first realize that it was a half eye for White. And I might not have in a real game instead of a problem.

I find myself leaning towards the idea that the most important thing for beginners to learn about, from a practical standpoint, is shortage of liberties. Very often there will be chances to turn a DDK game around during the dame filling stage.

The most interesting point was a section on "the number of stones". I hadn't latched on to it before, maybe because it sounds boring or maybe because it had the word "number" in it, but the basic premise is that when you decide to play in an area your choice depends on how many stones you already have there.

Back in the '90s I did some statistics on moves in pro games, which became the basis for my treatment of the tap, the step, and the touch. (See http://senseis.xmp.net/?BillSpight%2FbasicClosePatterns ). I also discovered that the balance of stones in an area seemed to be significant. Aside from plays that were not close to other stones, the most frequent plays were those where each player had the same number of stones, so that the player took a lead by one stone. Next most frequent were plays where the opponent had one more stone, so that the player caught up. Third most frequent were plays where the player went two stones ahead. (Here I counted a 2-0 lead.)

That may be a simple rule of thumb that you might teach to a beginner, that would help him to avoid over-concentration. But I have noticed that even some high dan amateurs will happily play to take a local three stone lead in the opening. (OC, doing so may not be a mistake, but the rule of thumb should alert you to think hard about such a play before making it.)

The only place that I have run across a similar idea is from Sonoda, who applies it even to a half board region.

[Bill Spight]

Specifically, I can't really remember being aware of the number of liberties in a contact fight, which is a pretty important consideration.

Hi John, I had no idea about this until I bought Bruce Wilcox's lessons on Contact Fights,
and very early on -- maybe on page 4 or 5 ? I don't recall now --
the lessons explicitly talk about the number of liberties in a contact fight.
I found it tremendously helpful at that time (around 10k ?).

The statistics that I mentioned in my previous post led me to recommend Wilcox on Contact Fights, even though I have never read it, because the majority of go plays were adjacent to a stone already on the board or diagonally adjacent to one. (I had read his EZGo articles in the American Go Journal, though.)

[EdLee]

Hi Bill, back in the Dark Ages, Bruce snail-mailed it to me on a CD.
These days I'm pretty sure one can order it online then download it from him.
I liked it a lot because he made it clear that the "beginner guidelines"
were exactly that, merely guidelines for beginners. In the later lessons, he emphasized
that reading and overall global considerations can override all the earlier "guidelines."
This was very helpful to me.

[RobertJasiek]

Numbers of stones (or plays made), as in the stone difference, can be relevant for efficiency considerations. Number of outside stones, as in the influence stone difference, can be relevant for development potential or fighting considerations on the outside, if the black and white shapes are similarly strong on the outside or if difference regions are compared.

[Boidhre]

Specifically, I can't really remember being aware of the number of liberties in a contact fight, which is a pretty important consideration.

Hi John, I had no idea about this until I bought Bruce Wilcox's lessons on Contact Fights,
and very early on -- maybe on page 4 or 5 ? I don't recall now --
the lessons explicitly talk about the number of liberties in a contact fight.
I found it tremendously helpful at that time (around 10k ?).

The statistics that I mentioned in my previous post led me to recommend Wilcox on Contact Fights, even though I have never read it, because the majority of go plays were adjacent to a stone already on the board or diagonally adjacent to one. (I had read his EZGo articles in the American Go Journal, though.)

His approach was interesting. The biggest issue is that it's a list of rules, this is excellent for programming a logic engine but isn't so good for human learning for most people. That aside, what he did was very good. He starts with giving a list of rules that essentially remove the possibility of tesuji or bad cuts from your close quarters fighting. These rules are not efficient but they're not supposed to be, they're aimed at 15-20 kyu players who are throwing away stones with abandon for little to no gain. Later he encourages people to read a couple of moves ahead in order to avoid situations where you need to use the rules. At the end, he does what Ed says, and basically says "if you can read it out that the cut is protected, you can ignore it for now" and similar.

Basically:

20k: Protect every cut, even if it is indirectly protected you might not realise this or be able to read this out.
2d: Protect what you think needs protecting. If in doubt, protect the cut.


Which seems a decent pedagogical approach. He explains that 6 liberty groups are super stable and can be left alone, later on he points out there are many shapes which have strings with fewer than 6 liberties but which are perfectly stable regardless of this either because they have unconditional connections to other strings, or the cuts don't work or similar. 6 is a maximum not a minimum for stability from what I understand to be his analysis.

[tchan001]

What's the Japanese name (not romaji) of the book?

[daal]

Yay or nay on these various points?

1. Like a few of the others, I learned to think about the number of liberties in a contact fight from Bruce Wilcox' Contact Fights program which had some fun heuristics such as when a string has a certain number of liberties, you must tenuki. (I've since forgotten the details). On my own, I came up with my own personal heuristic to be aware of a potential shortage of liberty problem, which is that if a string is enclosed and has 3 liberties, a fire alarm goes off. I suspect someone will tell me it's inadequate, but it has served me well.

2. Not in the bus, but while having breakfast I began to think about jump vs. cap. Big difference indeed!

3. I've never heard the word "kiki" before, so I certainly haven't internalized it, and although I try to force before defending, I've put off consciously thinking about kikashi until I'm stronger. Mistake? It's always seemed I had bigger things to worry about, but after taking a glance at SL on the subject, I can't help thinking that it's always good to have an extra stone or two when a fight breaks out. Despite this however, the only time I consider stone difference numerically is in the opening, and primarily restricted to the idea that if I am outnumbered 2-1 in a corner, I had better have a good reason to play elsewhere.

[Bill Spight]

The statistics that I mentioned in my previous post led me to recommend Wilcox on Contact Fights, even though I have never read it, because the majority of go plays were adjacent to a stone already on the board or diagonally adjacent to one. (I had read his EZGo articles in the American Go Journal, though.)

His approach was interesting. The biggest issue is that it's a list of rules, this is excellent for programming a logic engine but isn't so good for human learning for most people.

My pet peeve when I was teaching bridge was that people wanted rules.

[Boidhre]

The statistics that I mentioned in my previous post led me to recommend Wilcox on Contact Fights, even though I have never read it, because the majority of go plays were adjacent to a stone already on the board or diagonally adjacent to one. (I had read his EZGo articles in the American Go Journal, though.)

His approach was interesting. The biggest issue is that it's a list of rules, this is excellent for programming a logic engine but isn't so good for human learning for most people.

My pet peeve when I was teaching bridge was that people wanted rules.

I imagine the desired rules were of the format:

If X then Y.

Not:

If X but not Q, R, S or perhaps both T and V, then Y but maybe Z is better in this position, so keep an open mind.

[Bill Spight]

the only time I consider stone difference numerically is in the opening, and primarily restricted to the idea that if I am outnumbered 2-1 in a corner, I had better have a good reason to play elsewhere.

Not that evening the stone count is bad, but please let me disabuse you of that notion.

At the beginnings of modern opening theory, a few centuries ago, play often began with the sequence, 3-4 point, 5-3 approach, pincer, play in a new corner. Top level players had already learned that they did not have to defend against a pincer, that play in a new corner was usually better.

[TheBigH]

I'm a bit hesitant to count kiki as extra stones when my opponent also gets to play one stone of his for every one of mine. It is nice to have the additional flexibility of kiki, but too many turn out to be aji kikeshi.

Sure, but it can happen that the forced response to a kikashi is not all that close to the kikashi stone.

[snorri]

Liberties are of course important, but I don't think our brains work so well with absolute numbers at the speed we are required to evaluate positions, so over time we make optimizations through practice, and probably this is subconscious. I think strong players are pretty good at making decisions based upon relatively small differences in liberties even in rather loose positions that aren't clear capturing races yet. The large vocabulary of common shapes they have must help and makes it more efficient.

A book that explains fundamentals needs to explain them without relying on what later might be the most efficient way for us. I'll make an analogy. The other day I was on a plane and was kind of bored, so I amused myself by trying to remember fingerings for the major scales on an instrument I hadn't played in years. Some of the more common keys I whipped through without thinking. Although at one time I could do them all without thinking (internalization through brute repetition) there were some that I clearly fumbled and I knew I was fumbling them even though this was silent in my mind. Not only did they not feel right, but I could almost hear the same mistakes as when I'd first learned them. So I slowed down and went back to the very basics I'd been taught as a child---whole step, whole step, half step, whole step, whole step, whole step, half step...verified the shaky ones and felt the "aha" moment of familiarity where they felt right again.

Now if I'd put in a few more hours in I certainly wouldn't be thinking about the specific spacing on a major scale. It would just be finger memory again like the good old days. But it's a reminder that the fundamentals are still fundamental.

That example is easy. If I forget a major scale I can derive it, by understanding the fundamentals of the music and the mechanics of the instrument. In go, can I derive a joseki from my understanding of the fundamentals? On some occasions, yes. Other times, it's much harder. I think we underestimate how many "basic" josekis are not so obvious. I think if it took a couple of centuries of games before any professional thought to play it, it can't be that darn obvious.

As for the number of stones, it comes into play when thinking about overconcentration and efficiency. That has to be fundamental. Is this obvious? I'm not sure. Dosaku beat up players much stronger than me by exploiting the obvious, then. It's one of those things where if you take time to count (like whole step, whole step, half step...) it's obvious. If you think that you can trust your conditioning and you are out of shape, maybe you're in for a rude surprise, though.

Sometimes I think getting stronger is mostly about being able to see the stuff during the game that is so "obvious" to you after the game...

So from this lesson, I'll put forth another side to fundamental:

A fundamental is what saves you when your conditioning fails, yet conditioning is itself also a fundamental.