Aimed at DDKs / weak SDKs / anyone who doesn't know the shape already, because it's really useful.

I post this not because it's profound, but because it's an observation I made when I was 12 kyu or so that has proved surprisingly useful, and I still see some strong SDK players on KGS (mis)read it from scratch each time. It's simple and comes up a lot, at least in my games. I'm surprised it's not taught as standard in the same league as nets and snapbacks, because it's a very useful shape to be able to spot instantly.

Anyway, I hope this post will be useful to someone. It's longish, so I wrote it in dribs and drabs in a rush - sorry! - so corrections and comments appreciated.

Motivated by this diagram:

Click Here To Show Diagram Code

[go]$$c Atari on the second line
$$ . . . . . . . . . . .
$$ . . . . . . . . . . .
$$ . . . . . O . . . . .
$$ . . . . O X O . . . .
$$ . . . . . . . . . . .
$$ --------------------[/go]

in which black is almost always outright dead even with other stones lying around, I asked myself: does the following work for white? In this post I explore this shape and related shapes.

Click Here To Show Diagram Code

[go]$$c Atari on the third line
$$ . . . . . . . . . . .
$$ . . . . . O . . . . .
$$ . . . . O X O . . . .
$$ . . . . . . . . . . .
$$ . . . . . . . . . . .
$$ --------------------[/go]

Of course, as it stands the answer is no:



So what sorts of things might happen? I find the most useful thing is strengthening of the marked white stone:



Application:


Some silly cases and caveats:



Also worth a read: add a second stone and sacrifice both.

Additional material at Sensei's.

_________________
a1h1 [1d]: You just need to curse the gods and defend.
Good Go = Shape.
Associação Portuguesa de Go

I teach it as standard to beginners, though later than nets and snapbacks. It's good as a tactical example to explain the strategy of why it's good to start in the corners and sides rather than the centre.

Aha! I looked for it but couldn't find it. Thanks. Hopefully the discussion is useful anyway!

(Though, if I may say so, like most SL pages I find it mostly very badly written. )

a problem that emerges:

Click Here To Show Diagram Code

[go]$$c
$$ ----------------------
$$ . . . . . . . . . . . |
$$ . . . . . O . . . . . |
$$ . . . . X O . . O . . |
$$ . . . . X O O . . . . |
$$ . . . . X O X . . . . |
$$ . . . . X X . 1 . 2 . |
$$ . . . . . . . O . . . |
$$ . . . . O . . . . . . |
$$ . . . . . . . O . . . |
$$ . . . . . . . . . . . |
$$ . . . . . . . . . . . |[/go]

how does black separate?

_________________
Tactics yes, Tact no...

Click Here To Show Diagram Code

[go]$$c
$$ ----------------------
$$ . . . . . . . . . . . |
$$ . . . . . O . . . . . |
$$ . . . . X O . . O . . |
$$ . . . . X O O . 5 . . |
$$ . . . . X O X . 2 1 . |
$$ . . . . X X . X 3 O . |
$$ . . . . . . . O . 4 . |
$$ . . . . O . . . . . . |
$$ . . . . . . . O . . . |
$$ . . . . . . . . . . . |
$$ . . . . . . . . . . . |[/go]

Like this?

ah but then there's this:

Click Here To Show Diagram Code

[go]$$c
$$ ----------------------
$$ . . . . . . . . . . . |
$$ . . . . . O . . . . . |
$$ . . . . X O . . O . . |
$$ . . . . X O O . . . . |
$$ . . . . X O X 4 2 1 . |
$$ . . . . X X . X 3 O . |
$$ . . . . . . . O . . . |
$$ . . . . O . . . . . . |
$$ . . . . . . . O . . . |
$$ . . . . . . . . . . . |
$$ . . . . . . . . . . . |[/go]

_________________
Tactics yes, Tact no...

Creating good problems is not easy . This one has too many possible variations.

I never promised this was a good problem! Thanks for the variation, though, shapenaji - but I suspect it'd take me a week to read this problem out fully...

You can always improve the page. Go ahead, it's pretty easy!

Edit: Oh, and I think discussion on these topics is always useful, especially when it's a personalized look at something.

_________________
a1h1 [1d]: You just need to curse the gods and defend.
Good Go = Shape.
Associação Portuguesa de Go

I never knew this...

do you have any other of these things i should know? haha

Well, while I think the page is badly written, I think it's far from a problem unique to that one page. I find SL really poorly organised, to the extent where one person alone (even one person with a lot of free time and spare motivation on their hands - which I'm not!) couldn't fix it. but that's a discussion for another thread, perhaps.

Can you share what you take to be the problem with that page - or with SL pages in general? I've always found SL to be ultra-useful and, for a wiki with so much raw information, very well organized. I read the page on the cathedral when I was 14k or so, and found it very useful at the time.

(I agree that the way the main article blends into the commentary on the article can be a difficult style, especially if you don't know who the main "deshi" are. But it's very honest to the wiki form. On a wikipedia article, the editors are encouraged to write in the Ten Commandments style of an academic reference article, but then when you lift up the stone you see a menagerie of worms and beetles scuttling about in the moist soil of the discussion page. A SL page lays out the obvious first, and then contains a mini-seminar on the more confusing aspects of the topic.)

Sure:

• I couldn't even find the shape that this thread is about. The search function does its best, but if I don't know the name of the shape, and it doesn't happen to be indexed on a page I can find easily (both of which are common), the information is lost in the void.
• Here is an example game linked to early on the beginners' study page. It uses language incorrectly ("captures" to mean "puts into a situation where capture is unavoidable"), has confusing grammar (white changes from male to female even within the same sentence - "if he plays in her own area"), and rambles on about two scoring systems (when the main point to a beginner should be "pick one, it doesn't matter"). I think this page is great, by the way.
• Here is the link that Phelan gave me. At the start of the page, it notes that the shape made is sometimes called a "cathedral", and posts three diagrams of some stones getting squashed against the edge in the sort of detail that a 25k would appreciate. Then the dan players charge in and start a high-level discussion immediately after, which no 25k will be able to follow and no 25k will necessarily realise is not intended for them. A telling quote, 80% of the way down the page: "Both before and after reading this page, I confess I have no idea what cathedral is."
• Here is the page of beginner exercises. I might try them myself later. They're too hard.
• Here is the page on vital points, linked to from the beginners' page. I don't think this page explains anything very well: it doesn't explain how to find vital points, or list common ones, it just jumps straight into some pro games. Anyway, enough beginners' pages, because I could list those forever...
• Here is a bad page about honte. It gives a washy definition, a single example, and then a whole stream of discussion - in this case mostly about translation from Japanese, not about go. It's interesting, but completely unsuitable for someone learning for the first time what honte is. Here is a page about a common shape which was marked as "urgently" needing "wiki master editing" in June 2009.
• Here is an unnecessary and badly laid out page; here is a similarly badly laid out page but which could potentially be useful; here, more of the same; this one is okay; this is far better. They're all saying the same thing.
• For comparison, and to show it's not all doom and gloom, here is a (fantastic) page on common corner shapes, and all of the pages linked to from there that I checked seem good too.

I have deliberately avoided very short pages, pages about books (which are sometimes even reviewed by their own authors), pages about users, deliberate discussion pages and non-serious pages. There are lots of these - probably more of these than about go itself - and while I don't object to their existence, they're almost all useless, and are not where effort should be spent. I've also avoided pages about very uncommon shapes or plays because they are bound to be short and badly written.

Here is an old L19 thread about it. Here is an unreadable SL discussion page about it.

Okay, thanks for the examples. Some of those pages you linked to seem perfectly good. The pages on honte and vital point give a definition, rather than a study plan, because these are not concepts beginners will understand after 15 minutes on Sensei's Library (as opposed to eyeshape and nakade, which a sharp beginner might well master from a single page). "4463 enclosure with 34 contact move" shows a number of variations with a few comments - what more do you want? I don't want to delete half of the branches, do you? Other pages aren't very good examples of problems with SL - if Robert Jasiek goes around assimilating various pages to the borg, well, RJ is a problem for the go community as a whole, not Sensei's Library specifically. But let's try to improve the others.

I think it's kind of awe-inspiring that you can find a heavily-commented, 9x9 beginner game on sensei's library. Everyone whom I've ever met in the Go world uses "captured" to mean "put in a situation where capture is unavoidable." I'm not sure why you're so down on such a helpful page. Have you ever read a commentary on a professional game? Inevitably when, during the end game, one player "puts stones in a situation where capture is unavoidable", the commentary will talk about capturing stones. But the beautiful thing about a wiki is that you can change problems you observe. I have removed the one use of "her" on the page.

I simply don't believe that beginner exercise 1 is too hard for you. (Although perhaps this is why you find the eyeshape pages confusing? ) There are at least two hundred excellent problems there. When I was a beginner I think I tried all three hundred. It's very obvious that the problems are getting harder and murkier as you go through them, and beginners can take that in stride. I agree that the SL editors went a little wild adding in-game problems from amateur dan games - the page would be more elegant if it were restricted to the thematic unity of the first 250 problems. On the other hand, having an opportunity to try my hand at some challenging, creative problems after doing the didactic ones didn't hurt me - on the contrary, I think it help show me how tactically fascinating Go is. Anyway. Options: (i) Leave "Beginner Exercises" as is. (ii) Delete the last 50-100 problems. (iii) Move the last 50-100 problems to a separate section on the same page, to make sure there is no confusion. -- Does anyone have strong objections to any of those three changes?

Indeed, no one will understand what honte or vital points are if they only have SL to learn from. If it's not a site that can at least claim to have these two basic concepts reasonably well covered, what is it trying to be?

On another note, if these aren't beginner pages, and don't involve a study plan, why are they linked to from the beginner study section? (Edit: apologies - honte isn't linked to, but vital point is.)


Awe-inspiring? Introduce a beginner to the rules, and tell them what it means to win or lose, and the first things they will ask will be variations on "what does a game look like?", or "how do I know where to play next?", or "what are you thinking about when you play your moves?". It's very basic pedagogy. It's the same rationale as the one behind commented pro games, or the one behind Malkovich games. These things existed for players of all strengths except beginners - it was an obvious addition to be made.


Is it useful, in a beginner study section, to define what it means to "capture" a group of stones (i.e. remove their last liberty) and then instantly abuse the new terminology? Compare the uses of "capture" in the captions for diagrams 3 and 7. I think it's just confusing.

As I said, I thought it was a great page. There should be lots more written in the same spirit. But it could (and, if aimed at beginners, should) also be clearer than it is.


I didn't mean too hard for me (though in fact some of the later ones may well be), I meant too hard for beginners. On the beginner study page, the link to these exercises comes right after capturing and simple uncapturable groups (two eyes or seki). But exercises 2 and 4 already involve reading throw-ins; 5 would be oshitsubushi except for a shortage of liberties; I skipped ahead at random to 35 and found a double snapback! What kind of crazy-strong beginners are you expecting?

It's also "hard" in the sense of unsympathetic. The first time I tried some "elementary" tsumego and couldn't do them, I just muttered rude words about how stupid I was and stopped doing tsumego. That's part of what being a beginner is like. Most people do not see go for the first time and think "awesome! I want to get strong right now no matter what", they think "awesome! This looks fun!" - right up until it's not fun any more because they're being silently humiliated by a bunch of strong players who think double snapbacks belong on a "beginners" page. This is also basic pedagogy.


I have strong objections to outright deleting worthwhile content, yes, but I don't have a strong preference between (i) and (iii) because I don't really believe the page serves much use to beginners, and I don't believe the last 50-100 problems are the only hard ones. (As I said, I'll try the problems myself. They'll probably be useful for me - even the earlier ones. I'm KGS 3-4k. This should tell you something.)

Oh, I missed this:


Some structure. It's a big mixture of common, simple variations, and variations that were played once in some random pro game (or even references to them without diagrams), and variations that nobody on SL seems to understand. There are comments like "black will play b to start a ko" where I can't find a ko, and unanswered questions.

Concretely: I'd like to keep well-established variations at the top of the page, preferably sensibly ordered somehow; questions about these or other variations can then collect at the bottom of the page, preferably under a 'discussion' heading, preferably all separate, preferably all well annotated. It's just a case of readability.

Interesting way of illustrating the difficulties of maintaining a wiki

_________________
Patience, grasshopper.

I take your point, but I'm not quite sure I agree.

There's nothing difficult about maintaining order and structure on a wiki, or indeed a forum thread. Everyone's just got to want order and structure. It's perfectly natural that, if I say "another thread, perhaps", and I get my request immediately ignored, I will have to make a difficult choice. In this case, I chose without hesitation to allow the derailment: irrelevant discussion isn't getting in the way of the content of this thread, because the content is in the first post, and I will take responsibility to edit relevant comments into the first post, at least for as long as I am active on L19. And that is how SL should be too: I actively encourage chat and discussion (obviously!), as long as it's kept separate from the main content, which should be regularly updated, well structured and easy to read.

Of course, there's the separate issue that this discussion merits its own thread. If jts or others still want to discuss it, I may ask a mod to split the thread. If it dies soon, probably no need.

I don't want to go into trying to identify and solve SL problems, even if I like seeing people put their efforts to it.
However, I noticed that no one seems to have mentioned that SL has a Position Search, that searches for a given pattern in the diagrams saved in the DB.

Did you try that?

_________________
a1h1 [1d]: You just need to curse the gods and defend.
Good Go = Shape.
Associação Portuguesa de Go