I feel like I know embarrassingly little about how to make a ko. I am usually good at keeping big ko threats on the board, only to reach a critical juncture and realize that i can't find a way to use them. That is, I can't start an important ko.
So, two questions;
1. Is there a formula, a path that is used to make a ko? Particularly a somewhat valuable ko? Poking around on a board, I can make a ko, but i can't really seem to apply that to an in game situation. It seems like they only sort of arise naturally. Are there any resources that you know of where I can read more about this? (also, since I am in Taiwan, buying print books is probably prohibitively expensive). I tried and failed to search on this site, and Sensei's wasn't really too helpful for my exact question.
2. Am I thinking about this the wrong way, putting the cart before the horse, so to speak? I often think of a ko battle as getting something for free, if you have more decent threats, so i would like to use that tool more, but is that the appropriate way to think of it?

劫爭辭典
http://www.sanmin.com.tw/page-product.a ... qQLc188SxV

_________________
http://tchan001.wordpress.com
A private blog on Asian go books and interesting go sightings
Now open only for a select few.
Go is such a beautiful game.

Yes, no offense intended, but you are thinking about it the wrong way. A ko is not a goal. It is a compromise that arises between competing goals.

None taken, thats why i asked. But I can't escape the feeling, when losing a close game with some great ko threats out there, that I am wasting them if I can't figure out how to make a proper ko, which has happened before...

start with the shape, and work backwards

I'm actually not entirely sure about that - often the situation is such that unconditional life is not possible/desirable.

One example is the sansan invasion into a 4-4 with a keima extension.

I'm not sure if this is what brodie is asking, but often there are ko "shapes". For example, "5" below asks for the atari at "6" with the intent to start ko at "7". There are other ko shapes, but this tiger's mouth on the second line is perhaps the most common.

Click Here To Show Diagram Code

[go]$$Bcm1 Prisoners:
$$ ---------------------------------------
$$ | . . . . . . . . . . . . . . . 6 7 . . |
$$ | . . . . . . . . . . . . . . 4 3 . 5 . |
$$ | . . . . . . . . . . . . . . . 2 1 . . |
$$ | . . . , . . . . . , . . . . . O . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . O . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . , . . . . . , . . . . . , . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . , . . . . . , . . . . . , . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ ---------------------------------------[/go]

While what Joaz said is good

I think the answer is also sort-of-yes. To me it feels like drawing a ponnuki in a tense situation (that has room to run or live) is often tesuji.

_________________
Pers. Obs.

actually, illluck, thats excatly what i was looking for. the example you gave is great, where could i learn about more "ko shapes"?

Also, i dont understand the following:



nor do i understand what you said, joaz. Are you saying that a ko is only a compromise between two players, like a negotiation? I understand that, but also thought it could be used as sort of a tool.

the ko threat offers your opponent a trade -- if they win the ko, then you take what you've threatened elsewhere. And visa versa.

that's why you want to make sure you're threat is big enough.

for me the best place to start thinking about Ko is when someone invades a 3-3 position. If you are strong and your opponent invades 3-3, think about how you can make a Ko for life of the group. These kinds of Ko, which generally cost you little and cost your opponent a lot if they lose it, are good places to start.

You might also think about playing some problems on goproblems.com and doing a "text search" for "ko". There is a good chance you will get some interesting problems that may help you think about the patterns where a ko might appear for life and death.

Oh, that first sentence was in response to Joaz's comment that ko is not a goal but a compromise, I forgot to quote

Mmm, I think Loons has a neater description of "drawing a ponnuki", since ko will always involve a ponnuki-ish shape. Besides that second line tiger's month and the other example given by Loons, another common ko shape is to double hane on the first line (e.g. moves 4 and 6 in the first diagram of http://senseis.xmp.net/?KoFightExampleFromAProGame3 ).

I would be a bit careful about placing too much focus on these kind of moves - often there are clean ways to capture/live (e.g. http://senseis.xmp.net/?BellyAttachmentTesuji ).

As for learning them, personally they sort of arose from playing games and tesuji problems.

I wouldn't worry too much about it. Keep your eyes open and your chances will come. The three most important places to look for ko is in life and death, in cutting apart and connecting groups, and in endgame moves.

Remember that ko threats are just as good for a ko your opponent starts as one that you start.

Also remember, if you have superior ko threats, your opponent may be playing submissively to avoid an unwinnable ko. That would seem to you like you never got to use your threats, but in fact you benefit a great deal from them.

There are some standard techniques and shapes for making kos, yes.

But really, the heart of the matter is this: as you get better, you may find more and more frequently that you recognize the various possibilities in a position without having to play them out. And you may judge that playing any of them right away is bad, or not worth enough yet, so rather than force the issue now in a way that might not be in your favor, you might play elsewhere, deliberately leaving these possibilities in reserve for later.

Then, later in the game, you don't simply decide out-of-the-blue "I want to make a ko now" and only then look around for a way to do it. Rather you already know that a ko is possible somewhere because you read it out 50 moves earlier, and your decision is merely whether or not now the time to do it.

This applies to more than just ko. Practice recognizing your options as they arise, whatever they are. And as you get better and see more and more, then yes, every so often, you will recognize that one of your many options just so happens to be the possibility to make a ko somewhere. And every so often, you will decide that making the ko is the best of those options and then it will actually appear on the board.

Ko is a technique, but not a goal as such. There are many life&death situations where the only way to live is by making ko. Some groups can't be outright killed but can be turned into ko, which is always beneficial for the attacker. In the kind of situation where you could live unconditionally with correct play, making ko is inferior result. There's also a third group of situations where you have the option to play ko for bigger payoff or avoid it for smaller gain (example). Which you choose depends on ko threats for both players.

Here's an example of attack by making ko:

Click Here To Show Diagram Code

[go]$$W white to play
$$ ------------------
$$ | . . . . X O . . .
$$ | . X X X X O . . .
$$ | . X O O O O . . .
$$ | O O . , . . . . .
$$ | . . O . . . . . .
$$ | . . . . . . . . .
$$ | . . . . . . . . .
$$ | . . . . . . . . .
$$ | . . . . . . . . .[/go]

solution:

Let's start by looking at a simple ko, see what white wants and seeing what black wants, and how the competing desires generate a ko.

Click Here To Show Diagram Code

[go]$$Wc The starting situation
$$-----------------
$$. . . . X . X . . |
$$. O . O X X . X O |
$$. . . O . X X O O |
$$. . . . O O O O . |
$$. . . . . . . . . |
$$. . . . . . . . . |[/go]

Click Here To Show Diagram Code

[go]$$Wc What white wants
$$-----------------
$$. . . . X . X W W |
$$. O . O X X . X O |
$$. . . O . X X O O |
$$. . . . O O O O . |
$$. . . . . . . . . |
$$. . . . . . . . . |[/go]

Click Here To Show Diagram Code

[go]$$Wc What Black wants
$$-----------------
$$. . . . X . X B . |
$$. O . O X X . X O |
$$. . . O . X X O O |
$$. . . . O O O O . |
$$. . . . . . . . . |
$$. . . . . . . . . |[/go]

As you can see, they both have goals, and those goals are incompatible. Both of them want to have a stone at S19. If White gets a stone there - and can hold it - black is dead. If black gets a stone there, black is alive.

Click Here To Show Diagram Code

[go]$$Wc The direct attempt
$$-----------------
$$. . . . X . X 2 1 |
$$. O . O X X . X O |
$$. . . O . X X O O |
$$. . . . O O O O . |
$$. . . . . . . . . |
$$. . . . . . . . . |[/go]

The direct attempt by white does not work. Black lives easily.
So white overplays just a bit:

Click Here To Show Diagram Code

[go]$$Wc The ko
$$-----------------
$$. . . . X . X 1 . |
$$. O . O X X . X O |
$$. . . O . X X O O |
$$. . . . O O O O . |
$$. . . . . . . . . |
$$. . . . . . . . . |[/go]

And the result is a ko.

Note one of the essential characteristics of a ko: somebody has to overplay. One player must conclude that if he stops to fill/fix the ko - or the spot where the ko is going to be - he won't get as much as if he allows the ko to materialize. As JTS noted, having superior ko threats is very useful. You can overplay again and again, knowing that you can win the kos. Your opponent must submit to the overplays because he knows that you can win the kos.
So creating kos can be a tool in that sense: if you have good ko threats, it helps to know how to overplay in such a way that a ko results if the opponent attempts to punish the overplay.

( FWIW, one could make the semantic quibble that if you have the ko threats, it is really not an overplay. I do hope that nobody does that. )

I really think overplay is a bad way to describe it... Not only is it the wrong term, but also confers a weird connotation. Starting a ko is often the best move - to call it overplay makes no sense to me.

Exactly. Sometimes the best result from an invasion is a life&death ko.
If a ko is good or bad depends on the rest of the board, it's not always easy to figure it out.

I agree. But it is the best that I could do on the spur of the moment. I'm trying to convey the idea that the person who creates the ko is reaching for something that he could not otherwise achieve.

Try this thought experiment: suppose that you have double ko on the board - which, as you probably recall - can be a source of endless ko threats for one player. Furthermore, suppose that the ko is huge, and that it is to your advantage, such that you have an endless source of game-winning ko threats. In such a situation, any attempt by your opponent to create another ko elsewhere on the board is an overplay. Any attempt by you to create another ko is not.

Those are the extreme ends of the spectrum. Most kos lie somewhere between.

I'm not sure if this is strictly true. He has the double ko to use for ko threats; you create a ko, he activates the double ko and then takes the ko. At this point trying to play ko threats to win the ko is pointless, but eventually he will have to forgo a move elsewhere on the board to win the ko; otherwise eventually you'll find a way to make a third ko, and he'll have to give up one of them.

And if it's an approach ko you can fill in the double ko. And of course if the ko is really huge, you're allowed to ignore the double ko.

I think the thought experiment you want is "imagine finding this local situation on the board when the ambient temperature on the board is zero, and all the ko threats have been filled in." Unlike any other situation in the game, a line of play that would be improper under that circumstance can be excellent under any other circumstance - even when all the moves left are 1pt gote.

I agree to say ko must involve overplay is wrong. Kos and overplay are orthogonal. Ko is a shape that can happen in a game of Go: it's like saying hane is overplay or one-point jump is overplay. It's nonsense. Sometimes those moves might be an overplay, sometimes they might be slack, sometimes they might be bad, sometimes they might be good, sometimes they might be awful, sometimes they might be brilliant.

What you could certainly say is ko is complicated, and therefore is particularly suited when a strong player wants to get a better position than he deserves against a weaker player. That doesn't even mean he has to overplay, often you can just sit there and play good moves whilst your opponent goes and makes bad ko threats. Of often the weaker player will be a chicken and avoid playing a ko when to do so is the best move.