Concise Problem Notation
Table of contents |
Introduction
Malcolm The notational system described here aims to provide a concise way of writing down a tree of variations. For instance, one may wish to give the solution for a tsumego which has several variations. It's less easy to read than the normal go diagrams. However it seems to be a valid alternative. As well as being concise, it also seems easier to write (on SL, or in a notebook) than making several distinct diagrams.
Rather than giving a formal definition, let's give some examples.
Example 1: A Fan Hui 5D problem
I got this problem from the mailing list of problems by Fan Hui, organised by the FFG (Fédération Française de Go). It's supposed to be around 5D level.
d ( e g! ( f a | a f c b ) | a f ( g e | e g c b ) | b c ( ( e | f ) g | a f ) )
"|" means "or". "!" means "tesuji". Parentheses are used to group sequences of moves.
In english, one might write something like the following. It wouldn't be clear though, one would have to give two or three separate diagrams:
"Black should play at 'd'. White cannot prevent Black from capturing the cutting stones.
If White then plays 'e', then Black 'g' is a tesuji. If then White 'f', then Black comes out of atari at 'a', and White can't then cut at 'b' because of shortage of liberties. If White plays 'a' after Black 'g', then Black sacrifices a stone at 'f'. White captures at 'c', but Black ataris at 'b' and White can't connect.
If White plays 'a' after Black 'd', then ..."
This gives more detail to the solution, but it's much more effort to write down! And one has to choose how many diagrams to make, in which order to present them...
Example 2: A classical problem
This problem is from the Xuan Xuan Qi Jing problem set (Xuan xuan Qijing problem 38). I found it very tough.
k ( j g! b a i h ( l d | d l ) | l b ( g j i a | a i | j g) | h b ( a i | g j) )
Hopefully I haven't missed out any important variations or made any mistakes!
Comments
Comments here please. Also feel free to give other examples of this notation for suitable problems.
Bill: It is a nice idea. :-) Here is an SGF style notation for the first problem.
d ( e g! ( f a ) ( a f c b )) ( a f ( g e ) ( e g c b )) ( b c ( e g ) ( f g ) ( a f ))
It is not as concise, but a bit more readable, I think. For comparison, with indentation:
d ( e g! ( f a | a f c b ) | a f ( g e | e g c b ) | b c ( ( e | f ) g | a f ) )
And this way:
d ( e g! ( f a | a f c b ) | a f ( g e | e g c b ) | b c ( e g | f g | a f ) )
Even more readable than SGF. :-)
Malcolm Thanks Bill. I hadn't thought of the sgf style notation. I agree that well-chosen indentation makes it easier to read. Which style notation do you prefer?
fractic: It's going to be a bit hard for under the stones play. At least on SL. And you don't get the instant recognition like with a diagram. But it could work pretty well in print actually.
Also I can't find the m in the second example should it be d perhaps?
Malcolm Yes, I don't know why I typed "m". I guess my brain wasn't functioning. I've corrected it, thanks. As for under the stones play, maybe one can use symbols like as well as letters.
Gresil: Why enumerate intersections one by one using a, b, c and so forth instead of using conventional board coordinates? Isn't it just an unnecessary symbol layer? Eg. in the second diagram A could just be called M16, B N16 etc.
Malcolm: Perhaps so. I think it's a matter of taste. Personally I find it hard to get my head around the usual coordinates. And a, b, c is more concise than M16, M15, M14,...
On this topic, this reminds me, I hope to make a page about Audouard coordinates soon.
Dan: I like this very much. One technique that has helped my chess tactics a lot is to do difficult problems and write down all the variations I can calculate. Writing down the variations is "cheating" a bit compared to actually solving problems at the board, but it enforces a little rigor in my thinking process that I can then try to emulate during a real game. This notation (perhaps combined with Audouard coordinates, which I also like a great deal) could make it possible for me to do the same thing with go problems. I'll give it a shot.