Freitag, 18. November 2011

On Solving Tsumegos

I think I might write about tsumegos quite frequently over the course of the next weeks. I've neglected them complety for at least 8 months and now hat I've (re)started doing them, they begin to fascinate me more and more.

As I am not a particular strong player, the tsumegos I discuss in the weeks and months to come will be quite basic (even embarrassingly so). What is interesting to me is that I begin to develop a, well, certain mindset that helps me to solve tsumegos and that promises to improve my game quite a lot.

I share these thoughts here, because I believe this change might be of interest for other players around my strenght (10-14 kyu'ish). So at least today's post is more about the way I solve tsumegos and what I think about them than about the problem itself.

This is the problem I will refer to:

Im convinced the changes in the way I solve tsumego were initiated for three reasons.

First reason: I completely stopped solving tsumegos on any kind of computer screen, but lay them out in front of me on a real board. I do not touch the stones until I feel I have read the problem out to the very end. I can not stress enough how much I feel that helps me in getting stronger!

Second reason: I do tsumegos frequently for about two months now. 

Third reason: I concentrate all my Go studies on life and death instead of changing topics every other day

So what is this change I keep on babbling about? Of course, I learn new concepts and I am confronted with new patterns I can then recognize, etc.  But that's not really the heart of it. I think the most important part is that I get more methodically and more and more patient in my attempts to solve a given problem. In my mind patience and method go hand in hand. 

You need the former to develop and stick to the latter. I will demonstrate that with a problem I came across today. 

This is a fairly simple tsumego I might or might not have gotten right two weeks ago. The task is simple: Black to kill. 

Until recently I would have started to go for the first point coming to my mind and then read from there. Today I start problems with a quick look and mental note what other points look interesting.

diag. 2
I "discovered" A and B and started with A. Point A was ruled out in about two seconds, but was the first move I looked at. C did not enter my mind until much later - which means that, despite my bragging, I need to look for starting points even harder.

diag. 3
Point B looked more promising as it threatens to take E19. If white on 2, 3 kills the whole group. 

A few days or maybe weeks ago I would have considered that the solution. The thought would have been something like "So that is what they want from me." Today I do not proceed until I am sure that white has no other way to defend. 

diag. 4
In principle there are four other ways white could defend. I dismissed A and C without any real reading as they apply absolutely no pressure on the black intruder 1. 

diag. 5
So I read B. The first variation I read was a success for white, as it catches black in a snapback. So black has to do something different.

EDIT: Who spots the mistake ;)?

diag. 6
I read this variation two or three times to make sure that I do not get it wrong and the eye at C is false. It is - so white can not play B in diagram 4. Victory was near!

diag. 7
So let's try white on D in diagram 4! As you can see, this variation leads to a Ko for life. Black starts by taking at A in diagram 7. It is still a Ko if Black takes at A immediately after white 2. 

diag. 1 (again ;) )
Now I was taken aback. The task was to kill, not to create a Ko for life. At that point I would normally have given up and looked at the solution. This time I went all the way back to diagram 1. 

C had to be considered (I then could hardly understand why I did not in the first place ... but that's how it went).

diag. 8
Due to liberty shortages of both groups. White can neither atari at A nor B. That looks promising. 

diag. 9
If white ataris from the other side black gives atari to the marked group. Connecting at A would not help, so it dies. 

There is no other way to defend for white, so black has fulfilled its task. 

It is a fairly simple problem I complicated for myself. The whole process demonstrates my weaknesses better than my strenghts. BUT: I did not give up and I was able to hold all variations in my mind until the end. So when I went to the solution I was rewarded with not only having the right answer, but also finding the Ko laid out in the references as well. I had worked through it. Though still very basic, I hope that this new way of solving things more methodically will be a base for the development of my strength.


  1. I admire your patience. To solve tsumegos on the screen is really a problem as it encourages a trial and error approach that can be harmful to your game play. I'll find your experiences very helpful and an encouragement to get a better board in order to have more fun practicing ;-D

    By the way: can you provide a translation for "tsumegos"? I had to look for it on

    but I don't find it very satisfying...

  2. When it comes to a translation I do not know more than you. Wikipedia says that tsumego most likely derives from:

    Tsumeshogi (詰将棋 tsumeshōgi?) or tsume is the Japanese term for a shogi problem in which the goal is to checkmate the opponent's King. Tsume problems present a situation that might occur in a shogi game, and the solver must find out how to achieve checkmate. It is similar to a chess problem.

    But that does not really help ;)