Skip to Content

tsumego

A study plan

I used to think that if I just played enough Go, I'd be bound to get better. The problem is that you can keep on making the same mistakes, and losing for the same reasons. I decided to set myself a definite goal: to improve by 5 stones in 5 months (from ~20k to 15k) and to make a definite plan of what, when and how to study.

Solving tsumego has to be a key part of any serious effort to improve your game. However, I think it makes a lot of difference what level of problems you work on. Some advice I've seen says it's best to solve easy problems fast - this fixes the relevant patterns in your brain and makes it easier to recognise them over the board. On the other hand, more difficult problems have new concepts to impart, and force you to develop better reading skills.

I know that my reading is poor, partly due to indisciplined tsumego work. Because I mostly do the problems on the computer (either at goproblems.com or using GoGrinder or SmartGo Pro on the iPhone), I've a tendency to click first and think about it later. This doesn't exercise my reading skills, just my ability to keep clicking 'Back' and trying a different point until I get to the answer by brute-force search.

So part 1 of my study plan is to solve six problems, every day, from the Graded Go Problems books, in my head if I can, or setting up the problems on the board and trying to read them out thoroughly before playing any trial moves. Once I'm pretty sure I know the solution, I can play it out and double-check I have found the best responses to each move.

This is working out pretty well, except that I think volume 3 of GGP is too hard for me at the moment, so I'm frequently getting problems wrong even after 10-15 minutes of thinking and playing it out on the board. This is frustrating and demotivating so I think I'll switch to volume 2 for the time being.

The second part of the study plan is to play a game every day on KGS. Because I don't play over the board much, and usually play slow games at DGS, my clock discipline is poor and I tend to waste time at the beginning, or alternatively play too quickly and get into a losing position early in the game. So I'm playing a medium time control game (20-30 minutes each with byo-yomi); if I lose the game, I analyse it the next day and look for places where I went wrong. If I win, I play another game the next day, and so on.

I'm definitely improving - in one sense it doesn't really matter if I achieve the goal by my arbitrary deadline, because just having the goal is enough to motivate me to study and play every day, and that makes a big difference.

Syndicate content