Go Training – How to Improve at Go?

There are many possibilities of Go training, but it’s important to understand that the most efficient methods should vary and they should be based on your current level, your strengths and weaknesses, your style of playing and willingness of changing it, your budget and amount of time which you can dedicate for studying Go.

Every Go teacher might also use their own methods of training, which don’t appear in this article. And if you wish to get a tailor-made study plan for yourself, simply ask your teacher about it.

Guidelines below are dedicated for Go players looking for various ways to improve their skills. If you can’t play this game yet, learn to play Go first and then come back to this article.

Beginning of your Go career:

If you are a very beginner, but you already know the rules, the most essential part of your Go training will be through simply playing Go.
This way you will get an initial experience with this game and after playing (and losing) some games you should be able to slowly start recognizing some basic patterns for capturing stones.

Even though you can play Go online, it should be an interesting experience for you to also visit a Go club and play there live and talk with other Go players.
Go players are usually happy to meet other Go players and it shouldn’t be a big problem for you to find someone who will be willing to teach you a bit in a Go club.

You can also take part in a tournament. It’s not sure that you would lose all your games – it’s a big chance other beginners will be playing there, too. But even if you don’t win a single game, you will start to feel the special atmosphere of Go tournaments, meet more players, make friends or even… find there your future wife/husband.

The next part of your Go training should be solving life and death problems (so-called “tsumego”).
The reason of solving them is not to memorize anything, but to increase your ability of foreseeing the next moves.

How to solve tsumego (Go problems)?

  1. You find a problem in a book or in an electronic device.
  2. You create the problem on the real board with real stones and try to solve it while looking on the board (the recommended way) or you simply solve the problem while looking at it in your book, phone or computer.
  3. You solve the problem in your mind (without playing your solution on the board, checking answer in your book or clicking anything on your electronic device). While playing Go it’s forbidden to undo moves and the same should be applied while solving problems. The problem is solved when you know where you will play after any possible response of your opponent.
  4. After you solved the problem, you show your solution to someone stronger than you (e.g. your sensei), so that they would mark it only as correct or wrong.

If your solution is correct – you move to the other problems.
If your solution is wrong – you try to find your mistake, then solve the problem again and then go back to the teacher to check your solution again, etc.

Getting to know only whether your solution is correct or wrong is more helpful for you than simply checking the answer. If you check the answer instead, you will get to know the answer even if your solution was wrong. In such case you lose the possibility of solving the problem again and you waste the major part of your possible reading’s improvement from that problem.

Why is studying fuseki (openings) not so important in Go?

The ability of foreseeing the next moves is the core of your level.
You can think about it this way – if you make a mistake in the opening (so-called “fuseki”) or generally in an empty area of the board, you can lose 2-3 points.
But if you make a mistake in a fight, thus you let your opponent live with his group or you die with your group, it can be worth even more than 50 points.

Single digit kyu and stronger:

Playing games (getting experience) and solving tsumego (foreseeing further) are still important parts of your training.

The easiest way to improve at strategical thinking, at positional judgment and at adjusting the local situations into the global scene is learning from your own mistakes.
To improve efficiently you should know what was your mistake, why it was a mistake, where it’s better to play instead and why it’s better to play in that place.

How to find your mistakes?

Look for mistakes while analyzing your games (especially your lost games)!

If you’ve lost a game and you don’t know why, not analyzing that game at all would be a real waste of your opportunities to improve.

It’s good to review and analyze your just-finished games by yourself.
Even better if you can analyze them with your opponent (you will also see their point of view).
The most efficient should be a review where you and your opponent invite a stronger player (ideally a teacher) to help you with the analysis of your game.
The teacher will point out your mistakes, will explain to you why those moves were mistakes and will also tell you why his suggestions are better.

Analyzing other players’ games (if they have similar level to yours or if they are stronger than you) in general should be less efficient than reviewing your own games. However, this kind of method of training can help you open your mind for completely new ideas, which you might be able to use in your future games.
Analyzing weaker players’ games doesn’t have big influence on your improvement.

Dan players:

It’s still good to solve tsumego, play many games and review your games.

At this point it might start to make sense to analyze your games with Artificial Intelligence. Bots will show you the best moves and from this kind of reviews you should be able to learn various ways of creating good shapes and various ways of playing certain exchanges of moves helping you to end locally in sente.

However, the AI will not explain to you why some moves are good or bad and why it suggests playing somewhere else. Thus, because of the lack of explanations, you will need to spend additional time for trying to understand its suggestions.

The good point is that the stronger you are the more you will be able to understand (and learn) from reviews with AI. On the other hand, when you are a kyu player, using AI for analyzing your games isn’t very beneficial – it will be hard for you to understand much from this kind of reviews.

If you are a strong player and you would like to learn more in less time you can also consider hiring a teacher who would analyze your games with AI and then tell you the conclusions.

When you reach a dan level, you can also start thinking about dedicating additional time to study openings, joseki and endgame (2-3 points of gain or loss start to be a big amount for stronger players).

You will find plenty of books which will help you in your openings’ studies.
But to really improve your openings, whenever you play a game and your opponent chooses an opening, you should always play the variations which you are studying.
And as your opponent can always play a move you didn’t see in the book, it’s very difficult to benefit from learning only theories of openings without practicing them.

Btw. Some time ago Chinese professionals started to invent new openings in a way that when somebody played a “new” (unusual) move and other strong players considered it as a good one, they were parting into groups and in every next game each of them had to play that new move in their openings.
Then they were grouping for analyses of those games with a special focus on continuations of that new move. If someone played a response that other players considered as a good one, then they were parting into groups again and in every next game each of them had to play that response for that new move.
This system was probably one of the reasons why Chinese top players started to catch up with Japanese and then Korean top players at the beginning of XXI century.

How to practice yose (endgame)?

There is not that much theory in endgame (so-called “yose”) and it’s good to learn it (e.g. from some books about endgame in Go).
Besides improving at accuracy of counting which endgame move is bigger, you can also improve by solving endgame tesuji and by learning to count faster (speed of counting points matters a lot as endgame is usually being played when players run out of time).

To get experience in endgame and also to start counting faster you can e.g. have a look at a pro game with a close final result, start analyzing it since endgame, then decide who is ahead, then fork the game in OGS (with e.g. 20-30s byo-yomi), choose the colour that is in your opinion ahead and win the game against AI.
Other option would be to ask another Go player to play an endgame from a random game. You together decide at what point you start the game, then each of you has e.g. 5min to count points and write down a komi which you can offer to play the next move. Then the player who wrote the higher komi starts the game and the game is played with komi written by that player.