What are good ways to learn to "read" a sequence of moves mentally (i.e. visualize it in one's mind), in Go? Is practicing on a goban useful, or is trying to stick to mentally imagining the moves better? or a mix of both?

(I am asking this question because I feel that better reading skills can have a strong impact on the outcome of a game.)

PS: I now realize that the question of reading sequences of moves mentally is only part of the story: in addition to "pure" reading, it is also extremely useful to learn to not read certain moves (because they have little potential, or because the outcome of some board situation reached while reading is known). In other words, learning how to prune the move tree helps read faster: it is wasteful to read down many branches of the move tree when theory can prune these branches (the theory of killing shapes, of their vital point, of semeai,…) or give them a lower reading priority (useful moves tend to exploit weaknesses, or choke the opponent, or have multiple good effects, etc.). In essence, learning to read well involves both the raw skill of imagining a sequence in one's head and the art of pruning moves or giving them a lower priority while reading.

  • @Rainbolt This question is not about memorization but about imagining moves mentally by just looking at the goban. Commented Aug 5, 2014 at 16:57

4 Answers 4


TLDR: Putting stones on the board is not always possible. Reading is super important. Practice it with tsumego and apply it in games.

You're definitely right that it has a strong impact on the outcome. Reading is the most fundamental skill in Go, and improving in that area is absolutely required to become a strong player.

In serious games, it is usually not possible to put stones on the board "just to see what happens". You'll have to imagine the results in your mind before you put down a stone. This is a fact that can't be changed.

However, as a beginner, you should try some stuff by yourself or in non-serious games (with knowledge of the other player, obviously) first. Then, it is absolutely fine to put the stones on the board to check what happens. You'll then quickly become aware of typical shapes (for instance a ladder), and be able to spot it immediately (if not, just keep trying!), without the need to put actual stones down. This is the first step on the road to strong reading.

In teaching games, you'll usually also lay down important sequences (during or after the game), in particular when they are not easy to see. Try to remember them so you'll notice them next time!

Further, you will need to improve your reading abilities. Typically, players practice solving tsumego for this purpose, as they require reading skills of a particular level (yours in this case). First make sure you really solved the problem - do not miss a strong, difficult-to-see reply by your virtual opponent. Then, to check your solutions, it is fine to put down stones to make sure you did not miss anything. It is also beneficial to have reliable solutions at hand with which you can compare your own, for instance by using a tsumego book.

Over time, tsumego as well as simply practicing reading in real games will improve your abilities and you will be able to read faster, deeper and more precisely. In particular, as you memorize the involved shapes, you'll immediately know which moves are impossible without considering them. This will drastically speed up the selection process and cut down the variation tree to very few branches, each of which you have to explore as best as your visualization abilities allow you to.

As a beginner, it is actually often sufficient to always read only 2 or 3 moves ahead to avoid stupid mistakes. Once this does not help you anymore, try reading a few moves more, and you'll certainly do more than most lazy dan players :)

This last tongue-in-cheek sentence is actually true - when they are not serious, advanced players often rely on shape exclusively without reading much, as they can already tell the results from the initial shapes (often, but not always correctly). This ability is beneficial in very fast games, but in serious games you still need to actually read (deeply) to find the best moves. Use memorized shapes only to reduce the tree width.

By the way, aspiring pro players typically spend a few hours every day on tsumego. It is, after all, said that pros solved every tsumego in existence at least twice. If you enjoy it, feel free to do alike! If not (which applies to many people), well, just do it if you feel like it - Go should be fun! Tsumego are often more fun when hand-picked and given by a friend, possibly to many people at once, with a joint resolution session which typically includes a lot of "Oh!"s and "Aaah!"s.

Regarding "visualization" of moves. I did not go into that point yet. This is something I kind of wonder about myself, as I found it difficult to see clearly the moves very deeply, in particular when "new" stones touch stones from a while ago or are even played on top of them (this is called "ishi no shita" or "under the stones"). As far as I can tell, you cannot work on this specifically - it's simply a result of studying and will hopefully improve as one gets better at tsumego.

As a final, practical advice: Copy or print out tsumego from a good book on cards or cut little paper sheets. I liked this a lot, but definitely start with the easiest book, as the problems are often harder than they appear at first! Then put them in various places in your house as well as in your pocket. Every time you are forced to wait for a minute or more, pick one of them and try to solve it. Repeat until the wait is over. I often did that in waiting rooms (even though I have to admit that it sometimes made me rather sleepy).

Quoting tasuki on the above works:

Cho Chikun’s Encyclopedia of Life and Death is a wonderful progression which will take you from an absolute beginner to mid dan level.

Next, we present you with the three most famous classical tsumego collections. Gokyo Shumyo deals mostly with patterns occuring often in real games. XuanXuan Qijing focuses on feeling and vital points, and the insanely difficult Igo Hatsuyo-ron is here just to make you cry.

They come without solution due to copyright, but you can mostly check them with friends and/or stronger players. By the way, I can absolutely confirm his statements, in particular about Igohatsuyoron :'(

  • Thank you for sharing your experience! I understand that tesuji problems are very good problems to start with; I guess they work as well as tsumego problems, for practicing training? Side note: iOS apps are very convenient, for problems (one of them even uses a spaced repetition system!). Commented Aug 7, 2014 at 10:20
  • I count tesuji problems to tsumego (I think tsumego translates to (generic) "go problems"), and yes, they are very useful. TimK mentioned Davies' Tesuji book, which I found terrific, particularly for beginners. I actually started learning about tesuji from that book. Some of them are also often seen in games, or (subtly) they are the reason a seemingly obvious move was not played, and it's fun to try and find them. -- I have few experience with Go apps, I only tried a few Android ones and found them so-so at best. I don't have an iOS device, but it's nice to hear there are good apps for that.
    – mafu
    Commented Aug 7, 2014 at 16:31
  • It is called ishi no shita, not shita no ishi – but this is still a good answer. You might also mention goproblems.com, a good site for practising.
    – PJTraill
    Commented Dec 15, 2018 at 19:16

When you're actually playing, you aren't going to be able to put stones on the board, so you need to practice reading exclusively with visualization. The best way is to do go problems. It's important to do them the right way, though. When you think you have found the solution, go through every response the opponent could play, and make sure you have an answer for it. If you can't do this, then the problem you're using are probably too hard for your current reading level. The first chapter of Tesuji by James Davies has a detailed example of how to do this correctly. The more you practice, the better you'll get.


The easiest thing to do mentally is to count. That is, "if I play this and my opponent plays that, how many liberties does each side have?"

Some of the more complicated plays, like long joseki sequences are best played out for full understanding.


Regarding visualization, try playing through pro games, first on the board, then, memorize them until you can play through them in your head. You won't succeed entirely in this, but, you will create more "space" in your mind for reading in actual games. Internet go and clients have been great for easily spreading and seeing game records, but the downside is that it is TOO easy. Reading is like lifting weights, the more you do it, the stronger you get. Also, work on ladder problems. In my case, my eyes always get confused when I try to read out which stones are or are not ladder breakers on the other side of the board.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .