Reading Improvement


I’m a few months new to Go and I find myself struggling with reading a lot. I’m wondering, for those of you who have been playing maybe a year or more, if you feel your reading has improved or do you find that reading is a part of your game that hasn’t improved while things like direction of play, life and death, etc. have improved. If it has improved, what are some things that have helped you along the way?


The best way to improve your reading is by playing. Then doing life and death will help the most second. But, if you feel that you are still struggling.

I think you should try asking yourself some questions. “Am I trying to read?”. “Am I still playing nothing but 9x9?”

I only mention 9x9 because it is so popular on OGS and if your learning to read then its time to make the swap because in 9x9 if you see improvement in reading its going to look nearly non existent. The board is just way to small to actually notice any difference unless your simply just winning or losing comparatively to where in 19x19 there are many parts of the board where you can see if your actually making a difference

The “Am I trying to read?” question. Normally pertains to the players who play really fast. I don’t mean just simply the players who use blitz time limits. A lot of players will play move after move without really considering what they are playing. It is good to pace yourself. Just because your opponent plays at a faster pace doesn’t mean you have to play as fast as they do. One thing I did when I was learning was that I made a rule about spending at least 30 seconds at minimum before actually clicking on the board. So its good to think about how much effort that “you are really putting out” in your games on enabling yourself to see your own improvement in reading.


Thanks for the response! Yeah I often irrationally feel pressure to play faster if my opponent is playing really fast.

Participation in my study group! See OGS Study Group - Board 12

Seriously though, reading is a skill, and improving any skill takes practice, and my study group lets you practice reading in practical middlegame situations (12 so far) and get immediate feedback through answers from the professional players who played the game, superhuman AI, and other study group participants. Check it out.

Obligatory second answer is: solve Go puzzles.


Doing life and death problems properly improved my reading I think. The key thing is not to try out moves but only place a stone when you know it is the right answer. You know it’s the right answer because you’ve read out all the options, responses and paths to the solution. I feel life and death problems are better than actual play for reading practice as there are (in lower level puzzles) fewer options to read out. In a game I get overwhelmed with options to really practice reading many paths.


No, reading didn’t really improve, still struggling to go through very simple sequences. After several breaks it is my reading that got broken. Probably, I simply learned to not rely on it in games.


Improving in reading takes just one thing: practice

I think figuring how best to motivate and stimulate that practice is the key.

For some people, playing games works better, since the competitive context of the game might give them extra motivation to try to read and think deeper in order to beat their opponent.

For others, doing problems may be more effective, if solving problems outside of time and competitive pressures is more effective for them to learn.

For many, probably doing a mix of both helps.


Some people indeed claim they improved their reading skills by looking at a lot of book covers.


I have been playing for a year or more. And yes, I can safely say my reading improved. From my original struggle to read one move ahead, I can now safely read like 3 or possibly 4. I am not even kidding. And there were times when I was on daily tsumego routine.

Maybe I am just retarded, I did not really compare notes with other players around my level, but I think people tend to sort of idealize the image of “strong” player’s reading abilities.

Yeah, if the sequences is mostly forced and I concentrate really hard I can get to maybe 6 or even 7 moves, but realistically that’s about it for me :smiley: In terms of actually reading all the possible branches and variations of a local position even a crosscut can still mess me up.

For me it is more like I have seen this position enough times to just know what happens if opponent tries to cut me. I am not consciously reading it - as in placing imaginary stones on the board.

So yeah, it gets better with practice, I guess? But maybe the jump is not as significant as you would imagine. (Or maybe I just suck :smiley: )


Having issues with reading too, but it’s improving and much because of doing go problems in my phone while commuting.

I’m not going to try and put a number on it but I would guess that my computational reading ability hasn’t improved much.

Imagine you know nothing but the basic rules. You don’t know about eyes or liberties. It would take a significant amount of reading to handle the simplest nakade situation.

Understanding principles like eyes, ladders, nets, nakade, capturing races etc allows you to read deeper because you can more readily dismiss options along the way.

I have an untested theory that it might be possible to improve reading capacity by regularly practising Blind Go. Thanks to S_Alexander for putting me on to that one. I haven’t properly tested the theory. Any volunteers?

1 Like

It’s remarkably similar to reading text. If you’ve just started, all you see is strange configurations that look essentially random.

After you’ve learned the alphabet, you can identify individual letters.
After having learned vocabulary, you will recognize these words in a text.
After having learned synonyms and homophones, you will be able to understand some puns.

Go requires Go vision. Go vision is what I’d call the grammar of Go; recognizing the underlying structure. Which stones form groups, which spots are ‘vital’ or at least interesting?

Train yourself to hallucinate stones in sequence. That’s what reading is. In order to know what to read, you need shape intuition, which you get from observing and replaying games played by people who are WAY stronger than you, preferably professionals (or superhuman bots).


I think as you play more and more you reading improves naturally, but in my case it hasn’t improved as much. Instead I’ve been trying to focus on shapes, identifying bigger moves on the board, direction of play, etc. Even today I’m making so obvious mistakes (which of course I only realize afterwards) because I’m not really strong at reading. That is also another reason why I only play correspondence games, so that I have plenty of time to think of something, take step away and come back the next day…

In my experience, while basic reading skills come naturally as you play more, you’d still need to explicitly study it in order to gain a decent level of skill. However, it’s also not really needed if you can manage to compensate for it in other ways in your play!

1 Like

Well said, thanks for the response!

1 Like

In my opinion this is the most important thing: it’s very easy to play without trying to read anything. You have to get into a habit of reading out your opponents response whenever a situation gets questionable. Once you start doing that you also see you’ll make less of the kind of mistakes that completely turn the game around.

Unfortunately brains don’t like working hard, so you have to keep pushing them to work harder.


I’m barely SDK and I still suck at reading. Lately I lost some corners because of a wrong confidence they were already alive, so I tried to improve by doing Tsumego.
I hate them! I still struggle to solve easy problems and also I don’t like the fact that you already know in advance that there MUST be a solution. So I struggle even more and feel very dumb.
So instead I started again playing 9x9, which I had left for a while. Now I lose against DDK and my rank is harmed but it’s way more funny than Tsumegos and I have to read fewer variations in a smaller environment than on 19x19 board, so I believe is a good exercise anyway


I gave up on reading. My solution is: play fast games which only depend on luck and a bit of intuition. You will lose a lot of games, but the frustration is smaller.

Don’t underestimate intuition. Its in fact possible to beat 1d without reading. Luck is part of intuition.

Igiveup, to train intuition you have to play even opponents, not 20k.More time/move is also useful to FEEL position properly.

1 Like

Can you explain this thoery a little bit more? It would seem that practicing Blind Go would help you improve your memory. Much like the game of Memory helps you to improve your ability to hold onto information and recall it later in the game.


I thought that reading was anticipating how your opponent will react to the moves that you play. Being able to anticipate your opponents response for many of the following turns in the game. For example, reading is the equivalent of using Conditional Moves in a Correspondence game on OGS, plotting 10 moves in advance, and returning to the game later and finding you successfully anticipated (read) X amount of the moves your opponent responded with.

I understand that if you aren’t using a tool like OGS’s Conditional Moves or Analyze Match, you would have to hallucinate the following moves. But unless you are recording your projections on paper each turn, I find it very hard to believe that anybody could successfully estimate how well their reading has improved, based on their deliberate practice hallucinating X of your opponent’s upcoming moves.

If you deliberately imagine the next five moves the opponent will make, but do this each turn throughout the game, how could anyone be expected to know, successfully anyway, that their reading had improved? At best we can feel we have improved. But without some form of tracking data, I don’t think knowing is even possible. To illustrate my point here are some numbers.

Obviously real life games would vary a lot. But this isn’t meant to be a realistic simulation. More of a sound guesstimation. What Black is trying to imagine and maintain in their memory each game…

Game Number: 50 (reading takes practice, right?)
Board Size: 9x9
Player: Black
Black Moves Per Game: 35 (35[B] + 34[W] = 69, game ends w/80% board coverage)
Tries To “Read” X Move Ahead: 5 moves, (black + white moves = this total)

  • 35 Turns Per Game (multiplied by) 5 Moves “Read” Into Future Per Turn = 175 projected moves.
  • How many did Black get right on each turn?
  • 175 projected moves (multiplied by) 50 matches = 8,750 total projected moves
  • 35 turns per match (multiplied by) 50 matches = 1,750 total turns.

More importantly what Black is trying to truly remember is how many moves he projected correctly ahead on each of their 1,750 turns. And 50 games is super small when Go players recommend that you need to practice to improve. And that is just on a 9x9. The bigger the board the bigger the numbers. You’d have to track data to have any idea if you were improving.

@rishin I’ve been playing about 6 months. However, I’ve spent a lot of time researching the topic about how to get better at the game of Go. I’m currently at 15K strength. I learned a great deal about how to play better by playing Correspondence games and using the Analyze Game tool on OGS. I try to predict how the game might play out. I try several possibilities, playing 15 to 30 moves for each variation. Then I play the first move from the best variation. It teaches me to “read the board” better (predict how the game is most likely to progress). Also, it teaches me a lot about what kinds of move tactics and stone placement patterns fail the most and why. I cannot recommend it enough.

For the 9x9 board, practicing Go Problems is a quick way to train your mind to recognize common stone patterns and how to beat them. I recommend Cho Chikun’s Encyclopedia of Life and Death . It starts at beginner level and progressively gets harder. When it gets too hard, you can just start over from scratch, working through them again. Each time you do this you should be able to go further than you did before. There is definitely a proper way to do Tsumego, and a wonderful treatment of that topic can be found in the free Go book, 81 Little Lions: An Introduction to the 9x9 Board for Advanced Beginners , by Immanuel deVillers, page 20. This book is a gem and I recommend you give it a look.

I personally believe that playing Blitz games helps you to develop intuition in a way that “reading” alone does not. Perhaps they are two sides to the same coin. I play a lot of 10 second move blitz games. You don’t have time to analyze and think. You have to rely on your gut a lot more. When you first begin you will likely be awful. But you will improve over time. I rarely play normal timed matches.

On the opposite side of the spectrum I play lots of Correspondence games, analyzing each move I made, multiple times, until the end of the match. I also only play on the 9x9 board currently. I flirt with Tsumego on and off. But I can tell you, without a doubt, that my game has improved at living in corners and stopping my opponent from building two eyes. Before Tsumego my ability to stop an invasion was a joke. They really are important. However, I’m only on Puzzle 70 of Cho Chikun’s puzzles. But I’ve played through them 8 times. Before I master harder puzzles I want to have a quick reaction time on these early ones.

Overall, my ability to “read” has really improved. I regularly use conditional moves and will often read 5 to 7 moves ahead before my opponent does something I did not anticipate. In fact, you can use the Conditional Move tool to try and track your own progress at reading :yum:. I will say that I invest a fair amount of time into deliberately practice Go. I don’t spend much time in Go literature yet. I figure once I feel more comfortable on the board, or perhaps when I hit 9K, that I will begin furthering my Go education. For now I’m just focusing on the basics. But if you aren’t trying to focus and trying to remember and paying attention to your own progress, competing with yourself like it was a game, you will progress much slower than you have the potential to. Deliberate practice is essential.

1 Like

Are you metaphysiciansplurgist? :thinking:

Anyway. You’re conflating a lot of things here.

  • Technicality. When you predict something, that’s called a hypothesis. When you offer a logical, falsifiable explanation for said hypothesis, we call that a theory.
  • Reading comprises two dimensions; for the sake of argument, let’s call them distance and quality. Distance relates to how many moves you can accurately hallucinate in sequence. Quality relates to how good the moves are that you’re considering at each step.
  • You can improve both by doing tsumego as well as memorizing and replaying games - if you like, you can replay them blindly.
  • If you want to know to what degree your Distance improved, you will of course have to measure it (though by measuring it, you will inevitably produce practice effects). There are many possible ways to measure this, and you’d have to study which provides the cleanest or most reliable numbers, but one of the more practical setups would be to lay out problems whose solutions are trivial shapewise and entirely forced.
  • If you want to know to what degree your Quality improved, you first have to realize that Quality is bounded by Distance. If you can only imagine a single stone, problems requiring you to find out why the “most natural” move fails - three steps down the road - will be intractable. Then you have to realize that you need at least two dimensions to describe the difficulty of a given tsumego - the number of moves N required to arrive at the shortest valid solution and the unintuitiveness U of each move in the sequence. Generally, unintuitive solutions are called tesuji. There is no practical setup to test Quality for reasons I won’t get into here, let’s just say it would be a very complex undertaking.
  • The only practical reading test is to do n-move tsumego. The better your Distance, the more variations you can try. The better your Quality, the fewer variations you have to try.