Where do I learn go

It might have something to do with your timezone, but in general you should be able to use an incognito/private browsing window, that way the website won’t remember what you did previously. If you close the window and open a new private/incognito one, you can actually circumvent the daily limit entirely by doing it multiple times (assuming they don’t use browser fingerprinting on top of cookies).

Anyway, I think I see the symptoms of a “hyper-cerebral” Go student here (mostly because I am one myself): sometimes if you want to improve in Go you have to accept that what you need is not an explicit explanation or a rational understanding of what’s happening, but you just need practice to let your instinct hone itself to Go patterns that are unfamiliar to you; this will eventually enable you to “unlock” concepts that were previously inaccessible.


I love those books, I have some doubts though if they are in-between. They explain a lot, but they also ignore many many variants, because the reader is supposed to work them out alone.

Hello there htrobert1

I wish you luck on your journey as you wrap your mind around tsumego, and I’m glad that lots of people have pointed you towards good resources.

I guess I just wanted to share an opinion because this is something that comes up for me when I see beginners try to learn more about the game, and focus on tsumego exclusively.

If I may make an analogy - [learning to solve tsumego] as a subset of [learning everything about playing Go] is somewhat analogous to [learning to parallel park] as a subset of [learning everything about driving a car]

Yes, studying tsumego problems can help you develop your reading skills. However, those reading skills are most useful

  • in the endgame stage of a game where there is a close capturing race, or
  • in a fight-heavy 9x9 game where there are lots of groups that are at risk

There are lots of situation in the Opening and Midgame stage that also require reading skills, but solving tsumego won’t do a very good job of preparing for those situations.

In order to prepare for all those other situations, I would supplement your study of tsumego with some time to study the following:

  • sente and gote
  • settling your stones
  • how to play a balanced opening (what parts of the board are most valuable)
  • how to manage direction of play
  • the transition between opening priorities and midgame priorities
  • how to utilize shape in both attack and defense

So just like in my analogy, learning to parallel park really well as the first thing you learn might give you something concrete to focus on, and have a nice set of progress bars from beginner to intermediate - but it’s important to remember that there are LOTS of other skills you’ll need to learn to drive a car besides learning to parallel park, and the specialized skills you’re learning won’t necessarily generalize to all the other things you’ll need to learn.

Good luck!


tsumego is usually about destroying group of opponent or defending your group
if you don’t understand answer, just continue battle of black vs white by adding next moves yourself until capture or 2 eyes happen

tsumego is not necessary, you can replace it by 9x9 games, it would be less lonely


Tsumegos are not only useful for endgame, but also

  • In josekis. Dying in a corner early in the game would be pretty bad, so better have good tsumego skills if your opponent deviates from josekis.
  • In the middle game: how do you know if your attack is sente if you can’t read the life and death status of your opponent’s group? How do you know if you need to respond to your opponent attack if you can’t read the life and death status of your own group?

19x19 includes not only crazy local battles, it also includes large-scale strategy. Learning how to deal with crazy local battles will not help you with large-scale strategy.
But if you are very bad at crazy local battles, you may lose in the beginning, before you had chance to apply your large-scale strategy.
In fact battle for the corner in the beginning is not rare on 19x19. With some ninja opponents its impossible to avoid.


The answer to your first question is - not often enough. In the past few years, I’ve mostly been playing against GnuGo, and making Opening / MidGame choices that lead the game away from life-or-death battles early in the game, so that might have shaped my thinking on this topic.

Having said that, I can certainly go back in my game history, and find human games where my opponent forced certain choices that made me either fight to save my corner, or invade their corner early in the game:


MOVES 8-21 (I am White - fighting for the corner I just invaded)

MOVES 40-64 (Invading Black’s upper right and creating potential for a living shape)

MOVES 80-92 (Rather than running into the middle with my weak group, I realized I could use Black’s shortage of liberties along the edge for a big capture)

(Funny side note: I probably would have completely missed this opportunity if I hadn’t read THIS POST IN THE GO MEMES THREAD earlier that week! )

However, if you look at a different sequence, like

MOVES 101-116

This (IMHO) required a completely different set of Go skills (i.e. shape, cutting points, stages of the game, territory vs influence, giving your stones somewhere to run to, using direction of play, etc), and none of those were things that studying tsumego would have prepared me for.

But yeah, my “comfort zone” - if you will - is to mostly try and avoid situation where I need to worry about life-or-death edge battles, and where I rely on various other Go skills. Here’s an example I wrote up in my 19X19 FOR BEGINNERS: SHAPE article:

So yeah, while I acknowledge the importance of tsumego skills, it is also quite possible to play many games without necessarily ever running into situations that might require those skills.

Also, I remember a thread some time ago (6 months? 10 months?) where someone was using tsumego sites exclusively to learn Go, and was being told that they were solving tsumego problems equivalent to a 12-15kyu level. However, when they (a 25 kyu player on OGS at the time) tried playing higher ranked players, they routinely lost - and they were trying to figure out why the skill level that the tsumego site gave them wasn’t reflected when they played 19x19 games against human opponents.

I made a similar comment at that time, and I stand by it. The game of Go includes many skills that studying tsumego will not teach you, and to over-focus on those set of skills at the expense of others may lead to confusion down the road. This is why I urge beginners to generalize rather than specialize early on.

Your mileage may vary, some cars not for use with some sets, contents may settle during shipping, void where prohibited.


When I play chess, I find that playing by “instinct” just means repeating the same mistakes over and over. Only considering every move of every piece methodically avoids this.

It’s true that go involves a dozen skills, and just focusing on one of them is not enough. Indeed I know an EGF 6d player who said he didn’t practice tsumegos until he became 1d on KGS. On the other hand he does teach tsumegos to his kyu students…

Improving at tsumegos is a long process. Conversely, at DDK level, learning broad strategic concepts requires less efforts. So I’d say that, at that level, spending 50%-70% of your studying time on tsumegos is about right.

Concerning your game: you lived with N19 but if your opponent had played Q19 instead of R19 you would have needed to add a move, otherwise Black can kill with O18, then White captures, then Black P18. In the game, since Black played R19, Black only has a ko. So I think that living with J19 would have been better.

1 Like

If you are completely new, you may not know what happens in a tsumego, or what the problem is

I’d recommend

As it is very old, i think you will need firefox to open it

It is starting with absolute basics like, where are the dead stones, illegal placements to play, how to capture stones, and some easy life&death, so you can get a feeling for what you may look out for.

This difference between Go and Chess is actually somewhat of an old adage. In Go, it’s much more common for players, even weak players, to “feel” that a move is good before they have any conscious idea of why it is (though of course it frequently happens in chess too, but mostly to stronger players I think).

Go is said to be a peculiar example of an activity that’s both “left-brain” and “right-brain”, in that getting good at it requires a lot of effort both in a creative/instinct-based way and in a rational/calculation-based way (although I think instinctive players tend to be the ones that improve the fastest). One needs both; for now you seem to be leaning too much on the rational approach, and it would probably benefit you most to work where you’re most lacking.

Anyway, I could write a bunch of stuff about this to try to convince you that it’s better to “let go” (badumtss get it) a little bit, but it’s not really any of my business and I’m probably not the best person for the job :slight_smile:


In chess the concept of “good shape” is non-existent.
In Go, the concept of “good shape” and the “direction of play” (also absent in chess) can take you from DDK to mid-SDK.
They are different games.


I’m not a strong chess player, but chess does have a notion of (for example) double pawns being inefficient, which might be considered equivalent to bad shape. And there may be other things that are similar to bad shape, such as overworked pieces, overpushing pawns, a “bad” bishop etc.

And things like castling, a “good” bishop, “knife”(knight) F5 and batteries may perhaps be considered equivalent to good shape in go.


Instinct is important to find good moves, but reading sometimes allows to find better moves.


Intuition has a major role in go. It really helps a lot to play. It’s not sufficient and reading will be necessary too.
It’s very hard to try to solve problems in a exhaustive way, testing all the moves. A problem match a level of play. You look at and with your go culture you have a intuitive selection of moves that you consider. Then you read and check. Even when you read you will have more intuitive selection.
If you don’t see then the problem is too hard. Try a more easy one until you assimilate enough to cope with a harder one.

For books, there is quite a hole in the go litterature like some books mentioned before for the first steps with a lot not explained. Seems you should be quite a genius if you want to learn by books. For that approach of the game, there is one i recommend, especially to the folks coming from the chess world “go for beginners” written by Iwamoto Kaoru. But sadly out of print. If you can find it…


This is why I got into the idea of wanting to make a beginner’s book like that myself and now I am getting into the preliminaries of writting the book for the next step to SDK. There really is a big gap there, however, for those that do speak English, I cannot stress enough how wonderful Janice Kim’s “Learn to Play Go” series of books was (and still is).

I honestly read those twice, went to DGS and immediately claimed to be 12k and after a few games where I saw the practical application of what I had just read, I was indeed 12k and rising fast.

Amazing books. No genius required (I have none). I cannot stop recommending those books. :slight_smile:

@Groin @htrobert1
I also found the book you mentioned in Amazon:
Here is the link

It is still in stock and with a reasonable price :slight_smile:

1 Like

Nice you found it!
There are a bit more books by now yes and i do appreciate the ones you mentioned.

To develop a bit more on “go for beginners” i got some enthousiastic feedback from chess players to who i introduced go. In the old now basic series (tesuji, life and death etc… ) of ishi press all was very enjoyable but far too high to really grasp. That collection was smashed into a unique book, basic techniques of go covering all and even more (like handicap strategies) but once again still too high aimed.

Then Iwamoto Kaoru published that “go for beginners” which has a bit of same format, systematic, exhaustive, organized and aiming to give a good understanding of all basics. All at a much less advanced level as the other books that was published. And all these seemed to answer very well to what a chess player was looking for.

Because that book was out of print at that time, i did a bunch of copy from one of my town library.

1 Like

I’d like to reiterate my observation that there are actually quite a lot of awesome resources for beginners, even complete beginners – but very often the problem is that it’s surprisingly tricky to connect beginners to those resources.

In fact, I’ve heard floating around the idea that it’s actually even more difficult to find resources for intermediate players, though it might just be a false perception.

And I think the issues that make it difficult to connect beginners to good resources are quite multilayered and thus tricky to address.

Off the top of my head:

One part of the problem might be the “competing standards”, as in, the energy to promote learning resources is inefficient because everyone is promoting different resources and they can’t agree on which one is best.

Another I think is actually the copyright, there are a lot of old resources that we might like to share/revamp/modernize/incorporate in modern resources but we’re unsure if we have the legal rights to do so (and the internet Go community seems to have a really annoying adherence to intellectual property, for some reason), and promoting paywalled resources when there are so many free ones out there often feels like a lost cause to be honest.

Reading this forum discussion, I was wondering if it might be worthwhile to work on improving and expanding the “learn to play go” area of OGS, and give it more visibility for newcomers.

(though again copyright makes it trickier than it ought to be)


OGS at first has a very defficient way to offer fair and interesting games to beginnners as i pointed this out since long ago. That should be a first step to renovate.


After thinking about it, I have to say this position has a bit of “ideally there would be a much better solution, so we should do that instead of any much more feasible but inferior solution” vibe to it.

I mean, if you have any concrete proposition on how OGS could offer fair and interesting games to beginners, I’d love to hear it. If you have expressed such ideas in the past, please provide a link.

Personally it seems to me that it’s something that it’s easy to identify as a problem, but there’s no straightforward solution to it, though I’m just one person with no directly relevant experience.

If we can’t come up with a clear way to solve that problem, I think “the first step” is instead to look for other ways to improve the situation, i.e. to compensate for the presence of the problem if it’s too hard to solve it directly.

Personally I come down to: there should be some sort of initiative aimed at beginners. When a new user signs up to the site, the site could show them a message like "are you a beginner? if so, you could be interested in (link)! ".

“(link)” could be an improved “Learn to play Go” page, or essentially an automated beginner course; it could be a program to pair volunteering experienced players to beginners to give the beginners a “mentor” of sorts; we might even try to start an OGS community project to create a new bot aimed at teaching beginners (though if that’s even possible, that’s probably far in the future).

Nothing is stopping us from brainstorming and even starting to implement all of these things at the same time, including trying to think about a possible solution to the problem you brought up :slightly_smiling_face: but we shouldn’t let the fact that “ideally we would just solve that problem and be done with it” stop us from trying out other solutions.