How to Teach Go

The following is a description of what I do when teaching an individual how to play go. This is derived from my experiences when teaching beginners monthly for a couple of years at the Massachusetts Go Association's beginners' nights. It seems to work fairly well for me with most people whom I teach; on the other hand, teaching methods always depend on the personal style of the teacher, and my perception of how well this works may be different from the perception of those that I've taught. Thus, I offer this up more as a case study to ponder rather than as the One True Way to teach go.

I start out by doing a brief overview of the rules, along the following lines: "Black and white take turns placing stones on the intersections, you capture like this (with examples, including multiple-stone groups, groups in edges and corners, examples showing when stones are connected and when they aren't), you keep playing until there's nothing more to do, and you take score by counting territory minus prisoners (with examples of territory)."

I try not to explain very much; I find that beginners can play a game while knowing shockingly little, and playing a few games really helps them develop a context for understanding the game that makes it much easier for them to get concepts that I want to explain to them. In particular, some concepts that I avoid initially are:

It's too complicated, it won't come up, and even if it is about to come up in a game, I play in such a way to cause it not to come up.
Life and death.
They won't understand my explanation before they've played a few games, and they won't need to know about it.
It's another complication which you can avoid for a little while.
Territorial subtleties.
It'll come up soon enough, and they need to see examples in practice.
Taking stones off the board at the end of the game.
I talk about this at the appropriate part of the first game.
Exactly when the game ends.
I talk about this at the appropriate part of the first game as well.

We then start playing 9x9 games (at appropriate handicaps; see the next paragraph). I emphasize that the point of the first couple of games is just to make sure that they understand the rules, so that if they think that a move that one of us has made is illegal when it's legal or vice-versa, that shouldn't be surprising. At the end of the game, I tell them about taking stones off the board that can't be saved; there's no need to mention life and death when doing that, or at least no need to mention two eyes making life, since in practice it's always pretty obvious that the stones that are taken off could be killed. I also explain my algorithm for telling when the game ends, which is that you keep on playing until, for each empty intersection on the board, you can say "this belongs to black" or "this belongs to white".

We continue playing 9x9 games. I find it best to play at a handicap that allows me to play at full strength rather than trying to play at a weaker level, though I may purposely not notice that I can kill my opponent's groups in some situations. (So I certainly avoid anything extremely sneaky; I usually even avoid playing on vital points of large eyes.) I start by giving them 6 handicap stones, and they can usually beat me by around the third game; I usually adjust handicaps after three wins. Typically, we get down to 4 stones by the time we're done (which is usually about two hours later); occasionally the final level is either 5 or 3. Different teachers of course have different appropriate starting handicaps; if you don't have lots of teaching experience, though, be prepared to have to add more handicap stones than you think would be necessary. (By the way, a good strategy to use when playing against beginners with at least 4 stones is to use a "divide and conquer" strategy based on building a wall of stones more or less down the middle; if you do that, there's a good chance that you'll be able to kill much or all of one side or the other.)

During games, keep your eye out for "teaching moments" that allow you to introduce new facts in a natural way. For example, they will soon run into several situations like the following:

picture of black stone on second line that's in atari

After it comes up three or so times, they usually figure out that the black stone is dead, I confirm that and congratulate them for figuring out their first go fact. At some point a ko situation may naturally arise, at which point I explain to them about kos. I try to go out of my way to prevent them from arising in the first couple of games; also, even if one does arise, chances are that a beginner won't even think of trying to retake the ko immediately, it turns out.

If you can, it's probably a good idea to try to remember how the game went and to replay a bit of it after the end of the game, giving some simple strategic advice while doing so. For example, you may suggest that the beginner try to concentrate on connecting groups together, giving some examples from the game in question where they could have done so. After having a few groups die, beginners often get paranoid about trying to save their groups; I try to teach them that connecting groups will save most of their groups and that it's not important to save everything. So in a handicap game with at least 4 handicap stones, if they save 3 of the 4 corner handicap stones, they'll probably win; if they save only 2 of them, they'll probably lose but it will be close; and if they save fewer than 2, they'll definitely lose. Don't do too much explaining at one time: you don't have to talk about every aspect of every move in the game.

Depending on how much time I have to teach them, I may take some time at the end of the lesson to explain to them about two eyes making life. It depends on whether you think that they've had enough experience with dead groups at the end of the game to be able to understand that concept; also, it's perhaps more important to work it in if you won't see them for a while and they'll be playing other beginners.

One thing that is very important is: Do Not Overexplain. Don't give theoretical explanations until the situation has come up more than once in practice; don't play out long sequences (= more than about one or two stones) on the board as an explanation; don't talk about why you did this instead of that unless they ask or unless it's important and easy to grasp. I used to do that kind of thing a lot; I stopped doing it as much once I watched other people do it while teaching to beginners and noticing that the beginners had this kind of glazed expression on their faces. The beginners would say "sure" and "uh-huh" when asked if they understood; however, they didn't mean "yes, I understand" by that, but rather "I'm willing to accept that what you say is true but I have no context for it and basically have no idea what you're talking about". Once you realize that this can be a problem, it's not too hard to recognize its symptoms when you're doing it.

I think that's about it. I should give credit for the idea of not teaching that two eyes makes life or teaching about kos to somebody else, since I didn't come up with that myself; maybe it was John Fairbairn who suggested that? Somebody on (Also, it's much more important to avoid introducing two eyes makes life early than it is to avoid introducing kos early, so if you can't stand to say the rules without mentioning kos, go ahead and mention kos.) The content of this page applies to teaching individuals, or maybe very small groups; I don't have any experience with teaching larger groups, and I don't know what of the above would transfer to that situation. At the end of the lesson, it's good to recommend beginners' books (mostly because there are some truly awful ones out there that should be avoided). I wrote a short note about that that you should feel free to copy it and give to people. (If you do so, edit it all that you wish; in particular I recommend that you add sections to it about where to buy go stuff and local contact information.) People can also look at the relevant part of my go bibliography.

Dieter Verhofstadt had the following comments:

I almost never point out mistakes. On the contrary, I nod enthusiastically or say "good move" when a good move occurs. By "good move", I do not mean a splendid tesuji that happens to work regardless of their objective, but an ikken-tobi that connects stones globally, or a thick nobi that makes territory while attacking. I'm not alone in thinking that a positive approach makes them come back. At any rate, it is more fun to remember good moves than bad moves, so they will learn more from the taps on the shoulder than from the impression of hopelessness.

There's one teacher at the go-ban. Some time ago I started giving lectures on basic concepts, and preparing discussion sessions. The audience is whoever wants to be part of it. Too often this turns out to be a all-but-one-teachers-and-one-pupil session, with the lowest in rank getting all the explanation. Not only do I often disagree with some of the explanation given by others, but also the overload of good advice completely spoils the balance I had set up between practice and theory. I myself try to shut up with respect when another player gives a lecture, even if he's much weaker than I am.

The person that introduced me to the game left the technical aspects of go from time to time, to show me how to hold a stone or how to sit at the table. That works particularly well for those who look upon the game rather as an amusing oriental feature than as an alternative to math. Also, it can come in handy when you try hard not to overexplain and an uncomfortable silence lands on the go-ban.

I quite agree with the above, especially the "one teacher" comment. I'm embarrassed to say how often I've been guilty of this, and it's almost never useful. For example, once I was watching another teacher teach a beginner, and that teacher said in passing that some stones were alive. I was watching, and pointed out a tesuji that the teacher had missed, so after a few seconds of talking, the teacher and I agreed that the position was more complicated than that. The thing is, though, this interrupted the flow of the teaching and delved into issues that the student wasn't equipped to understand yet; so even though what I was doing was a simple, factual correction, not a judgment call, it was still counterproductive. In the future, I've resolved to sit on my hands no matter what; it's much more important for students to get a smooth, coherent explanation then that every aspect of that explanation be optimal.

Any feedback about this page would be appreciated.

Up to David's go page or teaching page.

david carlton <>

Last modified: Wed Jun 28 17:36:36 PDT 2006