Go of Ten Go stories, game reviews and essays

Ten’s guide to studying professional games

Being a go teacher, I often get asked about how one should study professional games. While it is difficult — likely even impossible — to create a universal guide that will work for everybody, I thought I could fashion a fairly helpful framework that shows how one should go about reviewing the games. As long as one understands a professional game on a fundamental level, it becomes possible to learn something useful out of it.

First of all, let's make a rough sketch of what a general go game is like. I find it useful to divide a given go game in five stages, all of which have their own characteristics (in fact, I derive this five-stage division from Jeff):

Stage 1: Opening game (序盤 = joban)

The opening game lasts until the corners and sides have been tentatively divided out, usually ending by move 20 or 30. Fighting is scarce or nonexistent.

Stage 2: Transition from opening game to middle game

In this stage, some territories and eye bases get secured, invasions and reductions are made, and players in general develop their positions towards the centre. If a fight breaks out, the game moves to stage 3; if no fighting erupts at all, the game may directly move to stage 4. The length of this stage varies a lot depending on the game: in a peaceful Japanese professional game this stage might last from move 15 to move 80, and in your typical Lee Sedol game this stage might just about not exist at all (as the game would directly go from the opening to middle game fighting).

Stage 3: Middle game fighting (中盤, chūban)

In the middle game, there are one or more weak groups on the board that fight for their life. Victory is often decided by how skillfully the running groups are handled. In some games, this stage doesn't exist at all.

Stage 4: Transition from middle game to end game

In this stage, the weak groups from stage 3 have been ensured their life, and play switches to closing the biggest open borders of territories on the board. The first few moves in this stage are usually very big points-wise.

Stage 5: Endgame (終盤, shūban)

In the endgame, the borders of territories are finalized. Skillful play in the endgame revolves around keeping sente as much as possible.

One should note that there are valid reasons why the game is structured as above. Some amateur mistakes (and punishments) are as follows:

  • A plays stage 2 moves in stage 1. As a result, P develops more quickly all over the board, and thanks to that gets a better position in middle game fighting.
  • A plays stage 3 moves in stage 1 (tries a relentless attack on a single stone, for example). P makes a small local sacrifice and again develops more quickly all over the board.
  • A plays stage 4 moves in stage 2. P's stones get better access to the centre, and middle game fighting again favors P.
  • A plays stage 4 moves in stage 3. P kills a huge dragon from A and wins the game right there. Alternatively, P gets sente attacking moves against the weak group that still needs to live, and profits significantly in another way (by for example increasing the size of his territorial framework).
  • A plays stage 5 moves in stage 2. P is happy to answer, and gets stronger shape for his stones than he should have received.
  • A plays stage 5 moves in stage 4. P ignores A's small endgame threat, and takes a bigger move that encloses (or opens up) a sizable territory elsewhere.

I'm sure some or all of the above generalizations sound familiar to most readers.

The reason I made the above distinction between the five stages was to make it easier to approach a given professional game. As long as one understands what stage the game is in, it is possible to point out some factors that should be observed in the professionals' play. Here are then my in-line tips on what and how to think:

Stage 1:

Follow how the professionals spread out their stones on the board, and try to get a feeling for the general flow of the game. If there is a rare joseki variation, ponder a bit on why it was played. Don't expend too much effort here!

Stage 2:

Make a prediction on where the final significant territories will be located at the end of the game. Then see how the professionals carry out their invasions, reductions and territory-expanding moves, and see if they had similar predictions as you.

In addition, note how the professionals deal with weak groups. Try to figure out why some groups are sacrificed, why some groups go for an early eye base, and why other groups escape to the centre.

Stage 3:

Make a plan for how you'd strengthen and attack the weak groups on the board — doing both at the same time, if possible. If there is no profitable way to attack, is there a way to quietly make profit while threatening to attack?

Try to get a feeling for what the professionals are attempting to achieve with their fighting. Usually an attack is carried out in order to strengthen a group or to enlarge (or finalize) a territory. Other times, a player may be attacking to defend his weakest group in sente, so that he can get the first move of stage 4 (which can be a really big territory-enclosing move).

Keep track of when the players deem their weak groups strong enough to play tenuki, and when they find it better to play a "slow", group-strengthening move with a smaller territorial gain.

Stage 4:

In this stage, all the groups on the board should be cleanly alive. Get a feeling where the biggest endgame moves on the board are located. They are often located near the corners and the edges, next to the weakest living groups.

Make a note whenever a player makes a slow-looking gote move. Often, the gote move will have a big sente follow-up that makes it worthwhile! Alternatively, the move may be making a weakish group absolutely strong, fixing all bad aji.

Stage 5:

Try to get a general feeling for where the biggest endgame moves are located, and then see how the players played the game out. Note that the players fight hard for sente, often paying a point or two so that they can turn elsewhere first!

With the above tips, observing any professional game on a fundamental level should become possible. It is still recommendable that one try to find commentaries of professional games, but for an average amateur, they may either be expensive or too difficult to procure — in such a case, I can recommend going with the study framework sketched above.

Comments (10) Trackbacks (0)
  1. Interesting! I had heard the general outline (joban / chuban / shuban) several places before (as I’m sure many have), but the descriptions of likely punishments for “wrong stage” moves was quite intriguing. I also like how you’ve outlined the primary goals for each stage here.

  2. This is great, man, thank you very much. I found particularly interesting the division in 5 stages instead of the classical 3.
    Thanks again, will try to use it!

  3. Very helpful, thank you!

  4. This was a great essay. Very good descriptions (punishments, what to think) about stages of the game. I found this also useful when thinking about how I play the game. Thanks.

  5. Thanks for the nice reception, everybody!

  6. Great Tien!

    Now I only need to find some time to look at pro games… Stopped it a while ago :(


  7. This is fantastic. I will be referring to this time and time again in the future!

  8. Very interesting indeed. It sounds quite different from the commonly accepted scheme (at least among French players) that there are only three stages, fuseki – chūban – yose, respectively beginning – middle – end game. I have to say I never heard about or read the words “joban” and “shūban”. Introducing these notions also helps to understand the big ambiguity that sometimes arise from the use of “fuseki” and “yose” notions, that are sometimes presented as stages of the game, and sometimes as categories of moves, and are quite elastic. I guess I still have to study lots of professional games before I really understand these notions…

    Another question: I guess “chūban” and “shūban” are not pronounced the same way. I now chūban is pronounced tchü-ban’ (with “ü” like in German “für”), but how to prounonce “shūban”? Is it shu-ban’ (with a “u” like the “o” in English “do”) ? Or shü-ban’ ?

    • Fuseki, chūban and yose are indeed the commonly used terms, and while fuseki and yose convey their meaning quite well, they are in fact faulty terms. Fuseki doesn’t mean the opening game as a whole, but it is a term for a whole-board joseki (like the Chinese fuseki or the sanren-sei fuseki). Yose, on the other hand, means an endgame move (or possibly a sequence), but not the whole endgame stage of the game in itself.

      Chūban is pronounced with a similar “ch” as in “champion”, and ū actually means a long u sound (spelling the word chuuban wouldn’t be wrong). The Japanese u is actually a sound that’s somewhat between u and y — at least Swedish speakers can get it spot on easily. If you cannot make the intermediate version of the sound easily, it’s probably safer to go with a pure “u” sound. Shūban has the same double-u, and it would indeed be pronounced quite similarly to the “o” in “do”.

  9. Printed this out and hung it up on the wall behind my board. Much appreciated – Thanks!

Leave a comment

No trackbacks yet.