Dieter: I see you are now using my term boundary play. Trouble is, the definition in SL is far from correct.
My fault. A better term would be boundary-settling play. A move that forms part of a moyo (putting down stakes prior to fencing off a boundary) could be called boundary moves. But should not. A minor endgame sequence such as hanetsugi on the edge could be called a boundary play. But should not.
Allow me to explain. The source word is the Chinese 收 (shou). I stress that because I think it's important to get away from Japanese yose, even though that - if used properly - can have the right connotations. We should remember that the ancient Japanese masters struggled with this term, too. In days of yore (e.g. Tale of Genji times), they used kechi. They switched to yose once they started to get the hang of the concept. I believe a very good case can be made that they made the concept through the influence of Chinese books.
This word shou has a wide range of meanings in ordinary language, but the prime one is to bring in a crop - to harvest. By extension, it can mean the crops themselves, or to collect or gather in. By further extension, it came to mean also to conclude, to end or to finish up. There were other meanings. A very useful one for some people may be "the horizontal floorboards of a carriage-box"
![Smile :)](./images/smilies/icon_e_smile.gif)
Try working that into a conversation!
But in go, the base meaning of 'harvesting' is the one that applies, with a strong connotation of 'finishing up'. There is no specific time element involved. I think boundary-settling move captures the necessary components. I use boundary play as a shorthand for that. Maybe I shouldn't. (The Chinese for the minor or pure endgames moves is 官子, incidentally - a separate category.
So why is this term important? It is, but not because I am pushing a book. I have no ulterior motive. I am trying to pass on what I observe. My observations have led me to the conclusion that the rendering of Japanese yose as 'endgame' is one of the worst afflictions that has ever best go in the west. Other serious ones are the confusions over thickness/influence and sabaki.
If we look at the corpus of commentaries in old Chinese go that I have compiled, the single most important term is 紧 which refers to a tight style of play in which you keep pressure on the opponent. I more modern terms, it might be seen as a way of striving for maximum efficiency, but I won't say more about it here.
紧is roughly of the same magnitude of importance as 先 (sente). Since pressurising the opponent tends to be sente of a kind, that's perhaps not surprising.
Next in order of important is 势 or influence (thickness is a subset of that).
Next in importance, numerically, is 收 (shou). Not just next in the list but high, high, high on the list of words in general. It is the 50th commonest word in old Chinese commentaries, and the ones above it - apart from the technical terms just mentioned - are common-or-garden words like big, small, at, of and so on, or purely tactical go terms such as connect. So masters like Huang Longshi and Fan Xiping clearly had this concept of shou very firmly in mind all through the game. The number of references to it early in the game is considerable, but it tends to cluster in the middle game. I have often remarked that yose can be used early in the game in Japanese, but it's not so common as shou. That's because the Japanese have tended to use other, additional terms. One of the easiest to understand, once you know it is a form of shou, is kakoi (an early surrounding move), a concept that Fujisawa Hideyuki tried to highlight.
This concept of shou corresponds, I think, to what you are calling 'early endgame' but I think that term should be deprecated. The two components are inaccurate, or mean very different things to different people.
Your next category was central shape moves. Again, I think this is an unwise choice of terminology. Shape means too many different things to go players. In addition, the concept can be widened by following the Chinese term zhaoying - call & response - which is a very common concept in the Chinese corpus. Possibly more so than in Japanese games because of the element of group tax. But we see this a lot of AI play. The famous move 37 by AlphaGo against Yi Se-tol (a shoulder hit on the 5th line) was one example. The point about a call & response move is that it impinges simultaneously on several potential boundaries in various parts of the board. It is a multiple-use move. A Swiss army-knife move. There are some other relevant aspects mentioned much in old Chinese commentaries. One is a strong emphasis on eye shape (allows later flexibility). Another is specifically 'outside influence' (i.e. atsumi thickness). Big points are high up in the frame. All of these are well enough known, of course, but one that is not so well known, and which perhaps corresponds to your concept of 'central shape' best of all, is the L-shape in the centre. Students of Honinbo Shuei will be familiar with this, but it is a major concept in Chines ego in the form of 'turning the head' (转头). Indeed, it ranks at the same numerical level as 'big points' and is above call & response.
When it comes to actually settling boundaries, well, obviously, that's the hard part of go. Where do we start? How do we proceed? But we do have some guidelines. Huang Longshi gave us a good steer with his 5 Lands theory. Each region of the board can be considered a "land", each with its own attributes (expressing things like degree safety, growth potential, attacking uses and so on). Time is also a major aspect (your 'urgency'). The presence of call & response moves can take a lot of uncertainty over timing out of the equation, because they imply miai possibilities. But then the most taxing part of the procedure is evaluating the result of a boundary-settling procedure. I find that there is too much of an obsession with counting in western evaluations. I'm not saying counting is always wrong. But it's often too hard to apply in practice, especially with time limits, and it focuses too much on territory (perhaps though thinking of shou/yose as the endgame?) and ignores evaluation factors such as safety, aji and the like. The Japanese cover this aspect of shou by using the word 'thick' (in a different sense from thickness in the opening). It is a
very common term. In fact, though, counting is often not even necessary. Once a boundary-settling manoeuvre has been played (even just in your head), you just need to eyeball the position to get a sense of whether his stones now look overconcentrated compared to yours. If they do, you have made a gain. No need to count precisely - you have gained at least one stone! If they don't maybe you should be looking elsewhere.
I find there is one other valuable benefit of thinking of 'harvest' as opposed to territory. I see many players map out a territory and assume it's theirs for ever. But that's just a field. You have to bring the crop home. You need to harvest it actively. Shou is pro-active. Yose, at least in western practice, is reactive - repairing a fence when you should really be bringing out the combine harvester.
So, I think you are on the right lines, but I recommend a wider remit and more attention to lessons of the past. And a correction for 'boundary play' on SL
![Smile :)](./images/smilies/icon_e_smile.gif)