I remain unconvinced by Bill on Kano. For the avoidance of doubt, I am not saying Bill's way of counting boundary plays is in any way flawed (nor am I able to say whether it useful or correct).
But my feeling is that Kano is being criticised for something he never said or intended. I have to stress it's a feeling, as I'm not qualified to dilate on the mathematics, but it's a strong feeling, and it's based on the following observations.
First and foremost the idea of double sente has been around since at least Shi Dingan referred to it in the Qing dynasty, and it is alive and kicking today in each of the oriental go-playing countries. Kano sticks to his guns in the 1985 edition of his book, and although it's a different example he still uses a large-scale one where the value of the respective sente moves is very different.
The huge Chinese "Practical Comprehensive Manual of Go" of 1997 gives an example with a totally different position but likewise a huge discrepancy in the value of the two sentes. One allows killing of a group if unanswered, the other just allows a non-fatal incursion, i.e. the same idea as in Kano's big 1974 example. Yang Jinhua and Wang Qun also give examples in Chinese, large and small scale, in all cases with different values for each side's plays.
The Taiwanese author Li Song gives examples, too, and also has a good introduction on the history of boundary plays going back to Guo Bailing, i.e. early 17th century.
I'll skip Korean examples, as I think the picture is clear enough: we have had in place a method of talking about boundary plays for centuries. It beggars belief that if this was flawed, someone - even if he had to be a genius like Go Seigen - would not have mentioned it. It's true we had quirks like people not noticing a Shusaku game had been miscounted until an amateur queried, or weird positions that previous rules couldn't cope with. We've even had Kano himself, as I recall the story, exposed by Matthew Macfadyen for a mistake in an endgame problem, much to Kano's embarrassment. But these are quirks and one-offs - double sente plays occur in every game, multiple times.
So has Bill defied the odds? Maybe, and the novelty of CGT gives some grounds for believing in a platform for new insights. But as with conspiracy theories, I always think Occam's Razor is a better tool.
I believe the Oriental usage of double sente is nothing more than a description, that works in the same rough-and ready way that I say my wife's dress is red but accept she may call it burgundy, cherry, salmon, fuchsia, etc. (and in the way of the world I have accept I'm wrong while knowing I'm right enough). In contrast, while Bill will have to speak for himself as to exactly what he means by double sente, what comes over to me is that he sees it as a cog in a mechanism, and if that cog isn't exactly machined the whole mechanism will grind to a halt. Great if he can do it, but it's hardly fair to Kano and the others to blame them for non-working cogs.
That reminds me of one of my favourite stories. A Japanese interpreter was working with a group of American senators in Japan. The Japanese side referred to something as the honmono, the genuine article. Keen to show his prowess, the interpreter decided to translate this as "the real McCoy". Whereupon a senator excitedly interjected, "Say, you know Scotland too? My family comes from there!" Faced with expectant Japanese faces the interpreter felt obliged to render this straight into Japanese. Result, baffled faces and muttering of "How the hell did Scotland come into this?"
My view is that, presented with the alleged mistake, Kano's reaction would be "How the hell did cogs come into this?"
If I'm right and the Oriental view of the term is purely descriptive, we still have to assume the description is made for a purpose. The main answer to that seems to be differentiate double sente from one-sided sente and double gote, the categories universal in the Oriental texts (some add reverse sente, others treat it as a sub-category), and to point out that, no other things being problematical, double sentes are played first. Implicit in that, of course, is an addition to the description of who has sente. Imagine the situation in the aforementioned 20-point + 7-point double sente at which Bill took umbrage where one side has an area large enough not to have its life affected, but where that position arises only because that side has just made a move to create that area. Obviously he has gote. Just as obviously the other side will grab the sente. As he plays it he will perhaps think of it only as sente play, but if he was given that position cold and told he had sente, and he wants to know where to play, it is useful to be able, descriptively, to give general advice along the lines of "give priority to double sentes".
That's a trite example, no doubt, although I think it is always implied in the Japanese that it will be known in each case which side it is to play, and RJ seems to be saying that this is not necessarily assumed in western methodology.
In practice, pros seem to operate more on the crude but practical lines sketched out by Uberdude, much as we can see in Genan's Igo Shukairoku. Moves are mostly categorised as X points gote and X points sente, while double sentes are taken for granted and reverse sentes require a slurp of tea or a cup of coffee. But pros differ from amateurs in two main ways. One is that they have memorised counts for large numbers of standard boundary positions, and the other is that they stress restraint in the timing of boundary plays, to ensure that full account is taken of all aji (either to use it or eliminate it), so that many sente plays we amateurs make are far too early for a pro's taste - pun intended
That seems to be the only way pros incorporate the concepts of local and global, and as far as I can recall, they have never used these specific terms.
I am very unsure about this, but I think it might be correct to say that pros have a bipartite approach. They make an overall count of territory and prospective territory and then decide on a
strategy. Quite separately, they calculate the values of boundary plays as X-points sente or gote and thus decide on a
move. In contrast, western researchers appear to me to be looking for a method that can integrate these two parts.
Login
Register
FAQ
Search
lightvector
is +4, and the next play gains 2 points.
- 
(not shown), gains zero points. But that just means that each move in the exchange gains the same number of points.