Although I strongly suspect I am wasting my time, you don't find out anything if you don't try...

Robert Jasiek has complained in a separate thread that no-one is challenging his theories, and intimates that he would relish a chance to defend them. This post may disappoint him a little in that I don't dispute his theories. I admit I do sometimes have a problem with the mode of presentation of them, but the research and attempted application are worthy tasks which I support. But I can neither refute nor support the theories themselves - beyond vague impressions based on experience or reports in related fields - because I have not studied his theories enough to be able to apply them. Specifically, I bought Volumes 1 and 2 of his series but I found them too chewy to go on to Volume 3.

However, while pondering whether to buy Volume 3, I did read its sample pages and serendipitously I came across a set of 214 games based on a specific fuseki that was mentioned there and that may be an ideal study package for perhaps any level of player. In other words, this post is really to offer a study project in which Robert's theories play only a part. It is not meant to be about the features of his books. But I do hope his ideas can be referred to as part of the study, and that he will indeed welcome the chance to air and discuss them in a concrete setting.

Why is Package 214 ideal?

* Of this set of 214 games in the GoGoD database, Goldilocks would have said "not too many but not too few".
* The games span the last quarter of the 20th century to the present, and so can be classed as modern.
* The games are nearly all pro but have (unusually) enough amateurs involved to allow a pro-am comparison.
* It covers both fuseki and joseki.
* The joseki is the ever popular Magic Sword (this will probably explain the unusual incidence of amateur games).
* The basic fuseki occurs up to move 7 (i.e. not too complicated) but there are enough games with colours reversed or with extra moves played first to allow interesting comparisons with the basic fuseki.
* Quite a few of the games have easily available English commentaries (these are listed in Kombilo by Ulrich Goetz).
* There are also quite a lot of commentaries in Japanese and Korean in yearbooks (i.e. short and so manageable even if you have to use a dictionary).
* The basic fuseki and related josekis are given in Robert Jasiek's Joseki Dictionary, so you can use that to help you study and/or evaluate his system.

There is nothing special about the fuseki itself. I just happened to recognise it because of the Magic Sword connection (I have put a history of this joseki on the GoGoD CD). That led me to look up the position via Kombilo. I was looking for something to do with the 3-3 point, which we can forget about here, but by a fluke the results seemed useful and interesting in other ways. Given the above list of attributes I thought it could make a valuable study project on L19 while possibly allowing some discussion of the new theories. Sticking to the same corpus of games and so staying within tramlines, people could share up-to-date insights, and some could even test them via Malkovich games. If enough people were to get involved, I'd consider adding some of the comments from the oriental sources.

Let me repeat that this is not directly about Robert's book. It was just the tiny source. The stream here is of my own making. If it turns into a river through discussion here, then I may not have wasted my time

The source position after Black 7 in the game below was the last one given by Robert in his sample. I have restored the orientation from his version to the original, added the extra moves, and also added the comments from the yearbook.



About this Robert said of White 8, with the necessary adjustments in [ ]:



Reading that in, say, Go World, I would have accepted all that (with the minor exception that "cut" would normally be "separation" for a native, but this is a known preference of Robert's).

When I trawled for the position in Kombilo after Black 7 I got just three hits. That was a small enough number to overcome my usual sloth, and so I decided to turn round (ouch!) and extract the relevant yearbooks from a bookshelf to look up the commentaries. There were commentaries on two of the games, including the one cited by Robert.

In the latter case (Komatsu vs Otake), there was no mention of a mistake or a novelty or an anomaly in the fuseki. The first move to elicit comment was White 80. In the other case (O Rissei vs Yamashita) O likewise did not play the move recommended by Robert and likewise was not censured in the commentary - the first comment was on White 14. It would be too easy to say that two of the very best pros in the world, plus a Taiwanese 6-dan in the other game, plus commentators, had disagreed with Robert and that no pro had, by selecting his move, agreed with him.

The search was on the full board position. If, however, you search on the top half of the board only, you get the list of 214 games I am talking about. In that case you get nine moves (including tenuki) tried by White. In the vast majority of cases the results are after seven moves (as in Otake-Komatsu), but there are enough after eight moves (i.e. the fuseki colours are essentially swapped) or after nine moves to make these differences worthy of comparison.

But ignoring the whole-board position and so limiting ourselves locally to the various White moves tried after move 7 (eight, I think) we have to observe that the most popular by far is A, i.e. firmly in line with Robert's recommendation (BTW the overall win rate is 50% so presumably that's not a major factor). This move is played in about 47% of cases. The second most popular is B, occurring in about 12% of games. The three moves C, D and E each occur about 7% each, and of course these include Otake's move, D (F also had a handful of adherents - the rest were sporadic).

If we look at the games where D was played with different arrangements of two white stones of the bottom (all were with a stone in each corner), those who apparently didn't think D was a mistake included Yi Se-tol and Ishida Yoshio.

There are many possible interpretations of this state of affairs. Amongst others:

1. There are problems with sample size or statistical oddities

2. Robert was right and D was a mistake - even pros nod

3. Fashion is a major factor and A was the new black (? most likely line to lead to a fight)

4. A is best most of the time but with the particular lower configuration of the Otake-Komatsu game, it's not.

5. It doesn't matter two hoots - all the moves are equally good (or bad)

Etc, etc - but that's what a discussion would explore.

Just to kickstart a discussion, I offer a couple of points. One is that while I accepted Robert's comment above, when I later looked at the full game I was confused. White did not seem to come under attack in this area, and in fact seemed to control the centre. Further the commentary suggests White did OK in this area but overall his strategy led to a confusing game and he went astray not in the opening but in the middle game. The other is that (as often surmised here) fashion is indeed an important factor. T Mark and I have just been asked to check a paper for publication in a scientific review which uses the GoGoD database to demonstrate that fashions do exist in go openings. (Actually "fashion" is my word and is not mentioned in the paper. There it is "strategic social learning" - part of evolutionary theory. I can't pretend to have understood the maths, but the pictures were pretty and so I was convinced. This all stemmed from our catchphrase "White 8 is bad" BTW. We'll leave it to the author to announce when it's published.).

There is a third point and one that interests me most, but it is probably too intractable for here. My recent readings in the chess world suggest that the better the player the more likely they are to be fuzzy in their evaluations, i.e. they shy away from condemning moves as unequivocally good or bad. Because of their great experience they can include more variables, which makes them more cautious but also more flexible. On that basis, Otake et al. may have seen something unusual in the position (the mokuhazushi being the prime candidate) that led them to prefer D. But even if you accept that, players who are not at their elite level still have to make choices. Are they better off just cutting down on the variables? And, if so, to what degree? That seems a sensible strategy but, as I've mentioned elsewhere, it seems not to find favour with the hottest thinkers in the chess world. The favoured strategy there is to just keep taking the pills: cram in more and more study of concrete positions. Cutting down on variables is akin to not taking your medicine. It's too apt to lead to generalisations. The beauty of Package 214 is that we can talk concrete specifics. But the beauty of L19 is that we can discuss any aspect we like.

Just a couple of initial comments, not really to the deep point raised (but perhaps fun nevertheless):

First, no one has played the one space jump since Otake in 2006. The 63 times (per GoGoD) that the right half position has come up in the meantime, White has chosen something else.

Second, Otake has played the left side of the board (3-4 and 5-3 combination) sixteen times as White. So quite possibly he does have some more than usual deep thoughts on what the implications are for that formation.

Third, yes it is true that Yi Se-tol played the one-space jump - the first time that he had White. It is also true that he played the large knight move the two later times that he found himself in the same position. Did he learn something the first time or does he just enjoy variety? Meanwhile the other Yi (Ch'ang-ho that is) has been up to bat seven times as White here and picked the large knight move every time. YMMV.

_________________
Dave Sigaty
"Short-lived are both the praiser and the praised, and rememberer and the remembered..."
- Marcus Aurelius; Meditations, VIII 21

Luckily the thread subject did not prevent me from reading the thread:)



This is an exaggeration of what I have said but the direction is right.

The study object you have chosen does not exemplify one of my major theories. It is about a specific strategic choice for White's move, which p.160+ discusses. This is still interesting for discussion, of course.



I use "cut" as a go term while I do not use "separation" as a go term. Therefore it is more a matter of whether one likes this choice of go vocabulary than a matter of not being a native's first choice. An ordinary native speaker of English would not use go terms at all...

It is basic advanced player's strategic knowledge to attack from a stronger position when having support and to avoid attacking from a weaker position when not having support. Therefore I make my book's comment that the lean and attack joseki choice was a mistake. One does not need advanced opening knowledge for that. Similarly, in a locally weak position, it is a good idea to settle one's group quickly; White A does that.

Since also top professionals make more mistakes than one might expect, referring to their average playing strength or their missing comments does not provide evidence for good play. Professionals are not above applying the fundamentals.

The strategic choice is essentially independent of the exact position of ordinary white corner stones in the two remote corners. The black corner stone in the adjacent corner on the pincer side dominates the decision.

Although the example is about opening choices and a specific white strategic choice and so does not apply Joseki 3 Dictionary's major theory, there is still applicable general theory in the book. It is related to functional joseki types and described in chapter 2. The theory is for either relevant joseki type. Note that the theory applies to both opening and middle game! In particular, the theory states all possible, in principle available major reasons for choosing a joseki of a particular general type. Among all the possible reasons, the following most likely candidate reasons can be chosen for the given position:

Most relevant candidate reasons for the joseki type 'lean and attack' (after White 1 in my book; White 8 here):
a) An opposing moyo is prevented.
b) A weak friendly group is defended.
c) Opposing stones are hurt.
d) Prevent opposing territory where the group is running.

Here, (a) and (d) are weak reasons because Black makes good moyo or territory elsewhere while White prevents them only where his group is running. (b) is an insufficient reason because White can defend his weak stone better: by choosing a 'quick settling' joseki, he makes territory and does not help Black to make as much moyo or territory in the 'lean and attack' variation. Here, (c) is a weak reason because Black has no problem to defend his temporarily hurt stones and build moyo or territory, while White hardly builds either so far.

Most relevant candidate reasons for the joseki type 'quick settling' (after White A in my book or here):
a) A weak friendly group is defended.
b) Opposing territory is denied where the group lives.
c) The group needs emergency life.

In the typical 'quick settling' josekis following White A, (a) is applied very well because the previously weak white group's life is defended very well. (b) applies while Black does not make a lot of new territory elsewhere. (c) applies but is an exaggeration in the given position; moves such as White 1 also provide life, so emergency life is not needed for only the reason of making life per se.

In the given positional context, (a) and (b) are good reasons for choosing a 'quick settling' joseki starting with White A. For comparison, the available reasons for a 'lean and attack' joseki are much less convincing when applied here to the given position.

Hence, although the game example in my book does not discuss my major evaluation theory, my functional joseki type theory of intermediate importance applies well.

For a more complete study of the example position, one would also at least consider Strategic Choice 2 as White's third alternative in his Strategic Choice 1 on p. 160. To simplify, unnecessarily Black would be given a major strategic choice at all. So White need not make strategy more complicated than choosing 'quick settling'.

Now compare all of the above analysis with the - according to John - mostly entirely missing comments by professionals on White 1 versus A ;) Not only does my book comment at all and discuss the strategic choice, but it also explains a background of generally applicable theory for both opening and middle game.

I won't claim to understand the fuseki in question, but given that top-tier pros have played this I'll take it on faith that the jumping out is not obviously a mistake to professional eyes.

The existence of the 5-3 stone in the bottom left is uncommon, and my first reaction upon seeing the line of play in the game was that White is trying to use this to best advantage. I can give you an (uneducated) speculation. The usual approach against the 5-3 stone allows White to press black down on the lower side, and I suspect White's play in the game is to make this approach unappealing to Black. By jumping, White starts Black off with a low position on the bottom side so that Black becomes less interested in the approach move and forced to play elsewhere.

I am reminded of one of John's books where he talks about the "breadth" of a game. I've forgotten the the details, but the idea was to various lines of play were selected to limit an opponent's strategic options. I wonder if this game was an attempt by White to do the same. The fact that is hasn't withstood the test of time might indicate that it didn't turn out as planned.

In contrast to the magic sword, White 1 (book) or 8 (here) and A limit Black's variety of options in the joseki corner. White 12 at K4 or K5 (here) would limit Black's options in the 3-5 corner. However, the played White 12 (here) does not follow this idea; Black is given attacking options.

Opinions are worth nothing. Reasons are worth everything. Which are theirs?