Life In 19x19

Segoe Kensaku, who as I keep trying to remind people must have been the top player in the world for a time, was one of the more lucid go writers. I was just looking at a very, very long series he wrote on josekis (30 pages in each of several magazines), and I found a couple of insightful things I thought I'd share.

One was his categorisation of the types of joseki, based on the viewpoint of the player selecting the joseki:

1. Those stressing profit
2. Those stressing solidity
3. Those stressing simplicity
4. Those stressing outside influence
5. Those stressing attack
6. Those stressing complications

I don't think many of us would have come up with the same sort of list, and differentiating between profit and solidity might surprise some.

Segoe adds that types 1 to 3 would be most useful in handicap games, especially with 4 stones and over. Types 4, 5 and 6 would be favoured by power players like Honinbo Jowa, but one thing that made Honinbo Shusaku stand out was that he was able to keep his josekis in balance and was good at integrating all these types.

One of the most interesting bits of advice, however, concerned the vexing case where your opponent does not cooperate and you end up in a long and/or difficult joseki. Segoe says that because that does happen you need to learn to cope by studying one such joseki in depth to glean the basic coping strategies. The one he recommends is takamoku. He doesn't say specifically why, but it seems easy to infer the reason: this stone is placed far enough away from the corner to ensure that, whilst the opponent can mess around to a large degree (he has a lot of corner to play with!), it is virtually impossible for him to change the basic outside-inside relationship. Moreover, Segoe recommends studying takamoku on its own and also White's ploys when Black has, in addition to takamoku on D5, an extension stone on C10. However, in the latter case he also adds a black stone at the centre point. Presumably this is to denote ladders favour Black (it's relatively rare for a player to choose takamoku when the ladders do not work for him).

Two opponents not cooperating reminds me of this game.

http://eidogo.com/#DOsdgv:0,9

You would think after all their years, Cho and Kobayashi would have come to some sort of under standing, but they barely seem to acknowledge each others existance.

John Fairbairn wrote: Presumably this is to denote ladders favour Black (it's relatively rare for a player to choose takamoku when the ladders do not work for him).

I didn't know this. In ama games you often see it played with the opposing corner still open. I guess that corner should become a priority then.

Also, is there an English language resource that discusses the pre-existing stone at C10?

Good stuff. Thanks, John.

wineandgolover wrote:

John Fairbairn wrote: Presumably this is to denote ladders favour Black (it's relatively rare for a player to choose takamoku when the ladders do not work for him).

I didn't know this. In ama games you often see it played with the opposing corner still open. I guess that corner should become a priority then.

Also, is there an English language resource that discusses the pre-existing stone at C10?

Good stuff. Thanks, John.

Surprisingly, eidogo returns more than one tengen takamoku fuseki.

http://eidogo.com/#search:nw:14x14:.46x.88x.60

John Fairbairn wrote:Segoe Kensaku [...]
One was his categorisation of the types of joseki, based on the viewpoint of the player selecting the joseki:

1. Those stressing profit
2. Those stressing solidity
3. Those stressing simplicity
4. Those stressing outside influence
5. Those stressing attack
6. Those stressing complications

When regarding both players' views, I have [5] suggested this functional classification (and another, evaluation-orientated classification):

- finished thick settling
- aiming at thick settling
- quick settling
- cutting while moving to the center
- territory versus side moyo
- territory versus influence
- mutual running fight
- cut and fight
- lean and attack
- simple forcing
- two simple unsettled groups
- two simple groups, of which one is unsettled
- two simple settled groups
- emergency corner life after playing elsewhere
- attach and chase after playing elsewhere
- tactically complex fight

Applying this to Segeo's list, his list is too simplistic despite being for only one player's view.

RobertJasiek wrote:Applying this to Segoe's list, his list is too simplistic despite being for only one player's view.

Or yours may be a bit too complex.

oren wrote:Or yours may be a bit too complex.

It is not too complex, but it is appropriate if one takes the both players' view. It is appropriate because a) josekis for every type exist, b) every type must be known and understood because of behaving differently, c) there are strategic choices and generally applicable objectives (expressed in principles) specific to every type. Knowing every type and its behaviour allows development of good strategy in the positional context.

Taking the both players' view suffices, while taking only one player's view yields incomplete information. The latter can be useful for some applications, but nevertheless the former is also needed.

RobertJasiek wrote:

John Fairbairn wrote:Segoe Kensaku [...]
One was his categorisation of the types of joseki, based on the viewpoint of the player selecting the joseki:

1. Those stressing profit
2. Those stressing solidity
3. Those stressing simplicity
4. Those stressing outside influence
5. Those stressing attack
6. Those stressing complications

When regarding both players' views, I have [5] suggested this functional classification (and another, evaluation-orientated classification):

- finished thick settling
- aiming at thick settling
- quick settling

Applying this to Segeo's list, his list is too simplistic despite being for only one player's view.

Quick settling does seem like an omission to me, maybe he considers it a function of attacking, simplicity, or profit?
Other approaches I would have added to the list would be ko fight and tenuki. Though ko fight could be some combination of thickness and complications, and tenuki could be some combination of attacking and complications. These may just be edge cases though.

John Fairbairn wrote:Segoe Kensaku, who as I keep trying to remind people must have been the top player in the world for a time, was one of the more lucid go writers.
....

Mr Fernando Aguilar says that the book that changed his view of the game was the "Go proverbs illustrated" by Segoe. And he treasures it as his preferred book.

I managed to buy one copy of the first edition, and I can't judge the accuracy of the translation, but even thru that filter, I must say that the content of the book reflexed the ideas of somebody who really understood what Go is about, and also loved Go.

Sure, you can make finer distinctions. But can you use them?

Even six categories may be too much for the non-pro player. I suspect that the ability to use such categories depends on one's strength: the stronger you are, the finer the discriminations you can make.

I remember when I was a DDK. I had three categories:
1) I die
2) He dies
3) We both live

Joaz Banbeck wrote:Sure, you can make finer distinctions. But can you use them?

Even six categories may be too much for the non-pro player. I suspect that the ability to use such categories depends on one's strength: the stronger you are, the finer the discriminations you can make.

I remember when I was a DDK. I had three categories:
1) I die
2) He dies
3) We both live

Had you been a DDK for longer you may have been able to find the ever-so-elusive 4) We both die

Joaz, first learning the 16 types greatly eases then learning by understanding dozens, hundreds or thousands of josekis. Proceeding vice versa wastes great unnecessary efforts, because one still lacks the related basic functional understanding.

Obviously, beginners knowing fewer than 16 josekis do not need types yet. SDK players starting learning of many josekis should first learn the types, so that immediately they recognise similar josekis and do not need to learn each afresh.

When learning my first 1000 josekis, I had almost no idea of types yet. This made the learning unnecessarily hard. Only later, on discovering the types, I realised just how much effort I could have avoided if the types had already been available before. The same can be said about the general strategic knowledge associated to the types.

Your advertising of 6 types as too many for the non-pro players is one of the most counter-productive advices one could give to players intending to learn and understand many josekis. Much less counter-productive would be an advice to learn only the types (and their natures), no josekis and develop joseki-like sequences ad hoc during one's games.

Here is an attempt to complete Segeo's view for one player

1. Those stressing profit
2. Those stressing solidity
3. Those stressing simplicity
4. Those stressing outside influence
5. Those stressing attack
6. Those stressing complications

given my types for both players' combined view

- finished thick settling
- aiming at thick settling
- quick settling
- cutting while moving to the center
- territory versus side moyo
- territory versus influence
- mutual running fight
- cut and fight
- lean and attack
- simple forcing
- two simple unsettled groups
- two simple groups, of which one is unsettled
- two simple settled groups
- emergency corner life after playing elsewhere
- attach and chase after playing elsewhere
- tactically complex fight

Completed and corrected types for one player (joseki stressing mainly the mentioned aspects):

1. profit
2. solidity
3. simplicity
4. outside influence
5. fight
6. complexity
7. solidity and outside influence
8. options
9. profit and outside influence
10. playing elsewhere

Explanations why the extra types for one player are needed or why types 1.-6. can be used:

- 7. is needed for finished thick settling, because solidity and outside influence are equally important, because the solidity of the outside thickness creates its outside influence.
- 8. is needed for aiming at thick settling, because 1.-6. do not characterise it.
- 2. can be used to explain the quickly settling player's group of 'quick settling' and the cutting player's group of 'cutting while moving to the center', however, these are two subtypes of solidity, which differ from the solidity in 7. IOW, Segeo pretends his list to be short by unifying related, but different things in one type.
- 9. is needed for the moyo (and its player), so that it is understood that the moyo of 'territory versus side moyo' aims at both territory and outside influence, while the influence of "territory versus influence' can forgo profit and solidity.
- 5. must be 'fight' (instead of 'attack'), because 'mutual running fight' and 'cut and fight' allow attack, defense, and (by the same player!) combined attack and defense. Again, Segeo pretends his list to be short, because he hides subtypes that are different types of fights.
- 3. can be applied to 'simple forcing', 'two simple unsettled groups', 'two simple groups, of which one is unsettled' and 'two simple settled groups'. Once more, Segeo hides these subtypes.
- 10. is needed for 'emergency corner life after playing elsewhere' and 'attach and chase after playing elsewhere'.
- 6. applies to 'tactically complex fight'. Complexity is a slightly better chosen word than complications, because the latter occur also for 5.

One might add 'ko' as a type, but my preference is to treat the topic ko entirely separate, so that one need not discuss the many possible subtypes of kos.

The completed Segeo-style list for the view of one player has 10 types. Now, one might make the rash implication that, for both players' view, this would result in 10 * 10 = 100 types... Luckily, my list of 16 types is much shorter, because most combinations of a type for one player and a type for the opponent do not occur in practice.

Where one might have fallen into the trap of believing Segeo that 6 types for one player would be enough (but 10 are needed), it turns out that considering both players' view lets his approach create too many types. My 16 as the number of types is actually the smaller number! (This is so even for the naive 6 * 6 = 36, when believing Segeo too much.)

Careful thinking pays off. In my case, I simply studied every joseki (roughly the Ishida) to identify all occurring functional types, and so I found the 16 types for both players' view.

(For my other classification, 6 basic evaluation types suffice. In this respect, evaluation is simpler than strategic functionality.)

RobertJasiek wrote:8. options

- 8. is needed for aiming at thick settling, because 1.-6. do not characterise it.

Options, would those be like, people who like probes? I think that is an omission as well.

The most famous options are called 'miai', but options can occur alone or in greater number. My definition is:

"Options are strategic choices kept in reserve."

Test moves (probes) are just that; the possibility for them is not necessarily an 'option', but can sometimes be.

Please note that a classifying list of functional joseki types is NOT a list of all strategic concepts (such as 'probes' or 'options'). It is a coincidence that 'options' can serve as a Segeo-like joseki type or, more generally, be a strategic concept. (Other strategic concepts also occur: fight, leaning etc. You have suggested ko, but I prefer to treat it as a strategic concept; as a joseki type, it would be an overkill, IMO.)