Life In 19x19

I'll start my response with a quote from an interview with Hyunjae Choi, the 2013 WAGC Winner from Korea:

At the moment I am still a student at Myongji University, where I am enrolled on the only course in the world for go, although I am currently taking leave from study. There I study go theory and issues in the cultural, historical and educational aspects of the game. My actual practice playing go is not done at college but rather at a famous go club, which I attend from six in the morning until nine at night almost every day.

First of all...sigh. What a luxury to have so much time! But of course he must sacrifice other things in life to get it, so it is not clear whether his situation is enviable on the whole.

So he studies a lot and plays an enormous amount. Whether the "go theory" he mentions bears any resemblance to that discussed in this thread is not known. Likely there is some overlap.

Roughly speaking, I think it's possible to think about different skills that contribute to go strength.

1. Calculation (i.e., reading ability).
2. Judgment.
3. Other, physical condition, time management, psychology, etc.

In calculation, I think there are some contributions from go theory. The best example I think of is in the area of capturing races, where knowing a little theory can save a lot of calculation. Texts in this area maybe don't propagate their entire level of detail down to most kyu players, because books like Richard Hunter's Counting Liberties and Winning Capturing Races can be extremely dry and can take some time to digest. Still, even books like Bozulich's The Second Book of Go, which is pretty popular, and Janice Kim's Learn to Play Go series touch on the theory of capturing races. The authors of the latter books are definitely targeting beginners, so they must believe even at that level some theory is useful, even if it is incomplete.

Another area related to calculation is endgame theory, regarding the sizes of moves and optimal move order, especially in the late endgame. I think there are some insights of practical value there. I'd never thought much about the values or corridors and rooms before reading the Miai Values List. Prior to that I would have played such moves rather randomly. Because such mistakes aren't punished in obvious way like life and death mistakes are, their cumulative effect is hard to expunge from one's game without some consciousness of theory. This is because teachers often ignore the endgame in their reviews, and so the student is often left with little feedback.

With respect to judgment, we need judgment because we can't calculate perfectly until maybe the very late endgame. IMO, this is hard, because anything in go can traded for anything else of equal or greater value, and often it seems like comparing apples and oranges. Is getting sente worth leaving this aji? Is black's thickess worth more than white's territory? And so on. Here, I think having a lot of examples helps. Players using tewari often compare game positions to known josekis, to see if there is some inefficiency that can be demonstrated. Here, it is harder to make concrete statements because the whole board position matters. RJ's books count outside stones, but not all outside stones are worth the same amount. A first order approximation is better than nothing, of course, but I think in the area of judging positions there is still a lot of work do.

In the 3rd area---other---the current go literature neglects this. If we thought of go as mainly a sport, it would be a huge emphasis, and there would be a lot discussion about practical things like mental endurance, managing time pressure, etc. Can theory help here? Maybe, but I think it would be a different kind of theory. I think one thing is the element of risk. How does one judge risk? The truth is there are some positions so complicated neither player can read them out. What do we do about this? Are there good ways to estimate the maximum risk of a fight that can't be read out? What do we know about when mistakes are more likely to occur and how can we use that information to improve our play?

the 3rd area---other---the current go literature neglects this.

Not true, unless you mean western go literature, but in any case there's a vast amount of relevant literature in chess.

But I've been quite surprised how sniffy many go players here are about learning from chess.

Bantari, although your descriptions of "your" versus "traditional" methods are roughly ok, I want to correct two important points:

- It need not be definitions and theorems, but instead it can also be principles and concepts of go theory. I.e., you portrait it as if it always were mathematics, while in reality it can be ordinary go theory language.
- A fixed order of first understanding textual theory and then examples is not prescribed. It is also possible to read both texts and examples, and repeat reading in flexible order them, until one understands.

***

John, whether making lists, stating principles, providing definitions etc. is study or research depends on how such knowledge is retrieved. If it is retrieved from already existing explicit (written or verbal) knowledge, it is "study". If it is retrieved from existing implicit context (such as game sequences), significantly improved from something existing, or found as something new, while having to make inventions, it is "research". There can be a grey area in between.

It is an ideal for research to precede it by checking all earlier findings. In practice, this cannot always be achieved, e.g., because of language hurdles. Classic science research has sometimes seen independent creations of same, new results. In go, one cannot check all games ever played before, just to test some new concept to each existing move. The closest I came WRT testing (where it was useful verification) was for the Japanese 2003 Rules and the Ko definition; I have checked each shape class I have had available. Such a high standard cannot be applied for all practically orientated research, if one wants to make much progress for the scope of practical application. Therefore, principle tend to be in the 90% or 95% class, only occasionally in the 100% class. If you don't want to call new principles "research results", then you'd better dismiss immediately all the traditional proverbs with their 55% "quality".

Surely, creativity is a must for research.

Since you even consider "market edge over [...] commercial rivals" and "we work because we just want to beat the guy who at present keeps beating us", you have not spent enough effort into understanding many results and their degree of usefulness. Compare the results of different authors with each other and with earlier go theory, and the success of application of such results! Then you can recognise what is research versus repetition of earlier knowledge, and appreciate the applied value of research. Meta-discussion about market shares or beating players do not let you recognise it.

pwaldron wrote:a formal definition of a nakade is clearly not necessary for go playing strength.

For that purpose, a formal definition is as unnecessary as classic problems with many stones forming a nakade string. What is necessary for go playing strength related to nakade is the ability to distinguish what is from what is not a nakade at least in every practically relevant case. This ability can, in particular, be acquired from the essence contained in (a draft of) a formal definition: can almost-fill, cannot partition, without seki. Players getting these aspects wrong have problems in improving their L+D playing strength.

Really, sometimes everybody here reminds me of the one recognising truth only after having formally defined it;)

John Fairbairn wrote:True research must also include testing and peer review, which I think RJ certainly strives for, but trying out someone else's study method does not qualify for research air miles. I think true research should also include consideration of all the prior art, something too often lacking here, as I no doubt tediously keep pointing out.

These are all good things, without which we'd have a lot less productive research, but I don't think peer review can count as part of a definition, unless we want to prohibit Perelman and Peirce from those who did research. Testing also is a strange case, since we could mean so many different things for different kinds of research (historical, mathematical, physical sciences...)

But certainly study is not generally research.

walpurgis wrote:If by more important go theory you mean playing games and getting familiar with common shapes and tsumego

No.

And judging by how practically all the top Europeans as well as top pros have learned, I even seem to be right.

I don't buy "all". Apart from that:
they had no chance, because the literature and teachers for them had forced them to learn that way, or autodidactically.

Why are you still "only" 4-5 dan and not competing for top 10 in EGC if go theory is so useful?

1) This has a chance of being considered a thread derailment.

2) I have answered this in detail many times before.

Never did I need the knowledge of formal nakade definition, perfect understanding of superko rules, or what "cannot necessarily permanently partition" means. (! Your theory just might kill people by overloading their brains)

Only those (like you) that pick specialised formal definitions to prove something very unrelated to other theory presented for application.

Then BAM, I only study formal go theory and behold! I become strong by knowing formal principles and definitions of go terms, rules, theory and not playing.

I have said it a hundred times, and say it again:

I do NOT recommed ONLY go theory. I DO recommend ALSO playing etc.

.. So, in 2nd paragraph you agree that theory doesn't really make people stronger, aye?

I see the contrary of you suggest.

If you mean (in the combination part) that one needs to know how to count right in mid/endgame (ko's) then that I agree with.

Fine. You are beginning to understand...

But theory the way you're advocating on these forums?

...that research theory (in its form expressed as such) is not the form of theory I am teaching (outside these forums) for the purpose of application.

Research theory, where it exists and is correct, can be the basis for developing good applied teaching. Readers and pupils need not be aware of the research itself. But a good teacher can improve his teaching by knowing and applying the research results in his mind and refining their form suitably for the teaching.

You'll be hard pressed to find even one person who has directly become stronger by theory research

This misses the point. Learners need not use it DIRECTLY, as soon as didactically refined versions are available.

Why is endgame CGT so unpopular among players? Because didactic refinements are not suitable yet. One still needs too much direct reading of original research.

You make bold claims, but don't show any proof to back it up.

My improvement from 9k to 3d:
I have described this earlier and elsewhere, providing sufficient proof.

Research more than a tiny area:
This requires extensive discussion of contents of papers and books of various authors. It is something for the Book forum or the unmoderated rec.games.go. We can discuss it later this year, but not know, because at the moment I lack the necessary time for such a presumably long discussion. For the moment, you can refer to parts of my contribution to see why "tiny" is the most inappropriate, especially when you recall that there are also other researchers:

viewtopic.php?f=48&t=7791&hilit=research+jasiek

Maybe the subject of this topic needs to be a bit more specific.

Most of the time I see "Go Research", it's discussing how professionals and others get together and study various lines of play and look for new moves and variations to get an edge. There is a lot of research in Go and most of it is practical.

What Robert discusses is research on various theories which seems to focus on definitions and formulas.

oren, both kinds are go research, although they differ in quite a few aspects.

RobertJasiek wrote:oren, both kinds are go research, although they differ in quite a few aspects.

I agree. I was pointing out the same thing.

I was just looking at the subject and noting there is a lot of applied value in Go Research that professionals do. You can see it every day.

What most of this discussion is about is applied value of theoretical research (which sounds kind of funny).

RobertJasiek wrote:What then do you call, e.g., my increment from 9k to 1k in about half a year? It was so quick, because I studied much go theory (and only a few problems) during that period. The major obstacle during that period was a necessity to learn only from examples about such topics, for which then go theory was hardly available at all.

walpurgis wrote:1k is not strong. And you didn't get there with just theory, as you said. Besides, anecdotes don't count for much.

Of course, 1 kyu is strong. Players of average ability and more than average interest can probably reach single digit kyu rank. There are some 40 levels of go ranking, and 1 kyu is in the upper levels.

walpurgis wrote:You {Robert Jasiek} just might be the unique gem who learns that way (or claims to), but you have been blinded by all your years of theorycrafting. Your general player has little use for the kind of theory you vouch for. And judging by how practically all the top Europeans as well as top pros have learned, I even seem to be right.

Robert is not the only player who has benefited from learning go theory. After all, theory condenses knowledge, so that a few concepts can take the place of much trial and error experience. In the 1970s Bruce Wilcox demonstrated the utility of learning theory through his teaching it in conjunction with US Go Congresses (IIRC). Many kyu players advanced two or more stones in a week or two. From my own experience I know that a single idea can be worth a stone at the dan level, and even four stones at the kyu level.

walpurgis wrote:
Why are you still "only" 4-5 dan and not competing for top 10 in EGC if go theory is so useful?

An ad hominem attack indicates the weakness of your argument.

Go pedagogy is not very well developed, mainly because, IMO, it has focused on talented individuals who devoted themselves to go at an early age. What has worked for them is not necessarily what works for amateurs who learn as adults. In other areas of skill and knowledge there is evidence that adults learn best by learning concepts first. Why should go be any different? But the thing is, most of the evidence we have about teaching and learning go is anecdotal, not the result of carefully planned and controlled research on many subjects. Anyone who thinks that they know the best way to teach and learn go is probably wrong. Moi, I think that there is no one best way.

RobertJasiek wrote:Bantari, although your descriptions of "your" versus "traditional" methods are roughly ok, I want to correct two important points:

- It need not be definitions and theorems, but instead it can also be principles and concepts of go theory. I.e., you portrait it as if it always were mathematics, while in reality it can be ordinary go theory language.
- A fixed order of first understanding textual theory and then examples is not prescribed. It is also possible to read both texts and examples, and repeat reading in flexible order them, until one understands.

Hmm... Interesting.

Here is what I believe: for sporting purposes, the traditional method is better because the knowledge, being in some/large part subconscious, is more efficient in execution, even if it might be less efficient in acquisition. So, it ultimately creates better players even if the learning might under circumstances be slower. Just like in Golf - practicing a swing and making you muscles 'remember' it 'subconsciously' by constant repetition makes for an overall better player than the one who has to consciously think about the mechanics of every move.

From what I see you are saying, basically, the more formal approach you practice might help create more efficient ways to eventually get to the same results by improving on the methods on training the subconscious? For example, a research into a nakade, including formal definitions and enumeration, might help organize the methodology of presenting nakade examples more efficiently or methodically, and this can speed up the process. Just like research into the physics of a swing in Golf might lead to designing better club or better ways of training your muscle 'memory' and thus create better players.

In a sense, it is the separation of theory and applications which I might not have considered here - since it is contained within one and the same player. Very interesting thought, although certainly nothing new (just my blindness, heh, which is also certainly not new.)

Well, we will have to see how it goes.
I am still not totally convinced, by far, but there is something new for me to think about, in this context. Thanks.

Bantari wrote:the more formal approach you practice might help create more efficient ways to eventually get to the same results by improving on the methods on training the subconscious?

Or conscious.

Bill Spight wrote:Robert is not the only player who has benefited from learning go theory.

This is certainly true, but we shouldn't forget that there are a lot of go players who haven't benefited at all from go theory, those who have been stuck somewhere in the SDK area (some even in the DDK area). Every method seems to work for some people, for some other people more than one methods works - these are the players with fast improvement - and again for other players no method works well enough to get beyond their barrier.

karaklis wrote:there are a lot of go players who haven't benefited at all from go theory, those who have been stuck somewhere in the SDK area (some even in the DDK area).

Being stuck at a level is not the same as not having profited from go theory at all. I think that, e.g., every 10k knows that "One does not kill an already dead group.". This is go theory, from which they have benefitted to become 10k. Part of the 10 kyus would improve if only they always applied the principle. (The other part does already apply this aspect of go theory correctly in all their games.)

for other players no method works well enough to get beyond their barrier.

This is often said, but actually many players can try a) to spend (much) more time on go and go study and b) apply other methods they have not been applying at all or enough (such as regularly asking stronger players for advice or getting teachers, and at best a variety of teachers emphasising different teaching methods).

(Of course, there are indeed players having reached their final barrier, where any amount of extra effort is lost due to a too flat logarithmic ratio of additional increment per amount of effort.)

Apart from the applied values of research mentioned so far, there are also the following important aspects:

Research expressing explicitly earlier implicit information:
- A player missing some particular knowledge need not study arbitrary amounts of arbitrary implicit information in the hope of filling the right gaps accidentally soon and need not rely on traditional teachers providing him with examples conveying such implicit information. Instead the player can study the particular knowledge immediately. If he sees keywords referring to the knowledge, they tell him nothing, and it is clear that they are not just different words for something he already knows, he knows that he has a related knowledge gap, maybe some he was not even aware of at all.
- A player is given a choice of learning methods, where previously there was only 'learning from implicit information'. This is so especially for dan players, for whom the ratio of explicit versus implicit knowledge available is still low. The research raises the ratio by providing more explicit knowledge.
- Different players can understand each other much more easily, when they, thanks to research, can use both text and diagrams for communicating knowledge.
- Different players can find out more easily whether they share the same or different ideas about a topic. Such understanding is not limited any more to only small selections of specific examples or vague references to books or databases somehow containing specific examples.
- More new knowledge can be generated from general knowledge provided by research than from knowledge represented by a few specific examples of diagrams or moves in diagrams.

Research expressing new knowledge:
- See above.
- Players acquiring this knowledge get an advantage on a topic over other players.