Dan Principles

[RobertJasiek]

Since discussion is becoming OT for the other thread, I continue it in this new thread. Citation reference:
viewtopic.php?p=206592#p206592

anyone can memorize basic principles, strategies, proverbs, joseki's and so on, but that doing so does not make you a dan level player. Thus, nothing stops a kyu player from learning a whole bunch of theoretical knowledge and still losing miserably to a stronger player if his reading skills aren't in place. Of course these two things are complementary and no truly strong player neglects either. However, I think it's possible to become a dan level player without learning alot of Go theory by simply getting strong at reading and then learning by playing. You see that type of player alot on wbaduk and tygem and they can be fairly strong. [...]

I think that most if not all dan level principles and strategies are merely extensions of basic principles and strategies that even a 10kyu could learn. That, along with mistake reduction and reading ability makes one a dan player, not partaking of the mystical dan level principle soup distilled in book form, should such a source exist. [...] I don't think any of the principles themselves can be ranked [...] As one becomes stronger, one can apply the principle more and more and with greater skill. [...] the principles aren't ranked, but that it is the skill with which we apply the principles that make them 'dan level'.

My point is that in theory a kyu player could learn all theoretical knowledge of go but still lose because she doesn't possess enough raw reading power and tesuji skill to carry these principles out. Learning go theory and principles is a great booster, but I believe that cultivating the raw power of calculation is the most important thing and that everything else is auxilliary.

As soon as principles (or other theory) are clearly stated (usually in books), then in principle everybody can learn the theory. Even beginners can learn such theory but can often be expected to have greater difficulty in applying it, e.g., if the theory presumes reading.

However, everybody can only learn the theory that he has access to. Essentially, theory is only available for everybody in writing if it is a) new or b) previously never or rarely spread verbally but almost only existing in subconscious knowledge of usually strong players.

Thus there are the following kinds of dan principles:

1. Even if written down clearly, their application is too difficult for kyus.

E.g., a principle, such as on p. 257 in Positional Judgement 2 - Dynamics, about long, complex fights also suggests to iterate reading until judgement is unambiguous.

2. The principle exists (almost) only in strong players' minds.

E.g., I had never heard or seen clearly stated the principle "Choose the most valuable fight or else one of equally valuable fights.". As soon as I wrote it down on p. 224 in Fighting Fundamentals, it became, as I'd imagine, an obvious principle for every reader. It is obvious only once stated. This is so with very many fundamental things. Before I (re)discovered the principle for myself, I was rather aimless in fights because I would often not know in which part of the board to play next when there were multiple fights. The principle can be understood by everybody but applying it requires more: an understanding of a value of each fight. My invention of the concept of 'value of a fighting region' (p. 227) stated clearly what previously every professional player had almost always applied correctly but apparently only subconsciously at each turn and what had been a "mystery" for most amateur players because they would not know the concept, not even subconsciously. Never before my invention had I read or heard about the concept. The closest related remarks had been professionals using mysterious words, such as "follow the natural flow of the game" or "appreciate the beauty of shape [construction]".

As the examples show, "such" sources exist for some principles.

There is much dan principle knowledge that is still mostly subconscious, that all 6d+ get right in their games but almost all kyus are not aware of. Of dan knowledge for 1d - 5d, much is also known to some kyus but clearly fewer than dans.

Principles can be assigned levels from which correct application can be expected, at least their explicit knowledge is given or - for still only verbal principles - at least subconscious knowledge is given. There are 'dan principles' for which the lower threshold is in the dan range.

EDITS

[Babelardus]

...an obvious principle for every reader. It is obvious only once stated. This is so with very many fundamental things.

I've been reading your First Fundamentals book. The principles are rather obvious:

"Play in the big spaces"
"Choose the largest gap"
"Don't make an alive group more alive"
"Don't make a dead group more dead"
"Seek life in the most open spaces"
"Don't connect already connected stones"

The first thought that flashed through my head was: Well... *duh* ... Captain Obvious. Going by the example material I thought this book was better

Then I went over some games I had saved, and noticed that the principles aren't actually that obvious while playing and are often violated. Darn... he *is* right. I don't choose the biggest space and I do sometimes play in a small gap while there's a bigger one available. I do try to make life by playing toward the smaller direction. Also, I *am* lazy and just add a stone to an alive or dead group to make 100% it's alive or dead so I don't have to read it all out. I also sometimes connect connected stones because I'm either too unskilled or too lazy to read out if they are connected even when cut.

So yes, principles are very obvious when stated and explained by the examples; so obvious that it even sometimes feels insulting. However, it looks like as if you (or at least, I) need to be bashed over the head with them before you/I actually start paying attention to them consciously.

I've been trying to implement some of the principles mentioned in the beginning of the book, against a few of the stronger engines I use, and while I still lose, the difference in points has already decreased by a noticeable margin, on average

It's one of those "it's so hilariously obvious... when you know it"-cases.

[RobertJasiek]

Every DDK (and many players a bit stronger) frequently violates some or all of these and similar "obvious" principles. Players 5k - 1k occasionally violate them, and at their level it still hurts because every mistake of the remaining fewer major mistakes gets a greater impact on the game.

[EdLee]

principles are very obvious when stated and explained by the examples...
However, it looks like as if...at least I... need to be bashed over the head with them...

Hi Babelardus,

As often is the case, it's non-trivial for people to understand that understanding is a continuum:

• Never heard it, never thought about it ( one level );
• Heard it, doesn't get it ( another level );
• Heard it, thinks understand it ( another level);
• Heard it, forgets about it ( another level );
• Remembers, but applies it wrong most of the time ( another level );
• Applies it OK sometimes ( another level );
• Applies it ~50% of the time;
• Applies it most of the time;
• Can help others understand it ( at above various levels )

( Just because we get bashed on the head, and we consciously think about something, it doesn't mean we get it 100% ... We can still range from 1% to 50%. )

Where are we on various knowledge items ?

[Magicwand]

"Play in the big spaces"

disagree. Urgent spaces first

"Choose the largest gap"

??? never heard that one

"Don't make an alive group more alive"

some cases you might have to

"Don't make a dead group more dead"

this should be Dont try to save alread dead group

"Seek life in the most open spaces"

case by case. sometimes closed space is better

"Don't connect already connected stones"

depends on how you define connection.

[EdLee]

Hi Magicwand,

These are only guidelines:

• Their applications depend on the exact board position;
• They are not perfect. Not only are they not perfect, they're far from perfect;
• There are practically infinite exceptions.

These are generally understood -- that's the nature of guidelines/proverbs/principles.

Play in the big spaces disagree. Urgent spaces first

See above list.

Choose the largest gap ??? never heard that one

If we look at your games, I'm sure you follow this in some cases. I guarantee it.

Don't make an alive group more alive
some cases you might have to

See above list.

Don't make a dead group more dead
this should be Dont try to save alread dead group

These are two different ideas.

Seek life in the most open spacescase by case. sometimes closed space is better

See above list.

Don't connect already connected stones
depends on how you define connection.

See above list.

[RobertJasiek]

Magicwand, what are "urgent spaces"? There is a proverb / weak principle "Play urgent before big moves". But what is "urgent"? Usually, it is urgent to defend (sometimes also to attack) a weak important group or to stabilise unstable important shape. Usually, people suggesting urgency mention the former but forget to also explain the latter. (Some explanation of stability: http://home.snafu.de/jasiek/Joseki_2_Sample.pdf ) Urgency is a weak concept because it in it itself does not contain an obvious comparison to size, as in the size of a move in a big space. To always apply the proverb correctly, such a comparison is needed (whenever the decision is not obvious) and can be achieved by assessing urgent moves in terms of size or value. Values for urgent moves or big moves are multi-dimensional: they include aspects of territory change, influence change and change of other aspects of dynamic positional judgement. The proverb is designed for intermediate players who cannot always do a good value assessment of urgent moves yet. Urgency is a concept not so useful for beginners who benefit much more from the more basic concepts. Therefore, First Fundamentals avoids "urgency" and instead uses basic concepts, as e.g. in the principles "Notice unsettled groups" (p.28), "Connect your important strings" (p. 141), "Defend your unsettled important groups" (p.163). These a beginner can understand without having to know what urgency is.

Maybe you have not heard "Choose the largest gap". It is a special, simpler case of "Choose the bigger space" (p. 12) and an implication of this. Surely you do apply it, though:) When all relevant spaces are just gaps (such as along the edges), it is sufficient to compare their 1-dimensional width instead of their 2-dimensional size.

"Don't make an alive group more alive" and "Don't make a dead group more dead" are Babelardus's paraphrasing of similar principles in the book, such as "Do not defend the life of an unconditionally living group" and "Do not kill an already dead group" (p. 44). And yes, there are exceptions for strong players; e.g., it can sometimes be correct to play a negative ko threat defending the life of an independently live group. However, First Fundamentals is a book for beginners, not a book for strong players; so the beginners are not flooded with the exceptions for strong players.

"Don[']t try to save [an] alread[y] dead group" is a related, also useful principle, but not the same as those before.

"Seek life in the most open spaces" is Babelardus's invention and a misunderstanding of maybe the principle "Move to the wider direction" (p. 17). Or maybe it is the consequence of trying to write down principles read in a book from memory without double-checking the wording in the book.

Connection is defined in books for intermediate and dan players. Maybe beginners need not learn "indirect connection" yet as long as they make more basic connection mistakes.

[Krama]

A wise pro once said when asked what the difference between eastern and western is.

Eastern player plays, reviews, studies.

Western player talks about all of the above.

Western player always looks for a shortcut.

This is why western player is weak.

[RobertJasiek]

The pro was not wise but sought an excuse for not explaining well exactly what to study. Talking about plays, aspects of review and things to study is an advantage (only talking instead of playing, reviewing and studying is bad OC). Shortcuts are an advantage wherever applicable (using shortcuts to avoid everything for that there is none is bad OC).

Western players are weak because a) there are only few so that only few strong can emerge in principle, b) there is little availability to regularly play against strong players in the real world, c) many Western players start late so that the time for study is short until they reach the necessities of job and family, d) most advanced go knowledge is still only verbal and contacts to that verbal knowledge are limited due to the limited access to strong players, e) some useful knowledge is hard to access due to the Asian language barrier and f) for players with the great ability to learn from explicit written knowledge, the Asian literature, which focuses on examples with implicit knowledge, offers too little.

This is changing but very slowly because a tiny number of Western players create explicit written knowledge. After some more decades, this knowledge will prevail because of offering shortcuts.

[Bill Spight]

A wise pro once said when asked what the difference between eastern and western is.

Eastern player plays, reviews, studies.

Western player talks about all of the above.

Western player always looks for a shortcut.

This is why western player is weak.

Wise? More like prejudiced.

[Bill Spight]

And about shortcuts: let's get real.

Go proverbs are shortcuts. Tewari is a shortcut. Counting liberties is a shortcut. Go is full of shortcuts.

[Bantari]

What I wonder is the longevity of any such principles as you grow stronger. I have always thought of such approach as a "crutch" - necessary for a while, but with enough therapy (=skill) eventually discarded. Let me explain what I mean.

As a beginner, I have been told something crucial but trivial, like "don't make empty triangles." This was the "principle”.

As I grew stronger, I realized that there are so many exceptions... in virtually every game I was forced to either play empty triangle or be prepared to play it as part of some hypothetical sequence. Of course, this was mostly due to my mistakes, often to my opponent's mistakes, but it was so nonetheless. I have even seen empty triangles in pro play. So I realized how useless the initial "principle" was.

So I exchanged it for a slightly more generic "principle", maybe something like "try to maximize your liberties" or some such. As with the initial one, I quickly grew strong enough to see how restrictive that was, and how many exceptions I have to constantly juggle... So I needed to replace it again, this time with something even more generic, maybe something like "strive for efficiency of your stones."

This process was going on for a while, until I realized that there is very little room in my play for "principles." Thinking about it deeply, what I do now is a simple two-step process which does not rely on much except of my feeling/experience/mood and my reading. It goes like this:

Step 1: Given a game or position, I first try to determine the overall approach - what is the underlying metric which I deem the most important in this case. I think of it as something akin to "global strategy", and it can be based and determined by all kinds of factors - the position itself, the knowledge of myself, the knowledge of my opponent, my mood, and whatnot. Often, especially later in the game, the play is already rolling along a specific path and there is much less freedom to decide on the metric.

Step 2: Having determined/decided the metric I place importance on, I read out which global/local move maximizes this metric.

Of course, we can argue if there are any principles which help me determine the appropriate metric, or if there are any principles which can help me decide on the number of candidate moves for subsequent reading. Thinking about it deeply yet again, I think that they might be, but I am not sure how meaningful they are. They are certainly not often conscious. Let me explain again.

For each move I make (or almost each move) - I can name a principle or a set of principles which guided this move - or which can be applied to this move. When I think about it after the fact. In other words - I can reasonably well explain my moves - after they are made - in terms of "principles" and "theory" and all the nice definitions and theorems we have. However - and this is important - I am not sure I could have derived my moves this way. It seems to me that most of the time, this is a one-way-only relation. Principles can ultimately explain moves, but you derive moves not by applying principles but by reading sequences.

It is the move itself which determined the principles applied, not the other way around. In other words, the move I make decides which of the many principles I could have applied are the important ones in the given position. Or, at least, which are the important ones for me at the moment. This is how I can bring order into the theory I know. The other way - by taking all the "principles" which might reasonably apply to the given positions, and using them to determine the move I make - I was never really able to play. I was never able to decide on which principles take precedence before reading out a whole lot and ultimately deciding on my move - based on this reading, not on principles. Maybe I though differently as a DDK, don;t remember anymore... But certainly not after I hit the dan levels...

It such light, to me, “principles” were something useful for teaching, not really useful for playing. But thinking about how silly this sounds, I wonder… Maybe teaching also should avoid “principles”, or at least embellish them with a whole lot of disclaimers. But this might make them too complex and confusing, so i am really not sure here. This might be a separate “teaching discussion”, though. Back to the subject.

So now I wonder - how does the principles in Robert's book(s) stack up to this kind of process. Are they being refined as students grow stronger? Are they being replaced? Are they eventually being abandoned altogether? Robert - you yourself indicated (to Magicwand) that at least some of your principles are for weaker players only, and as you grow stronger you have to find ways to accommodate more and more exceptions.

And this brings me to the actual topic of this thread: Dan Principles. Are there any? Ones which help derive moves, or only ones which help explain moves to weaker players? How about high-dan players? Is there a set of more advanced principles, or just the same ones but backed with more knowledge of exceptions and better reading? Are there "Pro Principles"? Or, more general - is there a level at which any such concepts are meaningless? Is this why pros, most often, when explaining their moves, give sequences rather than principles?

Based what I wrote above you probably suspect which was I think the answer is leaning. But I would like to hear other opinions.

The bottom line is, I think, that “principles” might be useful at times, but the real skill and strength lies not in knowing and understanding principles, but in knowing and understanding all the exceptions to principles and in knowing when the principles are useless and you just have to read deep. Which is most/all of the time when you reach mid-dan. So where does this leave us and our theory?…

[RobertJasiek]

As explained before, principles have different qualities of frequency of correctness between slightly above 50% and literally 100%. 99.9% to 100% principles are also for professional players.

I use principles (and strategic planning, reading etc.) to find the next move. I also teach what I use.

Below 100% principles have exceptions. This is often obvious but when this is less obvious and still important, I write a "Usually," in a principle. For some principles, I list the major exceptions but for most principles written statements of (maybe verbally known) exceptions is open research.

Another important topic is relative priority of principles for the sake of solving conflicts of several principles. Higher order theory overrides lower order contradictions. One such higher order resolution is to use reading in case of doubt. However, where principles alone provide the answer, reading becomes superfluous.

So far there is only limited documentation of which principles are for which ranks and have which frequency of correctness. This must be improved in the future. Principles for stronger players accompany refinement of principles for wesker players other than the 100% principles for all players.

[Bantari]

As explained before, principles have different qualities of frequency of correctness between slightly above 50% and literally 100%. 99.9% to 100% principles are also for professional players.

I use principles (and strategic planning, reading etc.) to find the next move. I also teach what I use.

Below 100% principles have exceptions. This is often obvious but when this is less obvious and still important, I write a "Usually," in a principle. For some principles, I list the major exceptions but for most principles written statements of (maybe verbally known) exceptions is open research.

Another important topic is relative priority of principles for the sake of solving conflicts of several principles. Higher order theory overrides lower order contradictions. One such higher order resolution is to use reading in case of doubt. However, where principles alone provide the answer, reading becomes superfluous.

So far there is only limited documentation of which principles are for which ranks and have which frequency of correctness. This must be improved in the future. Principles for stronger players accompany refinement of principles for wesker players other than the 100% principles for all players.

Can you give some examples of principles with 100% correctness?

Also, can you give examples of situations where principles alone can determine the next move, without the need for any reading?

And maybe, while we are at it, some examples of how can you determine relative priority of principles without any reading?

I am interested because from my experience, I really cannot think of any. Unless we point to principles so generic they always apply, like "make good moves."

[RobertJasiek]

100% principles:

"In class 1 types 1 to 3 semeais with equal numbers of exclusive approach liberties, the inside liberties are ignored for determining the status by comparing the group's numbers of liberties."

(See Capturing Races 1, pp. 62f, 183. Note the seki requirement of at least 1 or 2 inside liberties.)

"Under rules resulting in a scored game, an optimal move of a player maximises the score difference in his favour."

(The principle does not necessarily apply under rules with exceptional results, such as 'no result'.)

***

Principles without reading can determine the next move:

"If all available valuable moves are gote options, play one of the most valuable ones."

(A 'gote option' gains X points for a moving player but zero points for the moving opponent. See also Positional Judgement 1, p. 69ff.)

"Kill a group surrounded by adjacent pass-alive opposing stones, with one lake and one vital point by occupying it."

(A 'lake' is a connected part of a group's eyespace.)

***

Determining relative priority of principles without any reading:

Suppose we have principle 1 applicable to one corner, principle 2 applicable to another, independent corner and principle 3 relating principles 1 and 2 to each other. Principle 3 has the higher priority.

E.g., principle 1 "Call X the per move value of killing a group", principle 2 "Call Y the per move value of defending the independent life of another group", let the two groups be mutually independent, principle 3 "Among basic gote endgames, play one of the largest", application "According to principle 3 and the mutual independence of the two groups allowing us to speak of basic gote endgames, we dissolve the ambiguity of the applications of principles 1 and 2 by applying the higher priority principle 3: 'Choose X if X>Y, choose Y if Y>X, choose either X or Y if X=Y'".

EDITS: corrections of the 100% principles.