Forum for Positional Judgment

Not WME [#2624]

Back to forum     Back to page

New reply

 
reply
Dieter: Not WME (2011-09-03 12:35) [#8708]

The last edit was not a WME.

Introducing concepts which are unknown but to the author, like 1-territory, are not WME but an overhaul and a suggestion for improvement. Having good novel ideas is one thing, presenting them as if they were already established knowledge is another.

For me, having played for 15 years now, the page has become incomprehensible. I'm ready to take the blame for not having introduced myself to every author's terminology, but that's about it.

reply
RobertJasiek: ((no subject)) (2011-09-03 14:45) [#8709]

Ok, point taken. I have corrected the RecentChanges comment from 'WME' to 'major update'.

If you don't know concepts, then they can be explained. If you are worried about the impression that new concepts appear as if they were old concepts, then footnotes can be added with a remark that they are new and who invented them. I was more worried about avoiding a reference to my name in every second sentence. But if you think that is ok, then please feel free to add references.

I do hope though that information sharing is not limited by less prior knowledge but progresses by increasing enhancements of knowledge presented on SL!

Which terms, concepts, methods don't you know and should get extra SL pages?

reply
Dieter: ((no subject)) (2011-09-03 15:14) [#8710]

Previously this page talked about evaluating territory and evaluating thickness. I agree it wasn't very helpful in determining thickness but at least it wasn't biased completely towards counting territory.

Since I've come to realize that Go is about putting more alive stones on the board than the opponent, the fuzzy concept of thickness suddenly became clear in my mind. Since then, territory and thickness are two of a kind to me: alive statuses that can become stones, and stones which will be alive.

Positional judgment is foremost about the balance of stability and development. The analyses methods outlines are useful additions, but they don't explain the heart of the matter. Having something like 1-territory early in the explanation, doesn't do justice to the subject. It's probably very useful for counting correctly, but not crucial for positional judgment.

As always, your efforts to improve go theory are most appreciated.

reply
213.73.118.84: ((no subject)) (2011-09-03 15:35) [#8711]

When you read literature until 2011, it is too easy to get a false impression that positional judgement would do nothing more than counting territory. The previous SL version at least gave a hint that also territory and influence should be assessed but said that that could not be done. The new version now states means with which it can be done. Besides quite some other aspects of positional judgement are now mentioned.

Why do you see stability and development in a mutual balance? I agree that both can be included in a positional judgement. If this is done though, then I wonder why not add all strategic concepts? Judgement is really much more closely related to analysis, which is done by analysis methods.

I do not quite understand your drawn relation between thickness and life statuses. The status I use for that purpose is n-alive but which statuses are you using? Sekis do not create outward influence...

Why do you consider the relation between stability and development at the heart of positional judgement?! - Note that stability is related to quiet position, quiescence, unrest model and development is related to strategic choices and options (but development directions can be considered another aspect of positional judgement, maybe as one of the few strategic concepts worth mentioning here).

If you think that the current explanation is too detailed too early on the page, then you might wish to add a longer introduction?

1-territory is crucial for good positional judgement but I get your point that this is the one term in greatest need of explanation.

reply
Dieter: ((no subject)) (2011-09-03 15:47) [#8712]

Quiescence, n-territory, mobility ... all of these are terms you have coined and for which this page refers to your own book. There is a moral problem with that: either you disclose these terms fully over here, voiding your book, or you think your book does the best service and needs to be rewarded for that, but in that case the community should decide that it is indeed the best work.

I think it is perfectly fine to define these terms over here. It is probably also fine to suggest they are valuable in the treatment of the subject positional judgment. However, to organize a generally used concept like positional judgment, centered around terms which are mostly coined by yourself, is not so certain.

I think what we want to see is an expert rendition of commonly accepted knowledge and suggestions to slowly move forward. There sure is a lot of value in overhauling the system, but maybe not from within.

reply
RobertJasiek: ((no subject)) (2011-09-03 16:05) [#8713]

Quiescence, stability, mobility and some other words or concepts have not been invented by me, although I may have provided new discussion or definition of such concepts. Other terms, concepts or methods are indeed my invention. There is often a development though: E.g., n-territory I have first defined but it has long been in implicit usage during iterative endgame calculations.

When a system consists much more of gaps than of contents, then a major rewriting is essential.

I can provide some more explanations of terms here but don't expect me to do it all alone! E.g., everybody familiar with computer go should be able to explain quiescence.

reply
tapir: headache (2011-09-03 16:27) [#8714]

A bit the page now reads like a page on headache that includes the line "headache is the ailment cured by Aspirin from Bayer". I agree that positional judgment is an headache, I believe you can contribute in overcoming this condition, still I dislike the way you did it here.

I pioneered the ConnectionTermsByRobertJasiek? page, I believe to open that page for all your terminology or make similar ones for other areas of interest is necessary before pushing it on major introductory pages. And yes it is true, counting only half the value of a gote move or a framework that another move may turn into territory is common practice. But when it is a common practice why do you need to put a stamp on it and sell it as yours? Also, SL does not necessarily have to provide only definitions and more definitions, most readers would be better served with a guide to correct counting in my very humble opinion.

X
MrTenuki: Re: headache (2011-09-03 18:39) [#8716]

I think one possible way to fix this potential conflict of interest might be be "dumbing down" the concepts somewhat in the Positional Judgment page for a more general audience-- for example, limiting the page to only the commonly understood terms. For those who are interested in the technical details, a separate page could be created to discuss Robert Jasiek's terminology in greater detail (where a reference to Joseki Volume 2 Strategy would be appropriate).

RobertJasiek: ((no subject)) (2011-09-03 19:29) [#8717]

Of course, there can be pages PJ for beginners, PJ for kyus, PJ for dans, PJ introduction. Who wants to write them?

Moving all technical details is a bad idea though. SL still misses pages for central "technical details" like, e.g., Strategic Choice. Even a PJ introduction should mention strategic choices though because they are one of the major aspects of PJ. Rather where movement makes sense more easily is for the finer details of every topic of technical details. It is not just my terminology that needs explanation in greater detail but also much of the "public domain" terminology like quiescence.

References to inventors or literature are appropriate at different occasions: Inventors should get proper credit where their inventions are specifically discussed (like Zeijst on the QARTS page or me on a page like NConnection). Literature lists should include all most relevant media for deeper reading on the discussed topic. Therefore the current literature list is good, although there might be a few more books I don't recall right now.

MrTenuki: Re: ((no subject)) (2011-09-04 03:47) [#8724]

Well, it would be nice if you could find a few more sources that refer to to the specialized terminology used in the page.

(For the record, I fully welcome your willingness to expand the content of this page; I'll [ext] assume good faith here, that your intention was simply to inform. But again, as Dieter and Tapir has mentioned, people might get the impression that you're using this page as an advertisement simply because a significant portion of the page's content refers to information in your book. This would be less of an issue, though, if other sources are also cited.)

RobertJasiek: Re: ((no subject)) (2011-09-04 08:35) [#8727]

There are only few sources in Western literature and, even worse, most are very specialised. Cho's book is probably the second best available English source but already is so specialised that, IIRC, it explains almost only one kind of analysis methods: counting territory.

Quiescence we could point to the computer go mailing list archives but... which year, which messages? It would be a pretty useless hint. Maybe somebody knows some computer go research papers that happen to explain it?

Unsettled group average is my invention. Elsewhere you can find only related information. Actually you can find it on SL! Look for various pages about miai value, count etc. I forgot which of them was the most helpful. So maybe half a dozen of links would be needed. Other related information is available on rec.games.go, but which messages in which year? I can't remember. Search in particular for Bill Spight's messages. They are about miai value and counts and... well terms of thermography (But who wants to use such? Just using vales is much easier in practice.). My invention, and so far you can find that only in my book, is to apply endgame techniques to the middle game and to study sequences of successive either attacking or defending plays by the same player and to combine that with direct formulas for miai value and count.

Local positional judgement and all its methods are my invention. Even if I try hard to remember prior contributions by others (say, by professionals), I don't recall anything relevant. I am sure there must be something but maybe nothing has been translated yet. It surprises myself that so little was known because local positional judgement is an important method with also the potential to contribute to assess outside thickness. What you find in professional sources (like joseki dictionaries) is more in an informal database style, when they make unproven statements like "Black is slightly favourable". So I wonder: Am I really the first to assess local objective values of thickness? All that has been mentioned for others is at best in relative symbolic numbers style, which to me sounds like alchemy. I know that such is tempting; many years ago I used something similar to QARTS. Now I think that such models are by far too weak because they consist of only wildly guessed numbers independent of group sizes. Local positional judgement offers relatively much more precise methods and values.

Global positional judgement: Here chances are great that more sources can be found, at least practical ones without explain general theory, like game commentaries. You won't find 1-territory and mobility values though. Also I do not recall thorough comments on strategic choices and options, except in books specializing on these, but only in the examples teaching style - nothing general.

Tewari: None of the mentioned tewari methods is new but I do not know any other book explaining all of them. In the English literature, The Vital Points of Go with three or four tewari methods is already the second best.

Mobility values: They are my invention. Simple and powerful but AFAIK entirely new. As new as territory efficiency.

Unrest model: Also that is my invention. I think many players use something similar in a completely intuitive manner without ever having explained it in general.

Advertisement for my book? It cannot be avoided or more than half of the methods would be ignored. I do not know any other book that discusses at least two of the major methods. Why is that so? I have written the book also to cover every analysis method (applicable to josekis) I know, have worked specifically to create those methods as general tools and, where necessary, done specific research for that while writing the book. (Similarly for the other chapters of the book.) The result is, to me not surprisingly, that my book is the best (English, maybe even world-wide) general source in particular for the positional judgement topic. Why not...? I needed 4 months to write it. So I had time to make the necessary studies and inventions. Also I had done preliminary related research during the previous 13 years. Not publishing the book in 1998 may have been a mistake but it had the advantage that I had much time to reflect the topics of the contents.

reply
213.73.118.84: ((no subject)) (2011-09-03 16:52) [#8715]

Guides to correct counting are supposed to be counting and EstimatingTheScore.

Counting half of a gote move is a related but different concept from 1-territory or the more specialized yose iteration.

You are glad to have known counting only half of framework that another move may turn into territory. And no - 1-territory does not say to count only half of the framework but says to count fully that part deep inside that the opponent cannot take away (0-territory) and to count half the outer part where the opponent can still reduce if he plays first. So do you have to correct your prior knowledge or have you just been a bit careless about describing it? Common practice? AFAIR, I have not seen it anywhere else for territorial frameworks. I had to invent or reinvent it. If you do know earlier references to the conceptual idea, please state them! What I saw before (example: Cho Chikun, minimal territory remaining after reductions) was different: count the deep inside part 100% but do not count the outer part yet at all.

In ConnectionTermsByRobertJasiek?, I corrected a mistake by you. If you want a page for all my new or newly defined terms, that would have to be a new one because not all are about connection.

Pages are open for improvement.

reply
Dieter: ((no subject)) (2011-09-03 23:21) [#8719]

Later, when I thought over the resistance I put up, I didn't understand why I bothered. Robert, go ahead. At least you are doing something and not standing by and criticizing like I am.

Time will tell.

X
RobertJasiek: Re: ((no subject)) (2011-09-03 23:25) [#8720]

Thank you for your very kind words! I very much appreciate your insight!

Dieter: Re: ((no subject)) (2011-09-04 02:01) [#8722]

I think it has more to do with being tired of speaking for a probably non-existent community than with true insight. In other words, if you are one of a handful visible users here, why shouldn't I (we) let you define whatever page in whatever way.

If you're onto something, then we should let you move. If you're onto nothing, then we will either see a truckload of questions by a large SL community, the power of which will either move you to something or remove the nothingness, OR there's just no one here.

reply
emeraldemon: comments on page (2011-09-05 07:12) [#8734]

Overall, I think this page is probably an improvement on the previous version, it is at least better organized and more detailed. A few comments:

1) I'm not completely clear on the difference between Positional Judgement and Estimating the Score . EstimatingTheScore is pretty much a discussion page right now, and could really use a WME. Maybe these two pages should be merged?

2) There was another attempt to formalize positional judgement: Framework Theory . Has that died? Should there be cross links on this page?

3) There are still many undefined terms in this page. If the page is still a work in progress that's fine, but if an introduced term cannot be explained I feel that it should be removed. In general I'm hesitant to introduce new terms if a common sense explanation will suffice.

On specific phrases:

  • "quiescence" - I added a link to the wikipedia page, not sure if this needs more. The idea itself is a useful one, I think.
  • "1-territory" - unless I misunderstand, this is already included in the definition of miai value (or CGT value). Is this meant to be a simplification of CGT counting?
  • "Unsettled group average" - Again I guess this is a special case of miai counting? I think we could have advice like: "when counting territory of a group, include the effect of possible future attacks" without introducing a new term.
  • "excess influence stones" - I don't understand this at all.
  • "the method 'attacking an unsettled group'" - not sure if this is really a seperate method, or the same as "unsettled group average".
  • "assessing influence and thickness by means of their formal definitions" - I don't think thickness or influence have formal definitions. I'm fairly certain long and unproductive arguments have been had on this subject already.

Robert, Are you planning on adding more here? If you are I will wait to make changes. Either way I will try to make time to WME EstimatingTheScore soonish, and we can see if a merge makes sense.

X
RobertJasiek: Re: comments on page (2011-09-05 09:32) [#8736]

Estimating the score is basically just counting territory, isn't it? Positional judgement assesses all relevant aspects of a position: territory, thickness, influence, strategic choices, mobility etc.

1-territory please discuss on n-territory! There no reference to miai value or CGT is made in the definition.

Miai counting is a somewhat fuzzy phrase. So I am not sure if Unsettled Group Average is a special case. Surely it is a special case of determining miai values and counts. It is more than that though: It is a method: a combination of procedural advice and formulas for values. "after imagining an attacking sequence and a defending sequence" already describes the method roughly. More precisely: attacker or defender make successive plays in their respective sequence. As usual, for miai values, the number of plays made is taken into account for the values calculation.

Excess influence stones is another method invented by me. "which imagines additional territory constructing stones by one player and then estimates the 'territory efficiency'" says it all but is, of course, a very short description. One player makes successive plays to construct territory (e.g., adjacent to his wall). Then the number of his excess stones (more stones by him than by his opponent) is counted. Then the territory efficiency is determined, for which the page already describes "points per played stone".

Of course, "attacking an unsettled group" is a different method - therefore it does carry a different name!

Influence and thickness do have formal definitions and can be assessed in terms of numbers: I have invented them in my book. The definitions simply refer to n-connection, n-alive, n-territory. In case of thickness, one determines the parameters for its player. In case of influence, one determines the parameters for either player (although it can suffice to consider n-territory only for the player for whom an intersection is under his greater influence).

Currently I do not intend to add significantly more to the parent page but I think others might wish to add a more detailed introduction section to it. Merging with estimating the score is a bad idea because that is something much more specialized.

reply
tapir: ((no subject)) (2011-09-05 10:12) [#8739]

I would like to order this page according to purpose of the various methods. Right now it is one big list.

1. Counting

- basic counting
- tools and ideas for counting
-- quiscience
-- qarts (assigning negative points for weak groups)
-- unsettled group average
-- how to count a framework?
...

2. Assessing influence, thickness, etc.

- mobility
...

3. Assessing efficiency

- tewari
- territory efficiency
...

X
RobertJasiek: Re: ((no subject)) (2011-09-05 10:54) [#8740]

Quarts does not count but creates symbolic numbers.

Mobility does not assess thickness or influence.

Not all methods of tewari assess efficiency.

Currently it is not one big list but structured by classes of methods and within each class by methods.

What about creating another section of the page in which (sub)methods are assigned to assessment objectives?

tapir: Re: ((no subject)) (2011-09-05 10:58) [#8741]

Qarts does not count but is meant as an add-on to counting. You count then subtract points for weak groups - how you arrive at these numbers isn't important to list them among counting tools. Similarly quiescience does not count but clearly it is a tool for counting. So I propose ordering the page less by what those models, methods etc. fundamentally are but for what purpose they are used. Mobility may not assess thickness or influence but the "etc." I added there, but as a count of outside facing stones there is at least some overlap, no?

RobertJasiek: Re: ((no subject)) (2011-09-05 11:04) [#8742]

No, until you provide a formal definition for mobility in the style of the formal definitions of influence and thickness by reducing it to the lower level concepts connection, life, territory.

tapir: Re: ((no subject)) (2011-09-05 11:14) [#8743]

Robert, there are methods that function as additions / improvements for plain counting and there are completely different methods as tewari with an altogether different mechanic etc. this difference should somehow materialise in the page. Btw. can you provide a formal definition for mobility that reduces it to a lower level concept?

RobertJasiek: Re: ((no subject)) (2011-09-05 11:40) [#8744]

Quiescence uses preliminary sequences after which then territories are counted.

I do not see how QARTS or similar methods provide an improvement. Rather they are very rough substitutes for accurate methods like Unsettled Group Average or Iterative Averages (which is hard to apply during the middle game though).

Presumably I have the ability to create a formal model for mobility but such models are not found quickly; I would need weeks or months of research. For the formal models of thickness and influence, I needed years. Now that we have them, I know how to proceed in principle, so my guess is that I could define mobility formally after a much shorter time. Just not quickly. Finding accurate, general models always tends to be hard work.

tapir: Re: ((no subject)) (2011-09-05 12:48) [#8745]

Ok, then please don't ask me to do sth. that would take you months or weeks. Likely mobility as the count of stones on the non-territory outside has to do something with concepts like influence or potential, it need not to be the same. What is important here is that it isn't related to counting territory and should be put in a group with other methods to assess non-territorial properties.

QARTS is just a modification of plain counting, you need to count to apply it. It might not be the best such modification but likely it is an improvement over counting only sure territory without taking notice of weak groups. Similarly quiescience has value only if you count afterwards, obviously enough I believe. So they should be grouped together with other counting related methods rather than the opposite ends of one long list. Both are not counting itself but they prepare for counting or modify the count, that much should be obvious.

 
Back to forum     Back to page

New reply


Forum for Positional Judgment
RecentChanges · StartingPoints · About
Edit page ·Search · Related · Page info · Latest diff
[Welcome to Sensei's Library!]
RecentChanges
StartingPoints
About
RandomPage
Search position
Page history
Latest page diff
Partner sites:
Go Teaching Ladder
Goproblems.com
Login / Prefs
Tools
Sensei's Library