Dear Robert,

"being responsive to"

-- Main line
-- Variation A
-- Variation B

does not mean at all

-- Oh, I know a very interesting sub-variation B-C, let's discuss this.

I do not think that this behaviour of yours has any dependency to what is permitted to you to do in this forum.

_________________
The really most difficult Go problem ever: https://igohatsuyoron120.de/index.htm
Igo Hatsuyōron #120 (really solved by KataGo)

RBerenguel, without counting preceding research, writing the paper took about 2 weeks. If it were meant for easy consumption, I would spend another 4 or 8 weeks to create a book. However, I lack the necessary time. - Maths papers I have seen at university had less than one 20th of the amount of explanations in my paper. Or look at typical computer go research papers: they have much less explanation.

My bullshit detector just began screaming its alarm.
Robert, this is plainly ridiculous.

Would you say it is consciously evasive behaviour or would you rather say that you’re not aware of what you’re doing here?

_________________
“The only difference between me and a madman is that I’m not mad.” — Salvador Dali ★ Play a slooooow correspondence game with me on OGS?

Bonobo, OC I can explain the details of my paper. For everything else, see the TOS, or write me an email.

1. I think you are underestimating what (real?) research is, and probably the distance between publishable quality and meant for easy consumption. Writing (actually typing, revising, rewriting and a final burst of composing, not counting the previous time needed to decide what goes in and what does not) the 8 pages of the introduction took me a month. And it was a joint work with my PhD co-advisor, we were meeting almost daily to craft it. Previous research was probably a full year, shared with her. Revising the full paper for submission (when all research and correcting was done) for readability took another month. 2 weeks is usually the time needed to prepare a 30 minutes presentation on current research (unless you have already a presentation for it). This numbers are not an isolated example from my work: my department colleagues usually took the same amount, give or take a week.

2. I have looked at many papers. Complex dynamics and real dynamics, real analysis, numerical analysis, computer science (several scattered branches depending on what I was looking for.) I have seen good papers, and I have seen bad papers. Usually Russian authors fall into the "bad papers" basket because their style of explaining things is not the style I prefer, but you can get into them with a little patience. Few are so bad that I don't bother to pass the first pages (I can only remember a handful.) Oh, interesting: The Texas sharpshooter fallacy. I know there's a more precise fallacy here, but I can't remember the name.

3. I don't see why I should look at a typical computer go research paper to evaluate the quality of yours: when grading a student I used to grade based on the own merits of an exam/work done, not on what the student's best buddy did, or what the rest of the class did. In any case, I have read a few computer go papers about Monte Carlo tree search and related concepts to writing engines. They blow your paper out of the "research" water.

_________________
Geek of all trades, master of none: the motto for my blog mostlymaths.net

RBerenguel, after so much meta-discussion, how about discussion of my paper?

I just got here after your snarky remarks to Cassandra, since I was following the thread to see if you explained something interesting. After skimming through the first few pages I am pretty sure I am not interested in reading it. Probably I'm not interested in the subject, too, but with the current presentation I don't plan on investing any amount of time in reading it. Oh, and since this thread is supposedly about "research" I can't really see where the "metadiscussion" argument comes. Well, actually maybe I do. See you in other threads.

_________________
Geek of all trades, master of none: the motto for my blog mostlymaths.net

Then it is not surprising that you cannot appreciate the research underlying my paper.

Meta-discussion???

You've someone telling you that the quality (not the content) of your paper is lacking. Really, this is extremely relevant to your work. If you want it to be more widely read and talked about you need to present it well and as clearly and concisely as possible. Presentation of your work is equally as important as the work itself. Unless its readable no one will bother to spend time on it unless they feel they have to (unless it's Continental Philosophy or Literary Criticism inspired work, then the opposite applies).

The preface and abstract read like a book blurb and opening chapter not anything you'd find in a research paper. I appreciate you're not working in your first language but getting a native English speaker who can reasonably work as an editor (read: could get paid work in this area as opposed to a friend who speaks English) would massively improve your capture of readers past the first couple of pages.

I'm sorry if that's harsh Robert, but if you want to be taken seriously by people in general rather than a small circle of go theorists, you need to write better.

Boidhre, a typical research paper would have the abstract, the 2 pages of definitions, about one example per type of ko intersection and little else. The preface, reasons for the non-obvious conditions and most of the remarks would be omitted; all the other examples would have been omitted and replaced by a short remark "has successfully been applied to all known ko shapes and to important counter-examples"; the definitions would have used mathematical annotation, so that more go players have less chance to understand them. So, of course, the paper does not look like typical research papers, which are compressed to be short enough to fit in a journal.

The definitions are as concise as possible. (With the earier mentioned exception of one superfluous line added for easier reading.)

The paper is written first of all for go theorists. The reader, especially if he not a go theorist, needs to invest time and effort. E.g., if the reader does not immediately understand the strategy- and force-related terms, he should read related explanations:

http://home.snafu.de/jasiek/j2003com.html
http://senseis.xmp.net/?ForceMathematics

If somebody complains about an, in his opinion, missing / low quality of the paper, he should justify it by better reasons than his missing willingness to invest time and effort, or ask questions.

I do not expect the paper to be widely read, because it is not written for the non-theorist go player. I do not recall any go theory research paper that would have been written for the non-theorist go player. This is no coincidence, because translation for him requires more time than an author of a research paper tends to have. Nevertheless, ordinary players can appreciate at least the example diagrams.

So is go theory is for everyone, but the papers are just for theorists?

_________________
"You have to walk before you can run. Black 1 was a walking move.
I blushed inwardly to recall the ignorant thoughts that had gone through
my mind before, when I had not realized the true worth of Black 1."
-Kageyama Toshiro on proper moves

Dear Robert,

I strongly question the methodology, as well as the presentation, of your paper. I think that both aspects have to be seen as an entity.


# # # # # # # # # # # # # # # # # # # # #

Starting point is your definition of "basic ko" (source: http://home.snafu.de/jasiek/korules.html#BasicKo):


"Common understanding" would imply that the property "Basic-Ko" is connected to a board position, AS LONG AS the "Basic-Ko Rule" applies.

To make this more evident: BEFORE Black connected his single stone, there was a "Basic-Ko". AFTER Black connected, there is no "Basic-Ko" anymore.

"As long as" implies that the "history of moves" has to be taken into account.


# # # # # # # # # # # # # # # # # # # # #

But -- for a reason that I do not understand -- this is not the way of your argumentation (/ thinking). You cut move sequences that are inseparably connected into slices, and consider each position after each move as if it were an independent "starting position" with no relationship to any historical development.

To make it more evident: Here is my understanding of the variation tree of a "Basic-Ko".

Main line:
Black 1 captures the Ko.

Main line: White 2 plays a Ko-threat.
-- Variation:
-- White 2 plays Tenuki.
-- Black 3 connects the Ko. End of the Ko-fight.
++ Variation:
++ White 2 passes.
++ Black 3 connects the Ko. End of the Ko-fight.

Main line: Black 3 answers the Ko-threat.
-- Variation:
-- Black connects the Ko. End of the Ko-fight.

Main line: White 4 captures the Ko.

Black 5, White 6, Black 7 as White 2, Black 3, White 4, with colours reversed.
Black 7 results -- as far as the "Basic-Ko" is concerned -- in the same board position as after Black 1.

Now let us assume that there are more than one "Basic-Ko" on the board, let's say three. As a matter of course, I assume an appropriate combination of who is to capture the Ko.

It should be evident that the concrete occurences of the above described variation tree -- applied to the three cases of "Basic-Ko" -- are interwoven.

To make it more evident:
Black 1 captures the Ko in case 1 (== Black 1 in the variation tree).
White 2 captures the Ko in case 2 (== White 4 in the variation tree). With regard to case 1, this is White playing a Ko-threat.
Black 3 captures the Ko in case 3 (== Black 1 in the variation tree). With regard to case 1, this is Black answering White's Ko-threat. With regard to case 2, this is Black playing a Ko-threat.
and so on, and so on ...

Correspondent moves -- with regard to their order in the variation tree -- are played at different times.
But there is no change between "Basic-Ko" and "non-Basic-Ko", even if any of the Ko-captures would be an Atari.


# # # # # # # # # # # # # # # # # # # # #

Now let's come back to your move Black 1 (within "case 3", let's say) in the minimized position, which you call "standard triple-ko", that captures White's group.

This move does not have any representation in the variation tree that is desribed above. This rises the question whether "your" move is really applicable within a "Triple-Ko".

-- Explanation 1: Black played two moves in a row. This violates one of the core rules of the game and therefore can be excluded.

-- Explanation 2: White dispensed with her move in "case 1" (assuming the existence of three "Basic-Ko"), for a reason that only you understand. This does not provide any value to the research of the issue in question. And also violates your general "forcing-idea" within your paper.

-- Explanation 3: White played a Self-Atari. This might be an interesing topic for another research paper, but does not provide any value to the research of the issue in question.

Conclusion:
"Your" move is not applicable within a "Triple-Ko". Just because you chose an artifical position with an artifical player to move.

This is comparable with someone -- who has a rudimental knowledge of the game only -- taking a snapshop of the board accidentally, and then starting with Black, just because it is Black, who begins move sequences in the majority of diagrams this one has seen before.

I do not think that you would need to stoop to this too low level.

But you did -- repeatedly -- and this will not provide to a reputation of being a "reliable researcher".

_________________
The really most difficult Go problem ever: https://igohatsuyoron120.de/index.htm
Igo Hatsuyōron #120 (really solved by KataGo)

Robert, you are completely failing to understand what I'm saying and I can't be bothered with this, so grand, whatever.

moyoaji, go theory papers are for those willing to invest the necessary time and effort. A reader lacking specialised education needs to acquire specialised knowledge, before he can expect to understand a paper at all. A reader having studied some mathematics (or related fields) can expect to understand everything if he invests at least 30 minutes per difficult text line (such as the definition of 'strategy'). Everybody else might need more time.

Somebody wishing to apply roughly the core of the theory without understanding the fundamental layers needs to get at least a rough understanding of the last conditions of local-/global-ko-intersection. For that purpose, an informal understanding of "(answer-)force" and "prevent" might suffice.

Cassandra, my paper is not meant to be interpreted with common understanding, but understanding must come from the paper's definitions. There is: "A basic-ko-intersection is an intersection of a basic-ko." Therefore, your idea "AS LONG AS the 'Basic-Ko Rule' applies" does not apply. The definition of basic-ko-intersection does not refer to a basic ko rule or its application, nor does it refer to "as long as". A basic-ko is not defined by means of the basic-ko-rule. A basic-ko consists of the two intersections on which a single stone could capture a single stone and then immediately a single stone could recapture a single stone.

In "BEFORE Black connected his single stone, there was a 'Basic-Ko'. AFTER Black connected, there is no 'Basic-Ko' anymore.", you consider two different positions: the position before Black's connection and the position after Black's connection. For each position, one can determine whether the two studied intesections are a basic-ko.

"As long as" implies nothing, because you have made up this "As long as". It does not occur in the definition of a basic-ko-intersection. What is not in the definition is not applied.

To reduce confusion, it would be nice if you used the terms as I use them: "move-sequence", "start-position". For the paper, I do not write "move sequence" or "starting position". By using the paper's spelling, confusion is avoided as to whether indeed you mean my term or whether you mean something else.

What do you mean by "You cut move sequences that are inseparably connected into slices"? "cut", "inseparably connected" and "slice" are not terms in my paper.

For basic-ko-intersection or local-ko-intersection, history bans are not considered. For global-ko-intersections, one chooses whether to consider history bans or whether to study a particular position (with its turn and komi) as is.

When a particular position is studied, it is determined for this position which are the ko-intersections and which are not. For that sake, each of basic-, local-, global-ko-intersection is studied separately from each other and separately for each intersection of the board, because each of these terms' definitions speak of "...is an intersection...".

So, let is suppose a particular intersection of the studied position is studied for whether it is, e.g., a local-ko-intersection, then one studies move-sequences with certain properties. Each of these move-sequences is related to the studied position. In the simplest case, the relation is that the studied position is the "current-position"; then the "start-position" is this current-position and is the studied position. (There is another possible, more complicated case, which I do not mention now.) So we have some such move-sequence. Now, you wonder about the effect of this move-sequence changing the studied position. (Unless it consists of passes,) the move-sequence does lead to other positions, but those are not the studied position; the start-position remains the same. This is a purpose of using the term start-position: we still know which position is being studied for which of its intersections are the ko-intersections. During this particular study, we do not study "which intersections are the ko-intersections" of other positions occurring during such a move-sequence.

A basic-ko does not have any variation tree, because the definition of basic-ko-intersection does not refer to any variation tree.

A basic-ko does not consider any ko-threat, because the definition of basic-ko-intersection does not refer to 'ko-threat'. Same for local- or global-ko-intersection. The definitions of the three ko types are INDEPENDENT of ko-threat: you do not find 'ko-threat' mentioned in the definitions of basic-ko, move-sequence, left-part, strategy, answer-strategy, compatible, answer-compatible, can force, does force, prevent, answer-force etc. If you want to understand the terms in my paper, you must NOT apply something that is NOT in the definitions.

Same for 'ko-fight'.

Same for 'connect a ko'.

You write "I assume an appropriate combination of who is to capture the Ko", but, for the sake of my definitions, such an assumption cannot be made, because the definitions do not speak of "an appropriate combination of who is to capture the Ko". Instead, local-ko-intersection speaks of "a player"; global-ko-intersection considers the turn in "the moving player" and then speaks of him as "the player".

The definition of basic-ko-intersection does not speak of a combination of three (or however many) basic-kos, but speaks of "an intersection of a basic-ko", i.e., for each intersection of the board it is considered whether it is a particular such "an intersection". After applying the definition to your interesting six intersections, you will have determined that each of them is a basic-ko-intersection. This determination does not depend on any combination of basic-kos. All that matters is whether every particular basic-ko exists in the studied position.

Your "This rises the question", Explanations, Conclusion and "this will not provide to a reputation of being a 'reliable researcher'" are conclusions from immaterial assumptions, such as considering a variations tree at all. Why immaterial, see above.

For the sake of the definitions, it is immaterial whether a position is artifical. The definitions apply to each legal position, whether or not artificial, and regardless of the board's grid.

Dear Robert,

If you wanded the contents of your paper to remain completely "inside your world", and not let it diffuse slowly to the "common world", and let it be understood therein, why on earth did you publish it ? And where does your assumption come from that it would benefit the "common world" (e.g. increase the insight into what you call "ko strategy") ? Because the above will be also true for your other papers, why on earth did you start this thread ?

If you were a monopolist, who distributed an indispensable product, I could understand that you strove by all means to have the entry barrier into your market as high as possible. But you are the owner of a market nicher, who wants to sell a dispensable niche product, and strive by all means to not become engaged in any marketing, or promotion, measure that could support the launching of your brand.



For the rest of your post,



applies, too.

_________________
The really most difficult Go problem ever: https://igohatsuyoron120.de/index.htm
Igo Hatsuyōron #120 (really solved by KataGo)

Could you point us towards a "typical" go research paper?

Wrong. I have provided one example already. I could give around 20 more without more effort than opening my old department friends' homepages and grabbing their preprints from there.


Wrong. I have already commented on this.

Wrong, usually this kind of results get written in a paper to be submitted to a lesser journal, and available in the researcher's page, at the very least. Likewise, I could bring a few examples without much effort.

I guess the go theory researchers and the go players with mathematical inclinations are pretty much overlapping. I guess most go players interested in this paper are in this subset. I don't see then where mathematical notation would have hurted.

Wrong. I have already commented on the amount of work needed in a typical research paper to write introduction and preliminary sections. They can easily be more than 30% of the whole paper.


In my opinion (what a funny turn of events that I have to use this phrase!) your paper has a very low quality preface and preliminaries. It lacks a table of contents. The typesetting is sub-par (no named definitions, no cross-reference of definitions and hypotheses, etc.) It lacks a clear structure. The phrasing in most places I checked is unclear, with long phrases. The overuse of the passive voice should be avoided. The definitions and restrictions in the first pages lack immediate motivation (since a definition for a new term has to be motivated either immediately before or immediately after it is needed or introduced in practice.) The fact that the paper lacks all this has nothing to do with the fact I don't give a **** about its contents.


Almost no researcher expects his papers to be read by except than a handful of other researchers. And anyway, they care and try to make it readable. Also, above you said



and now you say



So, who shaves Bertrand Russell?

_________________
Geek of all trades, master of none: the motto for my blog mostlymaths.net

May be that you are the ONLY one who thinks that understanding of your paper must come ONLY from "your world" inside your paper.
But you massively overburden the inhabitants of the "common world". And additionally, what benefit should come from studying the contents of your paper, which is valid only within the borders of "your world", and cannot be transformed to the "common world" ?

Let me give you an example, which is not related to the game of Go.

Let us assume that there is a "common understanding" what a "car" is. Your claim is to have a very special "car theory", which will bring forward the ability of driving cars.

Your definition of a car, its elements, their properties, and behaviours, do not fit "common expectations". Even if is seems that some of your definitions etc. are largely similar to what is used in the "common world", especially if you use identical terms, properties, and behaviour, of the element in question will still differ from "common expectation".

"Your car" has four "wheels", as "commonly expected", but these are not circular, but 32-angled, not made of rubber, but plastic, not filled with air, but solid. "Your car" has a "steering wheel", as "commonly expected", but turning this wheel left makes the car turning right.

It might be possible that someone will be able to drive "your car", but in a very special, and sheltered, environment only. No one (maybe but you) will be interested in driving "your car" within the "common world".

_________________
The really most difficult Go problem ever: https://igohatsuyoron120.de/index.htm
Igo Hatsuyōron #120 (really solved by KataGo)

Cassandra, there is not "my world", but there are "my definitions". The definitions can be read by everybody, regardless of whether they are not understood by everybody.

In maths, the purpose of a definition is to use or apply this definition. The purpose is not: to avoid its use or application.

Theory goes through stages:

1. The theory is created.

2. The theory is understood by others.

3. Optionally, an approximation of the theory is translated so that more people can understand at least the basic idea.

Maybe you confuse the stages 2 and 3. On the on hand, you want to understand the theory itself, e.g., when you try to find mistakes in it in order to prove how bad a researcher I am;) OTOH, you want an approximation of the theory, e.g., when you complain that I insist on my definitions being used and applied at least when my paper as such is being discussed.

I published the theory in particular for the purposes 1 to 3.

That usually 3 comes after 1 and 2 does not mean that 3 would be impossible.

Elsewhere, I gave hints how ordinary players can profit from the theory or an approximative understanding of its basic idea.

My other papers do not all fall into one kind. I have written research papers and have written other (more) application friendly papers.

My "pure" go theory papers are not marketing papers.

pwaldron, examples of typical go research papers:

http://www.dumbo.ai.kyutech.ac.jp/~teigo/GoResearch/
http://typo3.cs.uni-paderborn.de/filead ... f13_cg.pdf
http://lie.math.brocku.ca/twolf/papers/semeai.pdf

RBerenguel, it seems we were confronted with different kinds of papers at university.

Mathematical notation would have hurted, e.g., my own research, because it could have proceded only much more slowly.

Concerning the research amount for my ko paper: did you notice that it took me 13.5 years?

Concerning your comment "has a very low quality preface and preliminaries. It lacks a table of contents. The typesetting is sub-par (no named definitions, no cross-reference of definitions and hypotheses, etc.)", I agree that these aspects could be improved greatly if I invested a couple of additional weeks. I lack the time to do so.

Concerning "lacks a clear structure", I disagree: the paper is devided into clear sections, as are the definitions.

Concerning "The phrasing in most places I checked is unclear, with long phrases." it depends on whether you mean the definitions or the commentary. The commentary I have not optimised for best English, because I lack time to do so. The definitions have a couple of long phrases or sentences, because this supports conveying the contents. If sentences were split, unnecessary ambiguity would be introduced about which sentence provides additional necessary conditions for which other sentences. Alternatively, the definition text could have been made much longer by using many lists; I have preferring to use lists for the central definitions, to emphasise them; if I had used lists everywhere, this emphasis would have been lost.

Concerning aspects of English style in that paper, I lack time to improve it. You need to a) not read, b) bear this / complain about this or c) wait one or two decades, until hopefully find time to publish everything in a book. Since I lack time to improve the English style in such papers quickly, my alternative would be not to publish such papers. I prefer publication with imperfect English style to no publication at all.

Concerning "The definitions and restrictions in the first pages lack immediate motivation", motivation is found a) in the application of the early stated terms and b) in the explanations on the restrictions. If, however, you want, per term, motivation - definition - application/commentary, you need to wait for (c).

I know that many papers are expected to be read by only a handful of other researchers. I do, however, think, that this situation can change. Not by motivating necessarily you to read, but by motivating at least a few more interested people to read.

Cassandra, your car example is no exception to the rule that all car examples are flawed;)