[...] you have yet to convince me that your work [...is...] worth something.
[admin]
Much of the above post was removed for being in violation of the TOS. -JB
[/admin]
A vague treatise on influence
-
Magicwand
- Tengen
- Posts: 4844
- Joined: Wed Apr 21, 2010 5:26 am
- Rank: Wbaduk 7D
- GD Posts: 0
- KGS: magicwand
- Tygem: magicwand
- Wbaduk: rlatkfkd
- DGS: magicwand
- OGS: magicwand
- Location: Mechanicsburg, PA
- Has thanked: 62 times
- Been thanked: 504 times
Re: A vague treatise on influence
"The more we think we know about
The greater the unknown"
Words by neil peart, music by geddy lee and alex lifeson
The greater the unknown"
Words by neil peart, music by geddy lee and alex lifeson
-
RobertJasiek
- Judan
- Posts: 6273
- Joined: Tue Apr 27, 2010 8:54 pm
- GD Posts: 0
- Been thanked: 797 times
- Contact:
Re: A vague treatise on influence
Do not expect me to spend months on repeating explanations in my work and let me concentrate on the thread subject.Magicwand wrote:you have yet to convince me that your work [...is] worth something.
My description of influence has newly introduced these aspects and added it to the previous understanding of being given due to the impact of stones:
1) Influence is said to depend on connection status.
2) Influence is said to depend on life status.
3) Influence is said to depend on territory potential.
4) Influence can be measured by precise and always meaningful degrees.
I can convince you only if you share my view that
1) knowing that influence depends on connection status is more valuable than not knowing this,
2) knowing that influence depends on life status is more valuable than not knowing this,
3) knowing that influence depends on territory potential is more valuable than not knowing this,
4) knowing influence's precise degrees is more valuable than having an only imprecise understanding of degrees.
Therefore you need to admit that you accept at least one of these conditions. Otherwise I can never convince you.
As the devil's advocate, you will of course also doubt whether indeed I have been the first to invent the mentioned conditions or whether they existed earlier. You have the chance to provide any possibly existing, still unknown to me evidence of earlier existence.
-
RobertJasiek
- Judan
- Posts: 6273
- Joined: Tue Apr 27, 2010 8:54 pm
- GD Posts: 0
- Been thanked: 797 times
- Contact:
Re: A vague treatise on influence
For the historians, my earliest public draft of such an idea might have been here on 2005-08-08:
http://www.dgob.de/yabbse/index.php?topic=1651.0
http://www.dgob.de/yabbse/index.php?topic=1651.0
-
Loons
- Gosei
- Posts: 1378
- Joined: Tue Apr 20, 2010 4:17 am
- GD Posts: 0
- Location: wHam!lton, Aotearoa
- Has thanked: 253 times
- Been thanked: 105 times
Re: A vague treatise on influence
Peace out, dudes.
Tangentially, I don't believe anyone's mentioned Ishida Yoshio's All About Thickness. I found it pretty lucid, and the style is pleasant.
Tangentially, I don't believe anyone's mentioned Ishida Yoshio's All About Thickness. I found it pretty lucid, and the style is pleasant.
-
RobertJasiek
- Judan
- Posts: 6273
- Joined: Tue Apr 27, 2010 8:54 pm
- GD Posts: 0
- Been thanked: 797 times
- Contact:
Re: A vague treatise on influence
I did. Since I do not possess the book (it appeared when I was too strong for it), I do not recall all its contents though.Loons wrote:I don't believe anyone's mentioned Ishida Yoshio's All About Thickness.
What DOES it say about what influence are thickness ARE? I recall that it gives basic (and IIRC useful) advice about how to USE thickness but that is a different topic from saying what the concepts mean.I found it pretty lucid, and the style is pleasant.
-
daal
- Oza
- Posts: 2508
- Joined: Wed Apr 21, 2010 1:30 am
- GD Posts: 0
- Has thanked: 1304 times
- Been thanked: 1128 times
Re: A vague treatise on influence
I suspect that most people buying go books are more interested in how to use a concept than what exactly it means. I personally would have preferred more of the former and less of the latter in Joseki 2.RobertJasiek wrote: I recall that it gives basic (and IIRC useful) advice about how to USE thickness but that is a different topic from saying what the concepts mean.
Patience, grasshopper.
-
shapenaji
- Lives in sente
- Posts: 1103
- Joined: Tue Apr 20, 2010 10:58 pm
- Rank: EGF 4d
- GD Posts: 952
- Location: Netherlands
- Has thanked: 407 times
- Been thanked: 422 times
Re: A vague treatise on influence
So this is along the lines of your researches Robert,
but I remember a while back I was doing some influence maps of my own, using reflection of an influence-generating function off walls. (Using a method of images, take the board and make 8 reflected boards around it. Then compute influence for this "mega-board")
Ex:
X X X
X O X
X X X
(I believed that a high density of influence could be considered territory, and this would be an explanation for why a 3rd line stone is so much better than a 4th line stone for territory. The reflected portion in that same region is much higher, because of the exponential attenuation)
Another step I took was to have each stone contribute its influence to every point it could "see"(compute the relative angle of every group of stones, for each stone, find the closest group for each angle, and eliminate any points with a greater radius at the same angle)
I'm not sure the second part is accurate though. Influence ought to bend around groups,
In this way it should be more like a flow of water and less like a luminous source.
Any thoughts of how to improve these maps? I'd like to sit back down with them again...
but I remember a while back I was doing some influence maps of my own, using reflection of an influence-generating function off walls. (Using a method of images, take the board and make 8 reflected boards around it. Then compute influence for this "mega-board")
Ex:
X X X
X O X
X X X
(I believed that a high density of influence could be considered territory, and this would be an explanation for why a 3rd line stone is so much better than a 4th line stone for territory. The reflected portion in that same region is much higher, because of the exponential attenuation)
Another step I took was to have each stone contribute its influence to every point it could "see"(compute the relative angle of every group of stones, for each stone, find the closest group for each angle, and eliminate any points with a greater radius at the same angle)
I'm not sure the second part is accurate though. Influence ought to bend around groups,
In this way it should be more like a flow of water and less like a luminous source.
Any thoughts of how to improve these maps? I'd like to sit back down with them again...
Tactics yes, Tact no...
-
RobertJasiek
- Judan
- Posts: 6273
- Joined: Tue Apr 27, 2010 8:54 pm
- GD Posts: 0
- Been thanked: 797 times
- Contact:
Re: A vague treatise on influence
In relation to josekis, concepts must first of all be understood for the sake of building corner sequences and why they are built in particular manners. There understanding the concepts as such is more important than usage. Usage is also very interesting but I will discuss that in much greater detail for books about the middle game to be written later. Usage often occurs at a distance or in relation to positional context farther away. Such requires lots of full board diagrams. Instead of splitting the joseki series into too many volumes, I consider it more useful to discuss usage for the middle game.daal wrote:I suspect that most people buying go books are more interested in how to use a concept than what exactly it means. I personally would have preferred more of the former and less of the latter in Joseki 2.
-
RobertJasiek
- Judan
- Posts: 6273
- Joined: Tue Apr 27, 2010 8:54 pm
- GD Posts: 0
- Been thanked: 797 times
- Contact:
Re: A vague treatise on influence
This seems too artificial to be useful. Interpretation would be hard.shapenaji wrote:I was doing some influence maps of my own, using reflection of an influence-generating function off walls. (Using a method of images, take the board and make 8 reflected boards around it. Then compute influence for this "mega-board")
I started with such light models in the 90s, then tweaked a bit with stone circle shapes on a truely square board absorbing light, then revised again to let dead stones pass light etc. The pure light model (black positive, white negative light) is good enough for a very rough intuition of where the spheres of influence are and whether the light emitting stones are near. Beyond that it tells us almost nothing useful. It also cannot tell us precise things because we are not super-computers which could add all the stones' light ray functions for various (not always orthogonal) angles.Another step I took was to have each stone contribute its influence to every point it could "see"
The light model is so attractive because we love light:) I needed 10 years to overcome that feeling and create the idea of connection and life degrees.
-
RobertJasiek
- Judan
- Posts: 6273
- Joined: Tue Apr 27, 2010 8:54 pm
- GD Posts: 0
- Been thanked: 797 times
- Contact:
Re: A vague treatise on influence
Forget it! It is not worthwhile! Instead I suggest to study applications of my model. E.g., it is interesting to ask when it suffices to consider only connection or only life and the other aspects are dominated (currently cannot provide additional relevant information).shapenaji wrote:Any thoughts of how to improve these maps? I'd like to sit back down with them again...
-
RobertJasiek
- Judan
- Posts: 6273
- Joined: Tue Apr 27, 2010 8:54 pm
- GD Posts: 0
- Been thanked: 797 times
- Contact:
Re: A vague treatise on influence
Thinking about it again, I could not quite believe it and so I have counted Joseki 2 Strategy's bold (= most important) text paragraphs (sometimes blocks of factually related paragraphs) with these results:daal wrote: I suspect that most people buying go books are more interested in how to use a concept than what exactly it means. I personally would have preferred more of the former and less of the latter in Joseki 2.
85 definitions (of terms, concepts, methods)
176 applications (principles, applying conditions for concepts etc.)
That is 67.4% application, although the many examples are not even included in the counts.
For the influence and thickness chapters, matters are different though:
19 definitions
8 applications
This is 29.6% application.
***
EDIT: So I think what you mean is maybe not the ratio definitions : applications but the ratio general_idea : number_of_examples? More examples per principle etc. would only have been possible if the number of definitions, concepts, principles would have been much smaller.
-
cyclops
- Lives in sente
- Posts: 801
- Joined: Mon May 10, 2010 3:38 pm
- Rank: KGS 7 kyu forever
- GD Posts: 460
- Location: Amsterdam (NL)
- Has thanked: 353 times
- Been thanked: 107 times
- Contact:
Re: A vague treatise on influence
I agree with Robert. Light in Newton's particle interpretation is a linear theory. Light from two sources just plainly adds. I don't think influence works in this way. In Huygens wave interpretation things are even worse. I can't imagine influence from one black stone to be annihilated by influence from another black stone.RobertJasiek wrote:This seems too artificial to be useful. Interpretation would be hard.shapenaji wrote:I was doing some influence maps of my own, using reflection of an influence-generating function off walls. (Using a method of images, take the board and make 8 reflected boards around it. Then compute influence for this "mega-board")
I started with such light models in the 90s, then tweaked a bit with stone circle shapes on a truely square board absorbing light, then revised again to let dead stones pass light etc. The pure light model (black positive, white negative light) is good enough for a very rough intuition of where the spheres of influence are and whether the light emitting stones are near. Beyond that it tells us almost nothing useful. It also cannot tell us precise things because we are not super-computers which could add all the stones' light ray functions for various (not always orthogonal) angles.Another step I took was to have each stone contribute its influence to every point it could "see"
The light model is so attractive because we love light:) I needed 10 years to overcome that feeling and create the idea of connection and life degrees.
Also other physical systems like two dimensional spin lattice models are linear and hence of no use as model for influence.
First it should be established that influence is a linear concept. I doubt this is possible.
-
RobertJasiek
- Judan
- Posts: 6273
- Joined: Tue Apr 27, 2010 8:54 pm
- GD Posts: 0
- Been thanked: 797 times
- Contact:
-
shapenaji
- Lives in sente
- Posts: 1103
- Joined: Tue Apr 20, 2010 10:58 pm
- Rank: EGF 4d
- GD Posts: 952
- Location: Netherlands
- Has thanked: 407 times
- Been thanked: 422 times
Re: A vague treatise on influence
RobertJasiek wrote:Influence is not linear:
I don't see this as an example of non-linearity of influence, I think the influence map shows how that formation can effect the surrounding stones. I just don't think the influence map is the only variable in play.
That's why I never made provision for dead stones, the influence map shouldn't be the sole metric for the game state, but I think it can give information.
As far as the validity of the light model. I think the reason why the light model is a good place to start, is that light operates on a shortest path principle.
The limitation of this approach is that it assumes continuity, rather than computing shortest paths on a grid, (the shortest path may be degenerate, and the degree of degeneracy may have value)
But I think that kind of path-ing approach is a good start, and is approximated by the continuous approach.
Tactics yes, Tact no...
-
perceval
- Lives in gote
- Posts: 312
- Joined: Thu Aug 05, 2010 3:35 am
- Rank: 7K KGS
- GD Posts: 0
- KGS: tictac
- Has thanked: 52 times
- Been thanked: 41 times
Re: A vague treatise on influence
did you tried something like collaborative diffusion ?shapenaji wrote:So this is along the lines of your researches Robert,
but I remember a while back I was doing some influence maps of my own, using reflection of an influence-generating function off walls. (Using a method of images, take the board and make 8 reflected boards around it. Then compute influence for this "mega-board")
Ex:
X X X
X O X
X X X
(I believed that a high density of influence could be considered territory, and this would be an explanation for why a 3rd line stone is so much better than a 4th line stone for territory. The reflected portion in that same region is much higher, because of the exponential attenuation)
Another step I took was to have each stone contribute its influence to every point it could "see"(compute the relative angle of every group of stones, for each stone, find the closest group for each angle, and eliminate any points with a greater radius at the same angle)
I'm not sure the second part is accurate though. Influence ought to bend around groups,
In this way it should be more like a flow of water and less like a luminous source.
Any thoughts of how to improve these maps? I'd like to sit back down with them again...
http://www.google.com/url?sa=t&rct=j&q= ... xpdYRTrQMA
i learned of it while playing on the ants ai challenge, seems to be close to what you want: it does find "path" "around" ennemy stones.
i Should NOT ener this discussion but concerning your examples Robert, the influence map can be quite easily corrected to avoid the issue you mention by tweaking the diffused influence if the stones are alive or dead (or almost one or the other), and collaborative diffusion would take care of not diffusing the influence of a stone blocked by some others.
In theory, there is no difference between theory and practice. In practice, there is.