Arimaa Forum
        (http://arimaa.com/arimaa/forum/cgi/YaBB.cgi)
      

        Arimaa >> Off Topic Discussion >> Bottomless wonders spring from simple rules
        
(Message started by: clojure on Oct 16th, 2010, 1:51pm)
Title: Bottomless wonders spring from simple rules
Post by clojure on Oct 16th, 2010, 1:51pm
As you probably know, Benoit Mandelbrot, known for Fractals, passed away.

Here's a quote from him in a TED talk: "Bottomless wonders spring from simple rules, which are repeated without end."

http://www.ted.com/talks/lang/eng/benoit_mandelbrot_fractals_the_art_of_roughness.html
Title: Re: Bottomless wonders spring from simple rules
Post by rbarreira on Oct 16th, 2010, 3:19pm
Sad to see him go...

If you don't believe the thread title, just see the beautiful patterns of the Mandelbrot set (which you can mathematically describe in one short sentence):

http://en.wikipedia.org/wiki/Mandelbrot_set

Cool animation (beware, 24 MB gif file)

http://upload.wikimedia.org/wikipedia/commons/a/a4/Mandelbrot_sequence_new.gif
Title: Re: Bottomless wonders spring from simple rules
Post by MarkSteere on Oct 19th, 2010, 4:19pm

on 10/16/10 at 13:51:23, clojure wrote:
"Bottomless wonders spring from simple rules, which are repeated without end."

Benoit certainly was not talking about game rules, but I can think of two games his statement describes: Dots and Boxes, and Flume. In Flume, if you form orthogonal adjacencies with 3 or 4 stones of any color, move again.  And repeat.

http://www.marksteeregames.com/Flume_Go_rules.pdf

It's easy to view one's game as having "but one rule" as Christian has demonstrated for us.  Flume really does have but one rule.

Dots and Boxes is known for its extreme computational complexity.  Flume's simpler rule set generates a larger (more complex) game tree.  The long chain parity principle of Dots and Boxes does not have an equivalent in Flume, depending as it does on the fixed wall widths of Dots and Boxes.

Dots and Boxes really is one of of the very best modern games.  It's not a kids' game like Checkers.  The "problem" with D&B and its improved counterpart Flume is that there's a big learning curve.  You have to invest hours of clueless play, but there's a big payoff.  You become skilled at a very robust, scalable game.

Oust is currently considered to be my best game, but as the decades pass I believe Flume will emerge as equally significant.
Title: Re: Bottomless wonders spring from simple rules
Post by clojure on Oct 19th, 2010, 8:49pm

on 10/19/10 at 16:19:23, MarkSteere wrote:
http://www.marksteeregames.com/Flume_Go_rules.pdf



Quote:
MULTIPLE STONE PLACEMENT When you form 3 or 4 connections (orthogonal adjacencies)


Do you mean non-orthogonal, or diagonal adjacencies (I'm no native speaker)?

I think the colors should be chosen differently, just because the usual color blindness problem. Especially if green circles are not part of the board (if this was manufactured).

You also need to restructure your writing. It's too chaotic. For example, "Color doesn’t matter when placing stones. Red stones, blue stones and the green "stones" surrounding the board are all treated the same. Flume uses the pie rule."

How does that last sentence relate to the one above it. It's awkwardly put.

I also think that the introduction part should include the goal. The introduction leaves some questions delayed, and is even contradictory with the goal ("There will always be a move available"). If the goal was part of the introduction, you could make it clearer.

Also there's problem with already mentioned "Color doesn’t matter when placing stones. Red stones, blue stones and the green
"stones" surrounding the board are all treated the same"

This is redundant (especially since it's in its own paragraph). If you changed the picture with purple stones to have different colors but same form, it would be enough.

Well, I think it's enough for the morning. What I intend to say is: if you have enough time to design those games, you should also think how to present them in the best possible ways.

You get plus points for examples and clear pictures.


Quote:
It's not a kids' game like Checkers.


8)
Title: Re: Bottomless wonders spring from simple rules
Post by rozencrantz on Oct 19th, 2010, 11:40pm
Mandelbrot's set in particular is exciting because it shows so much more variety than textbook recursive fractals like Sierpinski's, but its rule is not very much more complicated.

To me, Mandelbrot's set is the most beautiful and complex fractal with a generator I can (just barely) understand. I find L-systems fairly boring, and Dynamical systems are completely beyond me.

Of course there's a certain amount of subjectivity. Mark and I seem to have nearly opposed ideas of simplicity: Playing Dots and Boxes to me feels like an adventure into the baroque; even though I'm terrible at it, every new bit of Arimaa strategy I learn feels like a very natural extension of what I already knew.
Title: Re: Bottomless wonders spring from simple rules
Post by MarkSteere on Oct 19th, 2010, 11:56pm

on 10/19/10 at 20:49:18, clojure wrote:
Do you mean non-orthogonal, or diagonal adjacencies (I'm no native speaker)?

"MULTIPLE STONE PLACEMENT When you form 3 or 4 connections (***__ orthogonal adjacencies __***)..."  


on 10/19/10 at 20:49:18, clojure wrote:
"Red stones, blue stones and the green "stones" surrounding the board are all treated the same. Flume uses the pie rule."[/b]

How does that last sentence relate to the one above it. It's awkwardly put.

This is the closest thing to a valid complaint in your post.  The last sentence in said paragraph does not relate to the one before it.  There's somewhat of a premium on fitting the rules onto one sheet, and that wouldn't be possible if I had to give the pie rule statement its own paragraph, not without shifting down to a smaller font.  


on 10/19/10 at 20:49:18, clojure wrote:
I also think that the introduction part should include the goal.

There isn't room and it would cause an unnecessary postponement of the stone placement rules.  It's an extremely simple matter to glance down to the middle of the single page rule sheet where it says (highlighted in red),

"OBJECT OF THE GAME When the board fills up, whoever has the most stones on the board wins."


on 10/19/10 at 20:49:18, clojure wrote:
The introduction leaves some questions delayed, and is even contradictory with the goal ("There will always be a move available"). If the goal was part of the introduction, you could make it clearer.

Implicit in the statement "There will always be a move available" is the phrase "while the game is in play".  By definition, when the game has concluded, there won't be any more moves available.


on 10/19/10 at 20:49:18, clojure wrote:
Also there's problem...

[snip redundant complaint]

(Don't look now but you missed a definite article.)


on 10/19/10 at 20:49:18, clojure wrote:
This is redundant

The redundancy is intentional.  It tends to reduce the possibility of the reader making it all the way through the rule sheet with no understanding of the rules, which you seem to have managed anyway.


on 10/19/10 at 20:49:18, clojure wrote:
Well, I think it's enough for the morning.

I think it's quite enough for any time of the day.


on 10/19/10 at 20:49:18, clojure wrote:
What I intend to say is: if you have enough time to design those games, you should also think how to present them in the best possible ways.

I do.  And the best possible way is the most accurate, thorough, and redundant way - not necessarily the clearest but most ambiguous way.


on 10/19/10 at 20:49:18, clojure wrote:
You get plus points for examples and clear pictures.

Thank you.  The diagrams are more than just window dressing.  They're specifically designed to answer anticipated questions such as "What happens in xyz situation?"  Xyz situation will probably appear in one of the diagrams.
Title: Re: Bottomless wonders spring from simple rules
Post by clojure on Oct 20th, 2010, 4:14am
Let's start with this.

on 10/19/10 at 23:56:24, MarkSteere wrote:
"OBJECT OF THE GAME When the board fills up, whoever has the most stones on the board wins."

Implicit in the statement "There will always be a move available" is the phrase "while the game is in play".  By definition, when the game has concluded, there won't be any more moves available.

[snip redundant complaint]

(Don't look now but you missed a definite article.)

The redundancy is intentional.  It tends to reduce the possibility of the reader making it all the way through the rule sheet with no understanding of the rules, which you seem to have managed anyway.


First of all, it's different matter to create a rule sheet from giving some ideas how to improve it. I assumed that you get what I mean without being too accurate.

Also, it's sure possible that I didn't understand the rules but you didn't explicitly point out to me, in what case I did so (don't confuse it with the presentation disagreements). You must also consider the possibility that I did understand but it's possible that I didn't communicate it properly.

About the redundancy. Yes, valid point, and I did address that. If you had thought about it, I approved your pictures which are a form of a redundant information, agreed?


Quote:
"MULTIPLE STONE PLACEMENT When you form 3 or 4 connections (***__ orthogonal adjacencies __***)..."  


Ok. You are probably right since you insist but I must warn you that if you target your game to a wider audience, you might encounter problems with understanding that sentence without the help of guessing from the context. For me the pictures make clear what is happening. That sentence is making me confused. Could you at least describe what that means in formal way since informality is so ambiguous for me.

I couldn't find with my google-fu a proper definition but all the uses of "orthogonal adjacency" mean top-to-bottom or left-to-right adjacencies. Also Fritz's comment on a recent review on Arimaa agrees with my conception "Diagonals never count in Arimaa; "adjacent" always means "orthogonally adjacent"."

Now, the other option is that it is how you say, and how I interpret but that the pictures are wrong.

http://imgur.com/6wuBIl.jpg

In that picture, not a single stone is adjacent orthogonally to other stone according to what I've read.

Please explain?


Quote:
There's somewhat of a premium on fitting the rules onto one sheet, and that wouldn't be possible if I had to give the pie rule statement its own paragraph, not without shifting down to a smaller font.  


Ok. It's a valid reason but changing words from here and there, you probably could get away without changing the style.


Quote:
I do.  And the best possible way is the most accurate, thorough, and redundant way - not necessarily the clearest but most ambiguous way.


I think we somewhat agree but I should say that clearest explanation would be non-ambiguous.


Quote:
They're specifically designed to answer anticipated questions such as "What happens in xyz situation?"


Indeed.

About implicit rules. My intention was to say that one should minimize the amount of them since the reader need to carry those implications in his head. This is emphasized when those rules clash together (they are restricting part of a rule in distinct places). It's better to be explicit and think about how to relieve the pressure from the reader by forming paragraphs that are only loosely dependent on paragraphs that come later.
Title: Re: Bottomless wonders spring from simple rules
Post by MarkSteere on Oct 20th, 2010, 7:51am

on 10/20/10 at 04:14:53, clojure wrote:
In that picture, not a single stone is adjacent orthogonally to other stone according to what I've read.

Please explain?

Ok, I see your point of confusion.  The green check marks are where you'd place a stone to form 3 or 4 adjacencies.  I'll see if it's possible to clarify that in the available space, without redoing the whole rule sheet.
Title: Re: Bottomless wonders spring from simple rules
Post by MarkSteere on Oct 20th, 2010, 8:38am
Ok, try this:

http://www.marksteeregames.com/Flume_Go_rules.pdf

In fairness, you're not the first person to express confusion with the Flume rule sheet.  
Title: Re: Bottomless wonders spring from simple rules
Post by MarkSteere on Oct 20th, 2010, 8:51am

on 10/19/10 at 23:40:23, rozencrantz wrote:
Mark and I seem to have nearly opposed ideas of simplicity: Playing Dots and Boxes to me feels like an adventure into the baroque;

Baroque definition: "extravagantly ornate, florid, and convoluted in character or style"

This describes both the fractal and Flume game tree, both of which are generated by simple, recursive rules.  Where's the disagreement?  :)
Title: Re: Bottomless wonders spring from simple rules
Post by clojure on Oct 20th, 2010, 9:58am

on 10/20/10 at 08:38:32, MarkSteere wrote:
Ok, try this:

http://www.marksteeregames.com/Flume_Go_rules.pdf

In fairness, you're not the first person to express confusion with the Flume rule sheet.  


I see it now. I also had read a bit too carelessly, and understood 3 or 4 connections as 3 or 4 stones in connection.

It's not so weird to get confused with the mark you have used since it's meaning is so deeply attached to "correctness" right from first year in school. I would hope you instead use a dashed or dotted circle.

I still have lots of complaints about the writing and structure but it seems you don't welcome my input in that regard. It's fine. Your game, your rules. :)
Title: Re: Bottomless wonders spring from simple rules
Post by MarkSteere on Oct 20th, 2010, 10:21am

on 10/20/10 at 09:58:09, clojure wrote:
I would hope you instead use a dashed or dotted circle.

Ok, good point - now incorporated.

http://www.marksteeregames.com/Flume_Go_rules.pdf


on 10/20/10 at 09:58:09, clojure wrote:
I still have lots of complaints about the writing and structure but it seems you don't welcome my input in that regard.

I'll consider any specific suggestions you may have.
Title: Re: Bottomless wonders spring from simple rules
Post by rozencrantz on Oct 21st, 2010, 12:40am

on 10/20/10 at 08:51:03, MarkSteere wrote:
Baroque definition: "extravagantly ornate, florid, and convoluted in character or style"

This describes both the fractal and Flume game tree, both of which are generated by simple, recursive rules.  Where's the disagreement?  :)


Maybe we don't disagree after all, what you say there I do not disagree with. I was perhaps thinking sloppily, calling games with a smoother learning curve simpler. I certainly perceive them that way, but I should know as well as anyone that perceptions are not reality.
Title: Re: Bottomless wonders spring from simple rules
Post by omar on Oct 26th, 2010, 10:34am

on 10/19/10 at 16:19:23, MarkSteere wrote:
Oust is currently considered to be my best game, but as the decades pass I believe Flume will emerge as equally significant.


I had a chance to try Flume on igGameCenter shortly after you released it. I really liked it.
Title: Re: Bottomless wonders spring from simple rules
Post by MarkSteere on Nov 6th, 2010, 8:34am

on 10/26/10 at 10:34:42, omar wrote:
I had a chance to try Flume on igGameCenter shortly after you released it. I really liked it.

Thanks Omar  :)


Arimaa Forum » Powered by YaBB 1 Gold - SP 1.3.1!
YaBB © 2000-2003. All Rights Reserved.