Page: DesignArticle.Article-ChristopherAlexander - Last Modified : Mon, 16 Jul 12 - 5898 Visits

Author : Eric Pietrocupo

How it all started

I read a book called "The art of game design", it was a very interesting book and I even made a summary of it on this web site. In one of the chapters, the author talked about Christopher Alexander's theories and publication. One of the theories is that there are 15 properties that are shared by beautiful or well designed objects.

At first, I did not react to the theory, but when analyzing the properties in details I realized that many properties were actually present in my board game Fallen Kingdoms. Since my game had some of these properties, I thought that they were somewhat important to me since I unconsciously implemented them in my game.

So I decided to study the theory in depth because it could be the key to make a good game. So being consciously aware of these properties could help me make better games and avoid pitfalls.

Who is Christopher Alexander

There is probably more information on the Wikipedia page, but in summary he is an architect that has written many books about the "philosophy" of architecture. The book which I am going to talk about is called "The Nature of Order", it's published in 4 volumes and the 15 properties are located in the first book named "The phenomenon of life". The nature of order is much more accessible than the previous books which are much more abstract.

Christopher Alexander thinks that there are reasons why something is considered beautiful whatever are the personal preferences. There are rules set in nature that makes things beautiful or well designed, but these rules have been broken in 20th century architecture. So this is why all these new buildings are considered ugly on his point of view.

Christopher analyzed nature for more than 30 years before clearly defining 15 properties that are shared by beautiful or well designed objects. It does not mean that all object have all the 15 properties, they might have 2 or 3 properties, but they will avoid going in contradiction with these properties.

What does it have to do with board games

The properties of Christopher Alexander can be applied to anything created by man. It can be a building, pottery, tapestry, garden, music, etc. In his book, many illustrations are not actually buildings. Since games are also a creation of human being, the 15 properties could probably also be applied.

The graphic design of the game will follow some of these rules. Since most examples in the nature of order are applied to the visual aspects of the buildings and object. It would not complicated to transpose these properties to the graphic design of the game.

But I want to push the concept further by applying the 15 properties to the mechanic design. So game that use mechanics going in contradiction with these properties will make the game ugly. Personally, I think they are a lot of games out there that goes in contradiction with these rules. I often say that sometimes mechanics are not "elegant", I think that by "Elegant" I could mean "Goes in contradiction with one of the 15 properties".

Example of ugly mechanics

I'll take 2 example that should normally jump in the face. You only need to look at the game to know that there is something wrong going on. No play testing is actually required.

Hecatomb: This is a collectible card game designed by wizard of the coast that features pentagonal plastic cards. The cards are transparent so that you can easily stack them and see some information of the bottom cards. It's pretty unique an ingenious. Here is a picture of how the cards are used:

http://www.boardgamegeek.com/image/968707/hecatomb

Hecatomb has 4 suits/type/sphere of cards that player can collect to create their deck. In magic the gathering, one of the first CCG, there was 5 suites and it proved to create an interesting dynamic since it was huge success. But I don't think it would have been really odd if they would have used 4 suites instead like earth-fire-water-air.

I understand that sometimes people want to do something different, but hecatomb was not the place. The cards have 5 sides, why in hell would there be only 4 suites? This game is screaming to have 5 suites. With only 4 suites, it's simply inelegant.

Guess what! they eventually released an expansion that added a 5th suite. What a coincidence! Unfortunately, it was too late, the wrong was already done.

Advanced Dungeons & Dragons 2nd: In the 2nd advanced edition, each character can have 6 different stats ranging from 3-18 since they were determined by the roll of 3D6. But still, since there was artifact that could increase the stats of the characters beyond 18 and because there were also divine characters, the stats table were designed up to rank 25.

The only exception is the strength table. When a character had exactly 18, he rolled a d100 for exceptional strength that allowed a character to advance even more in the strength table. So the stat 18 was split up in half a dozen of level. But the worst part, is that if a character have a stat of 18 and get +1 strength, he does not jump to the next rank in the table, he jumps to 19 which is like half a dozen of steps higher in the table.

A copy of the table can be found here:

http://www.ancientscrossroads.com/adnd_tools/str_table.htm

This is not elegant at all, it completely breaks the symmetry with the other 5 attributes. If the strength table was too big, they could simply compress the table to make in fit into 25 entry.

The 15 Properties

I'll first list and describe the 15 properties like shown in the original book and then I'll try to adapt these rules to board game design.

The author say that objects are made of "centers". An object can have multiple centers which together create the wholeness. These center must share some of the properties below to make the wholeness beautiful or well designed. Centers also strengthen each other to make the wholeness better.

1. Levels of Scale: There must be centers of different size within an object. The difference between the center's size must not be too large or too small.

2. Strong Centers: Every object must have a center which is stronger or more important than the others.

3. Boundaries: Centers can be defined by borders or boundaries to separate them from other centers.

4. Alternating Repetition: Repetition of centers is beautiful if their position alternate with the position of other center. For example, Alternating objects in an hex grid style placement is better than in a square grid shape style.

5. Positive space: A shape has positive space the negative image of the object creates an interesting shape or prevent wasting space. For example, a floor made of octagonal tiles creates a square between the tiles, so the space is positive.

6. Good Shape: A good shape is composed of multiple centers which are good shapes themselves. Elementary figures are good shapes, so another way to explain it is that any shape composed of elementary figures is a good shape.

7. Local Symmetries: Some centers must be symmetric, but excessive symmetry is not better. So it's better to make multiple and different symmetries than repeating the same symmetry over and over again.

8. Deep interlock and ambiguity: This is when 2 or more patterns expand mutually inside another pattern making both of them bound together. The Yin-Yang symbol could be a small example of interlock. The idea is to link centers together.

9. Contrast: Having centers which are opposite to each other creates something new out of the contrast. The most common use is to mix contrasting colors.

10. Gradients: Centers must change or progress across the space. It can be created by a series of varying centers.

11. Roughness: Imperfection is better than perfection because the structure of the object is more important than the perfect placement of the elements.

12. Echoes : Centers must share a common property between each other to feel that they fit with each other.

13. The void: There must be a place where there is nothing, or where there is a low amount of details.

14. Simplicity and Inner Calm: Any center which is unnecessary or which does not enhance other centers should be removed.

15. Not Separateness: The centers must melt with it's surrounding centers, it must not be isolated.

Example: The flower

Let's try with a simple example. There are reasons why people find flowers beautiful. I'll make two simple drawing of a flower where one is beautiful and one is ugly. I'll list how some of the properties above can be applied to these drawings:

1. Level of Scales: On the left pictures, the outer petals are larger than the inner petal. While on the right side there is not much difference since both petals have approximately the same surface.

2. Strong Centers: The large circle in the center of both pictures indicates that they each have a strong center. The fact that all other petals are attached to the inner circle reinforce it as a strong center.

4. Alternating Repetition: On the left picture, the smaller petals are located in between the larger petals. So there is an alternation between a large and small petal. While in the right picture, they are both located on the same position, so no alternation at all.

6. Good Shape: Both pictures are made of primitive shapes like circle, rectangles, and ovals.

7. Local symmetry: The left picture, having 8 petals, allows various symmetric possibilities. The flower can be mirrored on many axis but that is impossible on the right picture since there is an odd number of petals.

9. Contrast: On the left picture there is a contrast between the large petal and the center of the flower since blue and orange are contrasting colors. But on the right side, the colors does not create any contrast.

12. Echoes: The left picture use only round shapes, so this is the property that links all elements together. While on the right picture, this link is broken since the outer petals are made of rectangle.

13. void: Both center circles are empty of details, it could create some sort of void. Put the picture is too small to see the effect of the void.

Reinterpretation for Games

Now what I want to do is to re-interpret these properties to apply them to game mechanics. So the list below is more a draft of how these properties could be applied. Some properties are easier than other to conceptualize, so the section below is somewhat incomplete. I'll gather comments and eventually build a guide where I will analyze each property in details with pictures and examples.

1. Levels of Scale: This could applied to the importance of the mechanics, it could mean that some mechanics are more important than others. Importance could be measured as:

• The amount of time spent to resolve the mechanic
• The amount of space taken on the board
• The impact of the mechanics on the game

Indirectly, is also means that games must have more than 1 mechanic to have levels of scale. So it is somewhat related to my side dish theory where a good mechanic needs to be accompanied by a few side mechanics to make the core mechanic better.

2. Strong Centers: This could means that there must be a strong core mechanic in the game that is more important than the others. Avoiding games where people are feeling that they are playing 2 different games in the same game.

3. Boundaries: It could means that a game must have restrictions and limits. One example that jumps to my mind is when one of my friend wanted to make games where you could play anywhere on the board. The problem is that people don't know what they are doing until the board is populated enough to see the consequences of their moves.

4. Alternating Repetition: I think the basic advice would be to avoid plain repetition. I had that problem in Fallen Kingdoms where every turn felt the same. So I decided to use use a gradient to make the game change as it progress.

5. Positive Space: This could be very hard to apply to a mechanic. It would means that a mechanic must have a "good space" around it to allow other mechanics to be adjacent to it. But since mechanics does not have a physical location, it might be hard to visualize. I think that if a mechanic is so intrusive that it prevent other mechanics from working, then there is a positive space problem.

6. Good Shape: Again, very hard to apply to mechanics. The basic idea is that a good shape is made of elementary shapes. Maybe the same logic could be applied: a mechanic must be made of elementary mechanics. But now the question is: Is there such things as elementary mechanics? Can a die roll or combat resolution table be an elementary mechanic? I also don't have an example of a "Bad Shape", so it's hard to compare and see the difference.

7. Local Symmetries: Symmetry is a property that I have used intensively in any design I made. For example, in Fallen Kingdoms, the number 3 can be found in many situations ( 3 buildings, 3 production tracks, 3 type of trophies, etc.) It could be applied as "avoid exceptions" because that has the effect of breaking the symmetry. The AD&D strength table example illustrate the lack of symmetry.

8. Interlock and ambiguity: I think that could be viewed as the shape of a mechanics allows another mechanic to expand. Like if 2 mechanics were dependent on each other. The first example that comes to my mind is in Settlers of Catan where the board setup will give more space to certain mechanics. For example, if a player is trapped or if there is a low amount of brick in play, development cards will be more important. The idea is that certain mechanics force other mechanics to react differently because they are interconnected.

9. Contrast: Having asymmetrical forces like in Starcraft could be a form of contrasts. Having players doing opposite things like building stuff and then destroying stuff could be another form of contrast.

10. Gradients: Games that have multiple phases like Civilization and 19xx have some form of gradient over time because the game changes and evolves. Using Number sequences (ex: triangular: 1,3,6,10,etc) could also be another way to apply gradients. As for breaking gradients, I always found that games which have scoring phases on specific turns breaks all possible gradient because it did not felt natural.

11. Roughness: Having a series of special power that are not exactly balanced is a form of roughness. I think this is best applied when there is large number of elements. For example, special powers of action cards is good if there is a lot of actions cards. But it players have 5 different actions, the amount of imperfection must be much lower else it will make some actions useless.

12. Echoes: This means that each mechanic must share some common properties. The first way to apply this could be that a game must use mechanics in the same category. So for example, trying to make an Auction-Dexterity-Wargame might not really be a good idea. On another hand, it could simply be a link between mechanics. For example, in Fallen kingdoms, there are 3 types production track which are related to 3 types of resources and buildings.

13. Void: This one is hard to implement since in theory it must be a simple mechanic that occupies a relatively good amount of space. But it's hard to define how much space a mechanic takes. It could also be seen as a lack of information rather than simple mechanics. Maybe and hex map that contains little information other than terrain could be a void.

14. Simplicity and inner calm: I think this is commonly used by game designer, it's the idea of simplifying the game. The idea is not to make only simple games, but rather remove useless complexity that has no impact on the game.

15. Not Separateness: This means that a mechanics must fit with it's surroundings. The best would be to make sure that a mechanic is not isolated from the others. It must not feel like if the players are stopping the game to do something else. Probably the Auction-Dexterity-Wargame goes in contradiction with "Not Separateness". I think increasing relations and interactions between mechanics will help improve the not separateness.

So this is it for now, if you have new ideas or example that reflect the properties positively or negatively, please let me know. The next step will be to gather ideas, comments, examples and make a more detailed guide about it.

Resources and Links

Here are a few links about Christopher Alexander's theory

Wikipedia: Christopher Alexander
http://en.wikipedia.org/wiki/Christopher_Alexander

Website dedicated to his books
http://www.natureoforder.com/
http://www.patternlanguage.com/

Applying the 15 Properties to other fields
http://www.truthtable.com/Fifteen_Princples.html
http://www.guyobrecht.com/15properties/front.html

Explanation of the 15 properties with some pictures of the book.
http://www.livingneighborhoods.org/ht-0/fifteen.htm

Christopher's Alexander theory of centers
http://www.regismedina.com/blog/2008/09/christopher-alexander-theory-of-centers

Discussion Links

This is the link of the threads where you can talk about the article.

Board Game Designer Forum
http://www.bgdf.com/node/6813

Board Game Geek
http://www.boardgamegeek.com/article/9610854#9610854

---

Back to the Game Design Article List