On the IA-Chess Analogy

Jessi Shakarian, writing in Medium:

When I picked up Information Architecture for the Web and Beyond, the so-called “polar bear” book, I didn’t expect to find a passion around chess. However, chess has become my lens of looking at information architecture in the real world.

In the book, the authors use chess is an analogy for information architecture — it’s a system of rules that doesn’t change based on where you play (on a wooden board in your living room, online against a friend across the country, or on an app on your phone).

The chess analogy is one of my favorite ways of explaining information architecture. As Jessi points out, the game has been around for a long time. Many people know about chess and — more importantly — are aware that it and its physical instantiation aren’t the same thing. As Jessi explains,

The player interprets the rules the same way regardless of a physical artifact or something in the digital space, and the value of the pieces on the board doesn’t change either.

In other words, you can play chess on different types of sets, and it is still ‘chess.’ This is also true of other old games such as checkers and backgammon. But chess is better known and widely acknowledged as a complex game of strategy, which make for a good illustration.

I first explored the chess-IA analogy in a 2009 email thread in the IxDA mailing list about the relevance of IA outside content-centric systems. The list’s archive doesn’t seem to be available online anymore, so I’m re-posting the relevant pieces here for reference.

In the thread, Dan Saffer wrote:

If you will agree that the field of information architecture is about organizing data/information so that it can be found and navigated through, most analog games don’t [fall] under this definition. Everything from chess to football to poker. But there is a lot of interactivity.

If you take a digital game like Simon <http://en.wikipedia.org/wiki/Simon_(game) > and interactive displays like Rosen’s wooden mirror <http://www.smoothware.com/danny/woodenmirror.html > there is no information architecture involved at all because there is no content to find or navigate through.

To which I replied:

The things that differentiate chess from (say) a pile of random pieces of wood on a table [are] precisely its information structures. Chess has a clear taxonomy (the different pieces, the colors, the layout of the board) and rules that define how those taxonomies interact. What makes chess in any way interesting is how the relationships between the items in that taxonomy vary throughout the game. I could go as far as saying that chess is primarily about information structures in a state of flux with each other.

The interactive elements of chess, on the other hand, are not core to its “chessyness”. This is illustrated by the fact that chess can be played with wooden pieces on a board, by correspondence on paper, by email, blindfolded, using a console terminal, by two computers playing each other using binary numbers, etc. (The same is true to football and poker, to a lesser degree.)

I’m not saying chess is only information architecture; I don’t particularly enjoy the game without its “tangible” UI (try playing http://www.craftychess.com/ using a terminal). But to say there is no IA there belies an incredibly closed-ended view of IA.

I went on to address some of Dan’s other points:

I agree with your stance re: Rosen’s wooden mirror, but that’s hardly a game, is it?

Simon, on the other hand, does have clear (if very simple) information structures. As with chess, they are what makes Simon different from a plastic cylinder with randomly blinking lights: there are only four colors (and not 19,202, for example), there are only four sounds, and there is a one-to-one relationship between these colors, sounds, and the buttons that the user interacts with. The rules of the game are also information: the fact that the sounds/lights are emitted in a random sequential order, and that said order is revealed incrementally one at a time. Someone designed these information structures for Simon. These are architectural decisions, and they deal with information being conveyed to the player. Information. Architecture.

However, Simon — unlike chess — is highly dependent on its interaction design. I would not be amused at all by Simon if I was playing it on paper, or on my computer screen. The “behavioral” aspects of the game are what make it successful.

My point: all these things have IxD and IA. I don’t know of anything that doesn’t to a degree or another. Even if we agree that IA is about organizing data/information, that is still a pretty big and pervasive area of concern!

Many things have changed since that discussion. One difference is the rise of content strategy, a discipline that focuses on content-centric systems. Another is a decline of interest in IA. These two issues aren’t unrelated. Given that IA was perceived to be mostly about content (as evidenced by that IxDA discussion), having another discipline focused exclusively on content has likely diverted attention from IA.

But IA and content strategy are closely related. (I see content strategy as applied IA.) And having a separate label for work on content-centric systems frees IA to focus on the high-level distinctions that all systems must present to users. Which is good, since the point I was making in that old thread — that all interactive systems have an IA, whether they’re perceived as content-centric for not — is even more relevant in our current multi-modal tech environment.

Cover photo by Marcobeltrametti (CC BY-SA 4.0)