Ivanhoe and Game Design

by Bethany Nowviskie

Jerry and Johanna invited me to join them this morning as the Ghost of Ivanhoe Present, to talk about the current state of the game and what's happening in our design meetings this semester. But I'd rather be the ghost of the conditional, so instead I'm going to predict a bright future for Ivanhoe if we do the kind of homework I'm about to lay out.

But to discharge my duty, here's what's happening: In a nutshell, we're implementing the game in a complex digital environment meant to organize and display player moves. That involves analyzing the underlying model, refining our notions of what gameplay in this context involves, and building a software framework which we call the Ivanhoe Game, but which is really meant to facilitate the game. What we sometimes mistakenly call "the game" is in fact a locus for Ivanhoe games.

Some other folks here this morning are much more involved in the practical undertaking than I am and can give you an idea of the various components they're working on and what our timetable is. I hope they'll jump in in a minute.

But I want to focus for a bit on how we can better understand what we're doing in our weekly meetings. How do we position ourselves in the field of game design, and what are our models for making Ivanhoe?

I'm going to tell you about a couple of these models this morning, less because I want to inform you and more because I'm hoping you can suggest some more avenues of exploration -- things we haven't considered yet.

The first of our predecessors is Nomic, which is a relatively new game invented by the philosopher Peter Suber. It was first described in a 1982 issue of Scientific American as "a game of self-amendment."

Nomic works like this. You start with small rule set. A few of the rules are immutable. For example: "All players must abide by the rules currently in effect." But most of the rules are mutable. A mutable rule might read: "Rules may only be amended by a unamimous vote."

The object of a Nomic game is collectively defined as the rules mutate. The subject of the game, interestingly, is up for grabs, too. At heart, it is its own subject, but there are instances of Nomic where the players have decided on certain themes, or to use the game to model certain real or imaginary systems. There's even one (I think defunct) game called "Garden Gnomic."

Ivanhoe, as I see it, is like Nomic in two ways:

1) First, we had so much fun writing and re-writing and fighting over the rules that it became obvious that rule creation and amendment needed to be built into the gameplay of Ivanhoe in a major way. The rules, for advanced players, can't all be immutable. It's only when we start to question the rules that we reach the real levels of self-awareness that Ivanhoe is supposed to foster.

2) The second way Ivanhoe is like Nomic becomes evident when you consider what constitutes "the rules." In very real terms, the rules are the entire environment of Ivanhoe. Self-amendment is the name of the game. The first rule of our Wuthering Heights session wasn't on the order of Rule #5.4.2, which determined the penalty for losing a challenge. It was much more simple, and was asserted not by the game designers, but by W.W. Norton and Company: "This is Wuthering Heights." Every move we made was an amendment to the discourse field surrounding that work as legislated by that particular edition.

We're looking at other games and classes of games as well. Interactive fiction, of course, is one, and we've hired a research assistant to outline some connections for us there, and with other text-based and shared-world forms. But we're also interested in more conventional game organizations. We had a minor revelation this week when Geoff Rockwell suggested we study the difference between a turn-based game like Monopoly and a game called Diplomacy, in which all players act simultaneously and let the game structure mediate when they step on each other's toes.

Of course, there is a thoroughly articulated field that analyzes play and conflict -- and that's not psychotherapy, but rather game theory. There are some concepts of agency swirling around mathematical game theory that could be useful to us in thinking about the role of the computer in reconfiguring the game state and forcing response from the players. The computer as a player -- a Denettian agent. This makes the programmers among us look incredibly nervous, but we're not really broaching an AI question here. It's more of a conceptual exercise in thinking of the computer as being just another agent or player in the little Prisoner's Dilemmas our users will undoubtedly set up for themselves. Imagining the random interventions Jerry and Johanna envisioned on the part of the machine as the actions of an irrational agent may free us up to apply some of the work game theorists have done in mechanism design.

Actually, psychotherapy's not that far off. There are some things Ivanhoe will probably never get close to, like behavioral psychology, but which have a lot to say about how to fashion systems to elicit certain responses, and how to use player response as part of a feedback system in the game design process.

A lot of that sort of discussion is found in the growing literature on digital game design, which I'm mining right now. The last model I want to mention comes from what the present game industry thinks of as its prehistory -- 1983. At that point, after Pac-Man, after Zork, a game designer named Chris Crawford realized that the industry would need a shared vocabulary and a common concept of gaming in order to make more than technical or aesthetic progress. That is, he wanted to advance the concept of game play and therefore needed to know what games are. And what they're not.

I'm going to talk about what they're not.

Games are not toys. Will Wright, the creator of SimCity and The Sims, is always careful to state that his creations are not computer games, but rather "software toys." They're more like basketballs than like basketball games. A basketball offers many obvious behaviors ("dribble me, throw me, slam-dunk me") and invites you to discover more: "twirl me on your finger like a Harlem Globetrotter." But there's no game built into a basketball. The game is a set of player-defined objectives that employ the toy and require using it to meet a goal. You can't win The Sims objectively, but you can have a personal victory.

Here's another one. Games are not puzzles. Game designers see jigsaws and crosswords as essentially static logic structures in which clues help you make changes to the game state with an eye toward reaching the ultimate stasis -- completion. A game (like chess) might contain or spawn puzzles -- chess endgames, for example. But games are fundamentally different because of their interactive or responsive quality. In a game, an opposing agent (whether human or machine), responds dynamically to your action. It's a puzzle if I'm putting a jigsaw together alone. It's a game if we're doing it together and you're trying to make a different picture.

The last two kinds of widely-accepted not-games are more problematic for Ivanhoe.

Games are not stories. Of course, there's a fundamental difference between playing any game and reading or viewing a linear narrative. But how do you understand this when you're making a game of storytelling?

And games are not simulations. That is, they're not serious attempts to represent accurately a real phenomenon in another, more malleable form. (Like Flight Simulator) But aren't we actually trying to model academic discourse and critical thinking for students? Although I heard them denied this week, the pedagogical aims of Ivanhoe have always seemed fundamental to me, and teaching is often simulating.

In fact, a game is an artistically simplified representation of a phenomenon. Chris Crawford tells us (in The Art of Computer Game Design) that "the simulations designer simplifies reluctantly and only as a concession to material and intellectual limitations. The game designer simplifies deliberately in order to focus the player's attention on those factors the designer judges to be important."

Eliciting and playing on that kind of focused attention is precisely what we're doing with Ivanhoe. And Crawford's simple statement is more than a philosophical distinction. It's a procedural imperative. I'd argue that it's this kind of insight -- the kind that we can gain from gaming professionals and from solid research in an array of related subjects -- that will help us implement the game in a digital environment.

[back]