For me, the mark of a talented game designer is someone who’s able to get to the point. The key skill is to distil complex ambitions into simple sets of instructions that an engineer can understand and implement.
Imagine a world in which tennis hadn’t been invented. You create a game involving two bats, a ball, a net, some lines and a scoring system. You proceed to play, but it turns out it’s not that much fun. As the designer, you might try a lot of balancing to do with bat strength, net height and so on, but it wouldn’t make the game much better.
It might be a while before you propose a win-by-two rule. The reason is that it’s a new rule, as opposed to balancing existing ones, and any time a game designer proposes adding something they will encounter resistance. Adding new rules means more coding, so you’d be asked to justify how it would be fun, because it would seem like a risk. That’s where clear communication is key.
Nobody, not even Shigeru Miyamoto, can really tell how a prototype will feel until it’s developed. You can write all the documentation you like, draw all the wireframes you please, but you still won’t know. There’s an X-factor. If you aren’t able to be crystal clear about what actions the player’s doll can perform and what rules enemies are supposed to follow, then you will not be able to find that X-factor and improve on it.
This is a similar problem to chaos theory. Using iterated rules and simulation software, mathematicians and scientists are able to create elaborate and often beautiful patterns. But they cannot usually prophesise what the simulation will do before they see it happen. They have to run it and observe, which is why it is still difficult to predict the weather.
The more I think on it, the more I realise games are all about chaos. Somewhere in the interplay between the agency of the player and the urgency of the scenario is this intermediate and volatile pattern. Its results can be beautiful, emergent things can happen, patterns of play and strategy can form. But we have to prototype it, then observe. So we should know more about chaos.
For a system to be considered chaotic, it must be sensitive to initial conditions. When developers talk about having a unique mechanic, they usually mean the player’s doll has some unusual ability, or the rules of the world empower it in some way. Gravity in Halo, for example, is intentionally floaty and that adds to the feel of play. Small changes to these rules have big consequences in-game.
A chaotic system must also have ‘topological mixing’. In lay terms, this means a constrained space in which the dynamic can repeat over and over, such as a chessboard, football pitch, PvP level or dungeon. Even in many open-world or action-adventure games, play is constrained into a range of tightened environments over the course of missions. This leads to variable action within that space, which has thrilling results in a good game.
Third, a chaotic system must have ‘dense periodic orbits’, which means being iterated with lots of repeated actions that build on one another to form interesting outcomes. This describes most successful games, from the simplest session of Tetris through to the complex back and forth of a Diablo III battle. Games deliberately limit the kinds of action that you can take, so they are often repetitive and the results are indeed interesting.
If your dynamic conforms to these rules, you have a great chance of making a conventionally good game. Will it be special, though? Not unless it has something extra: a ‘strange attractor’.
A strange attractor is an ideal point or shape toward which a dynamic system moves over many iterations. It arises as a result of one or more rules (let’s call them ‘strange rules’) applied to the system, which constrain or mould it in certain ways. Various exotic fractal images (such as the ones popular in early ’90s dance culture) are the results of strange rules. They give rise to beauty.
Strange rules are also the key to great game dynamics. That win-by-two rule in tennis is one example. That one rule changes the whole flow of the game, leading to many interesting decisions. The offside rule in soccer is another example, as is the single forward pass in NFL football.
Sometimes strange attractors arise from a couple of rules. Arguably, Grand Theft Auto‘s greatness lies not just in car-steering physics, but in the rule governing when police chases activate. A game using the same steering engine but without the police factor (LA Noire) is much duller.
The big challenge of game design could then be described as this: the more we try to describe dynamics, the less we are able to. They are ineffable. But what we can do is be much clearer on what we think our strange rules might be.
Strange rules are usually simply expressed, meaning players can comprehend them. In chess, you must announce a checkmate, for example. That can be explained in five words, but changes everything. A great game designer can find that rule and express it just so. Bad game designers make the rule too complicated, or get lost in endless documentation, justification and doubt.
So if you’re designing a game, look for the strange rule. But make it as simple as possible.