Game Grammar

I may end up saying the obvious here. I rewrote this talk three times, because I was told I'd be following someone who came out of a coma thanks to games... when i asked if I should tilt the talk more inspirational, they said "do the talk you want to do." Turns out that talk is dry and academic and full of diagrams. But there aren't any other sessions running right now, so tough, hope you like crunchy design talks!

This is the typical interaction loop that we've been using forever now.

I actually think it’s better to think this way… games almost like a piece of stained glass or something in between two people, mediating communication.

Both the ball and chess, however, have a set of common characteristics that enable play, which is that machine-like quality. In the case of the ball it’s manifested in physics… I am going to stick to game design more than play design here, but they really only have this one difference: the predefined objective...

Usually we say toy for the ball And chess is a game (or an “orthogame” if you use the Garfield et al)

There’s a common third type people speak of, which is puzzle...

Thing is, we can break chess into three parts The input mechanism The chess black box or machine And the feedback or visualization layer

The chess AI, though, is NOT in the black box Game grammar thinks every game always has an “opponent”

So there’s a core there. But let’s look first at how we interact We actually tend to have nested goals as we play Immediate goal, and long term objective, and we fill in or change interim goals as we go AI research actually formalizes this approach as a huge decision tree

Of course, in practice during gameplay we have to deal with pans going awry, and contingency plans...

Given an intent, we then have to march through this set of intermediate barriers: Intent is a verb.

The verb has to match an affordance

An affordance has to have an input

The input needs to have a command mapping or action The action goes into the black box

A major game grammar insight is that “games nest.” The classic interactivity loop applies just as much to a single button as it does to the overarching objective. In fact, it applies to every intermediate goal. In other words, games are made of games.

Take a racing videogame. Some folks like to say that “the game is the interface” (or the other way around). In some types of games, this is literally true; consider the racing game that shows you the optimal path at all times. The skill exercised is mostly one of applying the correct pressure to controls

Because the control is a game, it needs it own feedback loop. You need to know A) whether you executed the input (wheel turns) B) whether the input executed the action you were attempting Broadly speaking, we can call this “haptics” and “feel” When this type of task is completed well, we can often feel visceral fun, one of the four types Lazzaro identified.

Because of the granularity of physicality, the “turns” in games with a visceral or haptic loop are typically very fast and you operate on a time limit and miss your turn if you don’t execute. Effectively, it’s like phase-based combat.

We now have one subloop of our map of game. Funny how the simple act of moving a chess piece is so complicated! Feel alone is the subject of an entire excellent book by Steve Swink Now let’s talk about that black box. I am not going to spend a ton of time on fun and learning because I wrote a whole book about it.

This means that ludic systems have a threshold effect. TicTacToe is above it for a small child, but below it for an adult. Object permanence is hard for a baby. Self-awareness may not be present in all animals, etc. Important: we cross the threshold upon building our heuristic. That’s when boredom can strike, when we believe – rightly or wrongly – that we have a solid solution.

There are four big clusters of this complexity: Our physical selves, whereupon we see that the entire visceral loop is in fact a game in its own right Some term this category “sport” but the rise of esports belies that I personal think that the def of sport is about spectation and cultural practices Our mental serves, particularly other people Actual mathematical complexity, as exhibited by things ranging from go to the weather Lower end ones are often termed puzzles Higher end ones actually fall into mathematical complexity categories ranging from PSPACE-Complete up through various NP Hard ranges In common lingo, we often call just this “game” Fake math complexity aka chance

Layering a learning loop on something, aka "playing," turns a ludic system into a game. Sometimes this process is applied to systems that do not meet the complexity needed. Like say airline points. This is what we now term the field of gamification. Or pointsification.

If they are a machine, they need to be able to move in response to stimuli. You can think of “how much they can move” as “play” in the mechanical sense, like there’s a lot of play in the joint.

Alternatively, the ludic system might not have a lot of play to it in itself, but it might simply be mediating interaction with others, in which case you have to include the “play” that the interactor has. This is why multiplayer is always richer in possibility space than a fixed authored experience which has highly constrained (likely simple state machine) outcomes.

In the case of social problems, we’re dealing with emotional feedback on kinda hindbrain levels. Mentoring, pride, kvelling, naches, hierarchy, othering, identity These are still, in the end, about a mathematical graph, a social graph if you like In the case of math problems, there’s classes of math problems that are known to have the right complexity level Either way, building blocks that form an entire learning loop (aka atom) in their own right are termed a “mechanic.”

See "Social Mechanics for Social Games" presentation.

See "Games Are Math" presentation.

It’s important to realize that just as a ludic system doesn’t much resemble what it is displayed as, a mechanic or game pattern may appear in very different forms and even have multiple names. For example, party building in WoW and worker placement in a Euro game are both actually getting at the same NP Hard problem of set cover. And so is Sudoku.

I gravitated towards a system I could use on a napkin, because I usually want to do quick sketches for mental clarity or analysis.

Island Life had a pretty subtle difference from Farmville, which was adding the ability to sculpt your island by buying land.

Here you can see what parts of Frontierville were probably hard to balance -- right where it gets tangled--, and which were probably easier to cut if desired.

It was easy to see how the social games market evolved over time...

This was the graph for Metaplace, where we were trying to solve the issues of an incentive system for three groups: content creators, players, and socializers.

Doing this sort of analysis enables asking simple questions that can help a game quite a lot.

I recently worked with Buzztime to revise their "Countdown" trivia game. This is what the loop looked like for it. There was no way to learn to get better at the game -- it didn't loop.

We added a few orthogonal ways to play in order to encourage broader audience, a learning loop, and it paid off in terms of metrics.

It's funny how the graph isn't actually very different at all between Atari Tennis and Virtua Tennis.

In this view, metaphor is a severable part of the design. This doesn't mean it isn't incredibly important -- it is, and there are entire disciplines for each of these, and they tie back to game grammar in a deep way, especially since the interpretability of these tools makes them powerful playful spaces themselves.

Some argue that games are in fact all state machines and of finite space. This is probably true but only in the extreme sense: the state space of Minecraft is quite large, and most of it is totally undiscovered. Same goes for any virtual community. The MODE (as mode, median, and mean) is low choice architecture games, and that is because it leverages the power of feedback. Feedback and metaphor are commercially defensible, and mechanics are not.

This is about the author’s intent, from a crafts perspective…

For more on this see the presentation "Playing with Game."

I have a personal bias towards the upper right quadrant; the others are done well by other media, but the other media don’t do that quadrant well. So I see it as a sweet spot. Mention mythmaking versus authored story, plot BUT – the narrative quadrant, the upper left, is also rich with interpretability when done right, good authored stories And we shouldn’t denigrate acculturation (the role of most all comedies, fables, children’s books) or training (vital for survival). There is art to making them.

In the end this feeds back into our mental model of that black box. We finally made it around one revolution of the atom. ONE BUTTON PRESS did all of that. Now so will moving the knight to capture the pawn. Or kicking the ball towards the goal. Or deciding whether to stand pat or raise. Or adding a worker, or setting up a build chain, or deciding to attack or defend.

Practically every box in the map of game I am going to show you has an entire discipline behind it. It’s not surprising that we need UX designers and storytellers and mathematicians and community experts and so on. So: for every feature, every button, every interface, every atom, you can run this loop. You can ask questions. You can diagram the system and see whether it holds up. In this worldview, it’s very easy to sanity check concepts for coherence. Does the objective in the metaphor match the objective in the math? It’s also easy to tune. Does each system meet the complexity threshold? Is the feedback present, and multimedia? Do you provide social constructs? I’ve pitched THAT as a talk at GDC, we’ll see if it gets through. You may have noticed that not only are games made out of games, but that this entire thing is presenting a system comfortably above the complexity threshold. That’s why designing games is such a great game to play. My favorite, in fact. So I hope this little strategy guide helps you out!