What Makes a Good Game?
Key Takeaways from Will Wright's 'Lessons from Game Design' Lecture (2003)
Will Wright is an American game designer, probably best known for his work on the popular Sims franchise. That being said, I had no clue who Will was until I was pointed in the direction of his 2003 lecture on game design. By the time I was finished watching the nearly 2 hour talk, I had pages full of notes whose scope spanned far beyond what the title might suggest.
Here are some of the most broadly applicable concepts I walked away from the lecture with.
The Purpose of a Model
Before I highlight my 2 favorite models from Will’s talk, a quick note about models in general:
Models are valuable because of their ability to remove non-essential information and yet still convey the foundation of an important idea. They are valuable in abstraction, but only abstract insofar as they still have a purpose/objective for the person they are being offered up to.
When I recap below the idea of The Fitness Landscape and Nested Feedback Loops, think about how each is applicable to your life, and take the opportunity to layer additional context onto the basic model as you see fit.
The Fitness Landscape
Will starts off his lecture by talking about how games explore a ‘possibility space’. For a board game like monopoly, the spaces on the board all make up the so called possibility space. Within the possibility space of a game, there resides a multidimensional graph called a fitness landscape.
Picture a graph with 3 axis. No wait, I’ll do it for you:
On a fitness landscape in the case of The Sims, on the X-axis, you have something like social skills, and on the Y-axis, you have something along the lines of material or hard skills. Notice how in the above example ( and most examples to Will’s point ), the overall fitness of the player character is not maximized by optimizing for one skill or the other.
Instead, the player’s measure of fitness, or Z-axis, has greater potential to be maximized by pursuing a path of balance between the 2.
You can apply the concept to any number of dimensions. In addition to the idea not to lean too heavy into one or a few of a wider array of traits, Wright also highlights what he refers to as a Local Maxima. A local maxima on a fitness landscape, or specifically the example above, is any of the peaks that isn’t the highest of the landscape. The important thing to takeaway from the idea of a local maxima, is that before you can reach a new peak, you must first journey down from one before beginning the next hike up.
Nested Feedback Loops
Imagine a system where there is an input point, that takes something like your effort over time with a target outcome. If you succeed, you move to the next loop, and if you fail, you restart the stage of loop that you are on.
Wright immediately draws the analogy to game design, where let’s say a player starts of with the goal to understand basic controls. This may take them 10-20 seconds of button mashing to figure out, but once they’ve got it, the game will prompt them to progress to the next task, which may be something like navigating from point A to point B. If you fail by falling into a chasm, stepping on a trap, etcetera, you restart the loop. When you make it to point B, you may then be prompted to achieve something more complex like building an object or defeating an enemy, and so on.
Lessons from Games —
Players should experience increasingly difficult and far off goals.
Failures should become increasingly interesting and varied.
Player should always have a path to understand why they failed.
A final noteworthy takeaway is that players get bored and stop playing when the game runs out of layers. Picture a game like:
Learn Controls → Meet Needs → Acquire Skills → Build Relationships → ?
What level are you trying to beat?
Remembering what I said about models, layer context where you please. It might just help you play a more fun game. Fail in interesting and varied ways, understand why, and set increasingly difficult and far off goals. I watched this lecture because I heard @VanceCrowe and @Plantimals talk about it in this clip from their recent podcast together.
Next week, I’m going to talk about the thing that will (maybe) eventually kill us all.
Play on.
-Benjamin Anderson