The Fractal Nature of Making Games

Been reading and listening to discussions about crunch. There are lots of opinions on it and why it exists. Do I think that it has to exist? Maybe? Do I think it’s avoidable? Up to a point. Let me explain why without referencing mismanagement or passion tax.

Making games is hard. I really need to drive that point home. Like, really hard. Is it harder than other aspects of software dev? I don’t know, I don’t work in other types of software. What I do know is that the thing with games is that you’re not just “making a game”, you’re creating a world. An entire world. Let that sink in.

When you enter into a game, you are entering a completely new existence. It’s not even just games. It’s when you engage with any sort of content. It’s just that non-fiction content doesn’t require any willing suspension of disbelief and doesn’t require any sort of perspective change.

But in any sort of generated content, you enter into the world and become a participant in it. In a story, it’s not necessary to define every last aspect of the world because the reader brings along with them some concepts and beliefs and imposes them on the world. And for the most part, unless it’s relevant, it can remain in the background.

Visual media doesn’t really work that way, unless you’re going for a minimalist approach. Basically, if you don’t create it, it doesn’t exist. And that’s the crux of the problem.

You need to create everything. Like Carl Sagan once wrote, “If you wish to make an apple pie from scratch, you must first invent the universe”. And in our case, if you want to make a game, you must first invent everything else around it. Sure, we have engines now that do a bit of the basics, but think about everything that goes into a game. Can the main character walk? How fast? Do they collide with things? Can they jump? Can they jump over other things? Can they duck? Can they shoot? What do they shoot? What do they shoot at? Literally everything needs to be decided and crafted. EVERYTHING.

And all those things, they’re all fractal, able to be subdivided and analyzed and further defined. Take the walking character, how many feet are they walking on? Do they make a sound when they walk? What are you assuming when you do this exercise? Are those assumptions accurate? Is this character even walking on Earth? What’s the gravity? Is there gravity? You see how this rabbit hole of one thing just goes on and on.

A good game models the world that the player exists in, simplifying the elements that need to be simplified while modeling out the parts that need to be fleshed out. It’s a difficult task to balance at times, and it’s really where people start to get into discussions about simulations vs games.

I think a good designer is able to take a concept, distill it down to its key elements, isolating what makes a game fun and being able to capture the essence of the experience. On top of that, they need to have an understanding of how much something needs to be modeled and simulated so that it exists in the world in a way that the character can interact with it.

So the designer is tasked with all the nouns in the world. Every book, chair, table, car, door. And on top of that, they’re then tasked with every verb in the world. Every push, every step, every shot. (Cue every game designer that has used the door problem)

Even just taking a simple thought exercise reveals how much work there is. When you wake up in the morning, think about the steps you take from waking up to getting dressed.

Open your eyes.

Turn off your alarm.

Get up.

Walk to the bathroom.

Etc, etc, etc.

Of just those things listed above, I have the following obvious nouns:

Character

Alarm

Bathroom

But what about the non-obvious nouns:

Eyes

Floor

Bed

Nightstand

Does the player interact with them? In what way? Do they need to have interaction established? Are they animations (which then need a whole separate level of breakdown)?

Are the verbs that the character uses unique to certain nouns? Or do they need to be ubiquitous? Is there an animation necessary with the verb? Does there need to be?

Verbs from the above steps:

Open

Switch off

Get up

Walk

Every verb leads to another verb and another question. If I can open, can I close? If I switch something off, can I switch it back on? If I get up, can I go back to sleep? If I walk, how do I do so?

And this is just part of one sequence of a thought experiment. Modern games are made up of hundreds of sequences, interactions, and systems.

So then when it comes down to figuring out how long it’ll take to make a game, what is your best estimate for creating a world? Even if you’ve accounted for every noun and every verb, there’s still found work, unexpected issues that arrive, tech limitations, content pipelines, just an entire smattering of dependencies.

And then you set a release date because you want to ship at some point, but that means that the clock is ticking now. You’re finding bugs because you see the verbs behaving with the nouns in the way that you want, but there’s always edge cases, and with every verb times every noun, suddenly you have a giant matrix of expected and unexpected behaviors. Do you have time to fix them all? Of course not. Every game ships with bugs, it’s just whether or not the bug is bad enough to affect the player experience.?

But there’s a few really bad ones, and with the deadline in the sand, you start to work later and later in an attempt to shore up some of the worst offenders.

Suddenly, you’re crunching.

Okay, of course this is reductive. You can avoid this by changing scope, or by pushing the deadline back, or cutting corners on implementation. Of the three, changing scope is probably the most preferable, but ultimately, you have to make a decision somewhere and it’s going to piss someone off.

So yeah, making games is hard.