Chasing a nostalgia high, I recently installed Grand Theft Auto: Vice City (GTA VC or VC) on my computer. It’s a game set in 80s Miami, heavily influenced by movies such as Scarface. You could say it’s “Scarface: The Game” and it would be hard to disagree with you. GTA:VC was made in the early 2000s when compute power wasn’t anywhere near what it is right now. The PC requirements start at 128MiB RAM, which in this day and age is unthinkable. But back then it was used to create a rich game world that feels more immersive than some more modern games. The weather system is a gradient, it gets cloudy before it rains and it’ll go back to cloudy weather before the sun breaks through. Time works the same. Sped up for gameplay’s sake, the sun dawns properly in the morning and sets as it would in real life at night. The streets are busy with people who may interact with one another. Newspapers lamenting Elvis could be an alien are blown around. Palm leaves from the trees that line the streets are strewn all across the map dynamically as well.
Video games are, in essence, an abstraction of reality. As more and more compute becomes available, we are able to simulate more and more complex environments. At one point we have so much compute available, the abstractions can disappear completely. You can look at Microsoft’s latest iteration of Flight Simulator to see what that looks like. The game makes you turn on an airplane from a completely powered down state before you can fly. Compare this to the latest version of Grand Theft Auto, GTA 5, where a single keypress grants you access to an airplane. Pressing the arrow-up key makes your airplane go forward, arrow-down and your airplane goes backward. These abstractions serve to make the game more accessible and as a result more fun.
We can also compare these differences within the series of GTA itself. VC (and games before it) has the concept of “stores” to achieve certain actions. There are gun stores where you can purchase weapons. Clothing stores where you can purchase clothes. But there are big differences in how the interaction works. For example, in GTA 5, if you want to change your outfit, you have to go through the following:
- Walk into the store
- Go to the right rack for the item of clothing you want to change
- Press a key to open the menu
- Cycle through the available options
- Press a key to purchase the item
You’ll now wear this item, and the item will be added to your inventory. To change into a purchased item you:
- Go to a location that has a wardrobe
- Open your wardrobe
- Select the item you wish to wear
In GTA VC there’s also a clothing system. However its system is primitive.
- Go to a location that has a costume power pill
- Walk over the power pill
- You are now wearing that outfit
There’s no preview. The only hint of what the outfit will be can be deduced by reading a title that hovers over the power pill. “Bankjob”, “Suit”, “Beach”.
These abstractions can be found everywhere. If you wish to heal yourself you have to find a medicinal power pill in the city. They always show up at the same locations, so after playing for a while you’ll be able to know where you can recharge your health. Losing the cops can be done by running over a wanted-level power pill. In GTA 5 to recharge your health you would have to find a convenience store to buy items that will be added to your inventory, to be consumed then or at a later time. Losing the cops? You’d have to open your phone, call your hacker friend who would then hack the system and make your wanted level lower in exchange for a fee.
Vice City, similar to GTA 5, also has photorealistic graphics. By photorealistic I mean that they took photos of buildings, slapped them on cuboids and called it a building. However because of all the auxiliary items and events such as pedestrians, other cars, the garbage blowing around, street lights, et cetera, it all works really well. Unless you pay attention to it, you’ll hardly notice the streets you drive around on consist of buildings that are essentially cuboids. Whatever round geometry is in the game is reserved for roads and auxiliary objects. With buildings, almost nothing sticks out or in.
This all got me thinking about how as compute power increases, abstractions in video games dissolve but also affect gameplay. And not always positively. At one point, we will be playing nothing but simulators. Are we playing a simulator right now? Second Life but then for real? Games can become more and more complex until they mimic real life. Where will game developers draw the line? At what point does realism go from “fun” to “tedious”?
What makes the Grand Theft Auto series fun is I can drive over people. Something I can do in real life too, but won’t because it’s not right to drive over (most) people. It’s fun because I can crash through the gates of an airport, get into an airplane or helicopter, and fly away. Also possible in real life, but the legal consequences are insurmountable. On top of that, it’s also not right to do. Will GTA still be fun if my character gets taken into central booking, and I have to choose a public defender? Will it still be fun if I get away with running over people, but I have a mini-game where I powerwash blood from my car?
More diluting of abstractions is happening within a different video game called Call of Duty, a first person shooting game. The game initially launched in the early 2000s, with a very simple class system. When playing multiplayer there were 5 classes to choose from. Each class had its own weapon, so essentially you had 5 weapons. Now in Call of Duty, the system is much more complex. Instead of choosing classes you can choose your weapon directly. Each weapon can be upgraded, which turns into a careful balancing act. You can add a scope to your weapon, but this will decrease your close-range accuracy. You can add a foregrip, but this reduces the time it takes for your weapon to be ready to fire. Your weapon used to be static. The intricacies of handling were abstracted away into maybe 3 or 4 attributes that differ per weapon. Now if you want to play the game and be good at it, you need to leave the game, arguably breaking immersion, and read up on what the most balanced weapon is on the internet. Assume for a moment you have 10 weapons, with 5 different attachment points, with 10 different options. This gives you over a million combinations to try. Even eliminating the obviously bad ones, how can a casual gamer ever find the time to come up with the best weapon combination?
With limitless compute, games and the abstractions we create are a product of time spent. And although easy to measure, how “long” something takes is much more subjective. I do not have the time to check a million different weapon combinations. A teenager on Adderall will test it all over the weekend and have his report on YouTube the day after.
Is this the point where I simply say “I don’t have time to play games” like generations before me? Or is this an actual phenomenon in gaming that eventually will have to be addressed as even the most Adderall-addled teenager fails to extract enjoyment from a game where creating the best weapon requires a PhD in statistics.