Brain Dump: The rise of context sensitive actions (Patreon)

I’ve been thinking about something recently. Thoughts have been swirling around my head. And maybe they’ll develop into a full GMTK episode - but right now I just want to get my thoughts out there so you can think about them yourself, and contribute your own ideas.

So, I’ve been playing a random hodgepodge of games recently. Really new stuff like Assassin’s Creed Odyssey, Hitman 2, and Red Dead Redemption 2. And comparatively older stuff, like Sly Cooper and Castlevania: Circle of the Moon.

And something feels fundamentally different about these two collections of games. Like something shifted in game design between these two epochs. And I think it might have something to do with context sensitivity.

Okay, so back in the olden days, games would often have an all-purpose action button that would do different things depending on where your character is standing. It’s most famous in Zelda: Ocarina of Time, I guess. While the big green button is almost always used to swing your sword, the big blue button can make Link roll, pick up stones, talk to people, jump, and so on. All depending on what Link is doing or looking at, at that moment.

And this is, for sure, a smart idea. While old roguelikes on PC would have a zillion different keyboard commands to do a zillion different actions, Zelda can ram all of those actions into a single button and just have its use change to fit the circumstance. The only concession is that the button needs to be displayed on screen at all time, to let you know how it will be used.

But what seems to have happened, is that in a lot of big, ambitious, open world games there’s no longer a single context sensitive action button. No - almost every button can be context sensitive.

Nowadays, games will completely change their control scheme at a moment’s notice. Buttons work different depending on whether you’re on foot or in a car. Near enemies or near NPCs. Walking or running.

So in Shadow of Mordor, the B button can be a stun in combat, a climb down during parkour, or a “drain” when grabbing an enemy. In Red Dead Redemption II, the left trigger aims your gun, but also lets you talk to people. In Assassin’s Creed, the buttons do different things depending on how fast you’re moving.

And you can see why these games do it. They’re complicated. They span multiple genres inside a single game. They want the characters to be able to do loads of different things. But there simply aren’t enough buttons on a modern controller, so the devs just constantly change what the buttons do.

But I think the consequences of this can be quite negative. 

Like how these games are constantly telling you what the buttons do. And this means the UI becomes a mess of button prompts. Some are better than others. I feel like the Hitman UI, where the buttons hover directly over objects, is pretty easy to comprehend. But the Red Dead system, where the bottom right corner is an arbitrary list of new button configurations, is a nightmare mess.

And it’s confusing! You can’t just remember that X is jump - you have to be constantly redefining what X is going to do. And then there are annoying cases of conflict where you want to pet your horse but you accidentally punch it. Or you need to wiggle your character into just the right spot so that they pick up some ammo, and don’t get into a car instead.

But those are just the obvious drawbacks. It goes deeper than that, I think.

So going back and playing these older games reminded me that games used to be about mastering a consistent move set across a variety of challenges. Can you use Mario’s jump, or Doom’s guns, or Dante’s attacks to get through these mad obstacle courses filled with enemies?

These more modern games don’t really have that, and can end up feeling like a bizarre jumble of disconnected montages and simple mini-games. Shovel up pig poo for a bit. Look after your horse for a bit. Rob people on a train for a bit. Put a wheel on a wagon for a bit.

And in a game with dedicated actions, those actions can be deep and versatile and expressive. Think of Mario’s jump - you can modify the height and distance, move in mid-air, chain in and out of different moves, and adapt it with power-ups.

But if jump is just a contextual button press, it simply becomes “Press X to jump”, and that’s the only jump you’ll do.

This reminds me of games like Life is Strange, where Max can water a flower, play a guitar, look at a picture, take a photo, and so on. It’s just all simple button presses. And that’s fine - the game isn’t trying to give you 1:1 control over Max. But games like Red Dead and Assassin’s Creed certainly are trying to do that with their characters. 

There’s also the weirdo body horror of trying to manipulate a character whose actions are unpredictable. 

Video game characters should feel like an extension of our own body, or like we’re taking control of them for a while. And in real life, our actions are consistent. But controlling Arthur in Red Dead is like going to shake someone’s hand and accidentally kicking them in the shin. 

Someone on Twitter said that “you control Mario, but you direct Arthur”. I like this.

Also, there’s the troubling idea that the game constantly tells you what you can do, instead of the player deciding. If your attack button is currently the push ladder button, then I guess you can’t attack right now. 

The other version of this, I guess, is that the game interprets what it thinks you want to do. Like in Assassin’s Creed, where you hold the parkour button and run towards a building. The game decides for you how Ezio should move - which is either patronising, or out right wrong. No, I don’t want to climb up that drain pipe, I wanted to run up that wall!

Now you might think that the root cause of all this is games trying to put too much stuff in. And that’s probably true. But I think a bigger problem is that games are trying to be extremely realistic.

Because the solution for this problem has already been found. It’s to build the world around the player’s mechanics. 

Examples being how Luigi can manipulate the world with his vacuum cleaner in Luigi’s Mansion. How Kratos can smash pots, hit switches, and stop machinery by throwing his axe in God of War. How Sly Cooper makes wheels spin by running on them like a treadmill. How Samus opens doors by blasting at them, in Metroid. 

This is elegant. This is well considered. This even lets basic actions in the game have depth, because the main mechanics have depth. So a simple puzzle in God of War has thought and challenge behind it, because throwing the axe requires aim, and you have to account for the travel time of the axe, and so on. And then hitting a switch becomes training for throwing your axe at enemies. I love this stuff.

But it’s not realistic. God of War 2018 actually feels super odd because it’s all hyper realistic and everything - and yet it’s filled with goofy puzzles and all the switches are conveniently designed to work with Kratos’s axe and Atreus’s bow. Did anyone else feel this uncomfortable disconnect? It’s like a game that’s stepping between two eras of game design. 

So you could make a cowboy game that’s based around a few core verbs. Like maybe you’d have a brush as a weapon, and you could “attack” your horse to brush it. But then you’d be able to brush other characters and enemies and it quickly becomes a silly gamey game, and not a super serious HBO drama about the old west and I have a PLAN ARTHUR GOD DA-

So I think perhaps we’re slipping into a new era of games with worlds that aren’t built to accommodate characters, but characters that are built to accommodate worlds. And I’m not sure how I feel about that.

Would love to hear your thoughts!

New episodes of GMTK and Designing for Disability are both in the works and will be out soon. Thanks!