[In this reprinted #altdevblogaday-opinion piece, Vigil Games' senior designer Mike Birkhead offers several invaluable tips for crafting your game's boss encounters.]
There is no truer test of the combat designer than a boss, because they, more than any other cast member, thread those tightest of needles: challenge vs frustration. Your task is difficult, and, unfortunately, this topic of boss design expands far beyond one simple article.
I wish I had a simple solution for you. Every boss is different, though, which means there can be no magic formula. No, one thing, that good bosses have, and bad bosses lack.
What I have for you is Old Knowledge. A checklist, of sorts, that I like to run through in my head whenever I look at a boss. Like all great design tools I am borrowing it from greater men: Ollie Johnston and Frank Thomas.
They, and other animators at Disney, gave us the "12 Principles of Animation." And of those 12, which are all great to know, we designers are concerned with five: Anticipation, Staging, Timing, Exaggeration, and Appeal.
For animators, the principle of anticipation is the idea that you must prepare for any action the character is going to take. If someone is going to swing their sword, then you should have him tense his muscles in a preparatory motion.
We are not animators, though, and for designers this is known as making sure the attacks have "good reads" – meaning, the player can read, or understand, what the boss is about to do, so that he can react in the appropriate manner.
If you ever heard a designer say, "That attack has clean reads", and in your head you were wondering what the hell he was talking about, well now you know. You want clean — meaning easily indentifiable, the opposite of muddy — reads on the attacks, but what's more important (and most people tend to forget) is that it is not enough to have clean reads. You must make sure that your reads do not overlap.
Let's say I was designing a giant crab boss, and I want to introduce two attacks to the boss.
Attack 1: claw smash
Anim: raises right claw up above head and clicks claw (80 frames)
Sound: clicking claw sound
Anim: smashes claw down on the ground and creates a shockwave that expands in 360 degrees at a rate of 0.5 Hero Units a second for a total of 2 seconds.
Attack 2: claw swipes
Anim: brings right claw back (60 frames)
Sound: chittering sound
Anim: strikes out with the right, and then left claws in quick succession. The right claw covers a 95 degree arc on the eastern side, while the left claw covers a 95 degree arc on the western side of the crab
What's wrong with this set up? A couple things, actually. If you look at these attacks separately, then they appear to have clean reads: big motions that set up what's about to happen, and we even include unique sounds.
When you look at them together, however, you see that we have problems. Raising the right claw up and bringing the right claw back, while certainly different, aren't really different enough.
Remember in another article we talked about the Foveal vs the Periphery with regards to vision. The foveal (center) of your vision is great at picking out details, but terrible at sensing motion, while the periphery is the exact opposite.
In the case of these two attacks, your eye will sense the motion (the right claw is moving), but without shifting your full focus (foci) to the claw, you won't be able to tell the difference. What's more, the sounds of clicking claws and chattering, while cool, really do sound fairly similar.
All of this combines to muddy the reads between these two attacks. That isn't to say you can't set out with this intention in mind. Sometimes you may WANT to muddy the reads, but — very serious but — one does not go breaking rules, unless you understand the rules in the first place.
The principle of Staging can be explained succinctly: show what matters, and only what matters. When designing a boss you want to avoid extraneous details. This is a strong concept with regards to ANY cast member, but for bosses, especially; when you add puzzle mechanics and interactions — as you certainly will — you don't want to muddy up the presentation.
This principle refers not just to the boss you are designing, though, but also to the arena it takes place in. If you want someone to interact with something during a boss fight, but only at certain moments, then it needs to be either clearly usable, or very clearly unusable; similarly, if the boss has a weak point, then it should be displayed in an obvious way.
It had to be made obviously unusable
The first time we took the Scylla Boss (from Ghost of Sparta) to a play test, we didn't have the wooden railing around the crane wheel (it gets destroyed when she goes prone for the first time, you will notice).
The crane wheel served no purpose, until Scylla went prone, but you could still interact with it. Well, what do you think happened? People would do nothing BUT interact with it! Over and over and over again. Haha.
We even have special attacks Scylla would play to smash you in the face if you tried to use the wheel (that didn't work either). People didn't care. If someone can interact with something, they will. We forgot staging.
A large boss needs to move like a large boss, which means he needs weight. For the Disney animators, Timing referred to the number of drawings (frames) they would use for a given action. The more frames, the slower and heavier the action would feel.
Here's the funny thing about timing, though. You would think, because it is logical, that the longer your anticipation, the easier it is to avoid or parry attacks. Longer is better, right? Well, yes and no. Your ability to react effectively doesn't really follow a linear progression.
Longer IS better, but only to a point, and that point, from my experience, is around 2 seconds (with the sweet spot being between 1 second and 1.5 seconds). In my personal experience, anything beyond 2 seconds I need to start counting. I no longer can react based upon the feel, and that's usually always a bad thing.
So you need to have good timing on the attacks and other animations, obviously, but there is more to timing. Remember we are designers, so this isn't just about animation. Timing, here, refers not just to the timing on your attacks, but also to timing on your holds.
A "hold" is the short vulnerability window that you provide on a boss, so that the player can safely get their licks in — you know what I'm talking about. It's the belly up, "have your way with me" moment.
Anything less than one full combo is not enough — NOT ENOUGH — and remember to take into account the time it will take for your player to traverse over to the boss, too. You don't want to constantly keep giving the player a halfie.
However much exaggeration you think is too much, it is not enough for a boss. Bosses are not about subtlety — not at all. They are about larger than life shit happening to larger than life you. Here's the great thing about exaggeration: people don't actually like dying on your boss. They just want to feel like they are getting their ass kicked. Really.
Go play a God of War boss. You can pretty much just pick ANY of them, as they all nail this, but this time instead of watching what is happening, just watch your health bar. You'll be shocked, I think. Kratos gets smashed, flattened, thrown, burned, and electrocuted about a billion different ways, but that little bar won't go very far, and every Vulnerability moment rewards tons of health, just in case.
If someone has to constantly look up at the health bar to see how things are going, then you're pretty much doing it wrong. Remember, your job is not to prove how awesome your boss is, but to prove to the player how awesome they are, and you get there with lots of exaggeration.
Last but not least we have appeal. This is a tough one, but what was true for Disney back then is true for bosses you design today:
Appeal in a cartoon character corresponds to what would be called charisma in an actor. A character who is appealing is not necessarily sympathetic — villains or monsters can also be appealing — the important thing is that the viewer feels the character is real and interesting.
If I don't want to look at it, then I'm not going to enjoy killing it. Simple as that.
I always find Devil May Cry's bosses appealing
An aside to this: this principle of Appeal is important to keep in mind when designing weak points for your bosses, too, for if the weak point on a boss is a bunch of spikes, then I'm pretty sure most people will avoid it, as spikes are the very definition of threatening.
Devil May Cry has always been very good about designing weak points. The frog boss from Devil May Cry 4 has a very specific weak point — his tongue — and he only sticks it out when he goes into a well timed Vulnerability Window.
Designing a boss is a topic that spans far beyond the scope of this article, and as I have said there is no hard rule that you require.
What I have provided you with are some simple tools that I like to keep in my back of my head for when I look at bosses. It helps you to build the right questions to ask. Are they building anticipation? Are the important parts being staged well? Are the timing windows generous (or not generous) enough? Is there enough exaggeration? Do I want to kill this boss?
It's not like answering yes to all those questions is going to prove the boss is fun to kill, far from it my friend, but they will illuminate your path; and in a world as murky as boss design, you need all the light you can get.
[This piece was reprinted from #AltDevBlogADay, a shared blog initiative started by @mike_acton devoted to giving game developers of all disciplines a place to motivate each other to write regularly about their personal game development passions.]