Reposted from www.throwthelookingglass.com
In a Mechanical Breakdown, I look at a particular mechanic utilised in games and examine when it works and when it doesn't, from both the gameplay and the narrative perspective*. The goal is to use existing opinions from around the internet and my own experience to find new ways of looking at existing mechanics.
This time, the boss fight.
This article is going to go a little differently to other Mechanical Breakdowns, as there's a single principle which I found to underline the subject area.
The key to effective climactic events is paying off what you've set up.
That's it. Why? Find out after you click "Read More" 3 times while dodging all other links with the special Mouse of Clicking...
Set-up and pay-off is a staple in all storytelling media, and the idea that your story's climax should pay off you've set up is no big surprise. The challenge in games is that the gameplay must also do this for an ending to be effective. This makes the climax an interesting microcosm of a challenge that developers constantly face: that of satisfying story and gameplay equally.
What do I mean by paying off gameplay? Essentially, it's that the boss battle should be logical extension and expansion of what has come before, and that if you set up a gameplay element as being important in your game, you should pay it off by using it in a new way in the climax. These set-ups and pay-offs can cross between gameplay and narrative (so if an object is set up in narrative, it can be paid off by allowing you to use and experience it).
What, you want some examples? It's your lucky day: I've got some examples for you.
I'm going to start this off by talking about one of the games I talk about a lot: Half-Life 2. Well actually, I'm going to let Yahtzee do the talking for me.
The game ending comes (crucially) after all your weapons have been taken away and you're left to resort to the gravity gun that the rest of the game has been patiently teaching you how to use to best effect... It brings together elements the entire game has been building up to, rather than teleporting you to a different galaxy altogether to fight a couple of the big things that live there for some slightly arbitrary reason.
Half-Life 2's final chapter is, in a word, stunning. The genius of it was in unexpected wish-fulfillment: as soon as we grab hold of the gravity gun, we think "wouldn't it be awesome if I could throw people and these bigger objects around with this", followed quickly by "they'll never do that - it'd unbalance the game". That's the set up. The pay off is doing it anyway.
That, and making it fun.
In gameplay terms, there's nothing more fulfilling than using the gravity gun to throw combine soldiers at each other, and using it to effortlessly kill hordes of the tough bastards. In narrative terms, the final chapter brings you to the citadel which has been (literally) looming over you for the entire game as the source of the Combine's power, and allows you to confront Breen, the traitor whose condescending mockery of loyalty has haunted you throughout the game.
Yet it's not a perfect ending. The very end of the game, when you're on the roof of the Citadel, stopping Breen from escaping, is rather uninspiring. You're using the gravity gun to take out a grand total of two gunships in a way that's less fun than before (a.k.a the guided missile) and you're then using the same mechanic to shoot at a rotating shield thing that protects Breen. While this is preferable to Breen donning a power suit and coming at you Iron Man style (that would have been horrible, as it would have broken the climax from anarrative perspective) and it pays everything off in a narrative sense, the gameplay didn't quite feel bigger than what had come before. As a completely untested alternative: what if you'd had to gravity-gun the thing that creates the plasma orbs and use it as a projectile weapon against more enemies, thus paying off all the gunfire-style gameplay elements as well as the gravity gun and the overwhelmingly massive combine army.
Half-Life 2 teaches us a really important thing for the boss fight: you don't need a big bad for a climax: hordes of bad guys will do just as well. Or, in the words of Yahtzee:
concluding a game experience with a big fight when a game might have had you doing all sorts of thing besides fighting up to then, may be missing the point
This, I feel, is where an RPG can face difficulties. Often, the gameplay of an RPG is twofold: fighting, where you have to outwit, outsmart, outshoot and out-eat-health-potion the bad guys; and conversation, where your choices of how to interact with others in non-combat situations can drastically change the outcome.
The first is (relatively) easy*. Bosses (or hordes of bad guys, or both) work pretty damned well. But how do you pay off the gameplay of making choices?
*I'll get to the 'relative' part of that in a minute, as RPG's bring their own issues to this.
The RPG can take two approaches to this: they can ask you to make another choice which is more important than the rest; or they can hand you the consequences of earlier choices on a silver/blood-stained plate.
While both of these can work, I prefer the latter approach, and I'll use my own experience from Fable and another's from Fable 2 to explain why.
I do believe the "boss" for Fable 2 is a noted example of suckatude. You walk up to a guy with a mcguffin and either blast him in the chest with a pistol or wait for Stephen Fry to do it.
So what's wrong with this boss? In order to make the choice-making an integral part of the boss fight, the game ignores the fight: that other integral part of the game. While I'd argue that the choice given after the death of Lucien paid off the choice-making quite splendidly, the game robs itself of paying off the build-up to a confrontation with Lucien.
Meanwhile, the original Fable contained a (not final) boss who wanted to kill the character's sister. In order to make that choice important (and therefore 'climactic'), it had a large effect on your character's good/evil scale, to the point where my evil characters (horns and all) suddenly became good (halo and all), just because he didn't want his sister to die. By making the decision 'bigger', it had the ability to undo previous decisions without any real character basis for doing so.
On the other hand, making the player eat earlier choices can also be fraught with difficulty. The final mission in Mass Effect 2 is climactic in that it draws upon the loyalty and ship upgrades you may or may not have obtained to decide who lived or died, and yet the whole exercise felt rather false, since the choices themselves were false. There was obviously one right way to finish the game, and the game pretty much tells you what it is. The only reasons to make the wrong choice were impatience, misunderstanding or wanting to see how it affected the game. Meanwhile, meaningful choices whipped past at a rate of knots, with no repercussions in sight.
While I'm bagging on Mass Effect 2's boss (which I found to be a far better, though sadly far less ambitious, game than its predecessor), the final boss is an example of lazy writing in a boss. As soon as you see the human reaper you know you're probably gonna have to fight it, but by the time you actually do, you've been told that it's dead and not coming back. It's resurrection is then ignored: the game brings it back only because it needed it as a boss, and decides that that's a good enough explanation for the rest of us.
Regarding the fighting boss in an RPG, Yahtzee says it best in his summation of Alpha Protocol's boss:
if you haven't been cultivating the specific combat abilities necessary to counter the boss's various attack phases, then get ready to open wide and chew on the overloaded spoonful of shit that's coming your way
The moral of the story: if you've got a bunch of different skills and stats in an RPG, make sure that your boss can be killed by players who take each available path. Or, you know, have different ways of approaching the boss (or hell, different bosses), and let them choose - that way, it's the player's fault for not recognising their own strengths.
Difficulty is often a problem in bosses: not only in whether they're actually difficult, but what developers do to make them difficult. Take this quote about Metroid Prime's Boost Guardian:
The boss itself is not hard, it's just that you have to fight the boss in an area filled with poisonous air that slowly kills you. It takes to damn long to kill the boss so you can die easily.
In this case, the designer has made the boss harder using two simple factors: a time limit (albeit a creative one), and a super-dose of health. I'd argue that unless time limits or dispatching specifically tough enemies has been a staple of the game so far, that these are a symptom of lazy design trying to make the climax of the game bigger by making it artificially harder, instead of following the gameplay thus far to a logical conclusion.
Which begs us the question: does a boss actually have to be hard?
Galaxy pits the mustachioed hero against a fairly normal Bowser, but the game's gravity bending physics are what set this battle above every other Mario-Bowser face-off -- it really puts this Bowser battle in a league of its own. It's not the most difficult confrontation, but that very fact that also makes it so satisfying.
This implies that no, it doesn't. Galaxy's boss isn't about punishing the player in the hardest thing ever, but about combining the set-up of the gravity-bending game with the set-up of every other Bowser fight ever to create a stupendous, fun (and big) experience. The climax is almost a celebration of the wonders of the game thus far.
Personally, I try and make the boss fight on par or slightly easier than the final level in that chapter. I don't want to defeat the player just as they are about to unlock a new set of levels.
Super Meat Boy's Edmund McMillen puts it stupendously here. By ensuring that the boss is tough, but not so tough that you can't defeat them if you finished previous levels, you can ensure that the boss battle is a great experience.
In practice, the exact difficulty of a boss battle will depend on the mechanics and purpose of the game at hand, but the basic principles of set-up and pay-off remain.
Well that's about it for this installment of Mechanical Breakdown. The moral: if you set up story, pay it off. If you set up gameplay, pay it off. Think of it as Checkov's Gameplay.
What's your favourite boss fight? Is there a case where the rule of set-up and pay-off didn't ring true? I'd love to hear your thoughts.