A version of this article first appeared in the October 2010 article of Game Developer magazine.  It is effectively a ripoff — er, a REVISITING of a previous article I wrote for them here.  Hey, crunch sucks.

I’m a pretty mellow guy, for the most part.  I’m relaxed and easy going when it comes with design. But some things can make me go purple with rage.  One of those things are boss fights apparently designed as afterthoughts by otherwise capable and talented design teams.

I find it inconceivable that truly terrible boss fights still infect our games.  You would think we’d be better at this now.  Our genre is now middle-aged – Pong is 40 years old, for Pete’s sake.  We’ve had decades to hone our skills and practice.  And yet still, I’m playing triple-AAA games with boss fights pulled straight out of amateur hour.

A poorly designed boss can cripple or kill a game.  This is even more true nowadays, where many games are linear – which means the unfortunate player can’t move on without finding some way past your design abomination.  In such a scenario, stumped players have no choice but to reach for the strat guide, dial in the cheat codes — or quit playing altogether.

Disdain for crappy boss-fights is not new.  Some designers nowadays think that the idea of bosses are obsolete, and should ultimately suffer the same fate as the dinosaurs.  I disagree – a well-tuned, well-balanced boss fight can provide an epic capstone to a chapter or game, and can help create an emotional flow through the gamespace that makes the entire experience more compelling.  Great boss fights provide epic, memorable game experiences that will often live with the player longer than grinding through the cannon fodder to get there.

Unfortunately, a lot of boss fights are still a long way from ‘great’ or ‘epic’, and lousy bossfights can kill a game.  Too many design teams are still failing the basics.  And so, written in the glint of my incandescent rage at a game-that-shall-remain-nameless , here are a couple of things for designers to remember as they put together the ultimate showdowns in their own gaming experiences.

Show me my progress

Recently, a frustrating boss encounter in a platformer had me so apoplectic with rage that I’m fairly certain a blood vessel in my brain was about to burst.  My quick jaunt over to GameFaqs didn’t help matters.  Don’t get me wrong – the guide was quite helpful.  It pointed out that the Boss I was fighting was pretty much impervious to all of my attacks, but quite susceptible to super secret special attack C.  You know, the one I almost never use.

Which all would have been fine – hey, I like trying new things – except that the game didn’t tell me that.  I –had- tried using the specified attack and, near as I could tell, it made no difference.  There was no health bar that marched down (or not) based on my attacks.  No bits of armor that fell off.  No significantly different impact animations or reactions.   No red flash, no flytext, no distinctly different cries of pain.  Nothing.  The only way to tell if you’re succeeding is if your enemy falls over dead.  The only way to tell if you’re winning is to find out you’ve won.

This flies in the face of what video games do well.  Video games, as a media, are magical because they are experiential – they are all about learning by doing.  Players identify the gameplay pattern, and continue to optimize it as the game throws new challenges at it.  But when players have no feedback, they can’t learn.  They don’t know if they’re doing something utterly futile, or within a cat’s whisker from utter victory.   They might be doing absolutely the right thing, but get an unlucky streak and die without knowing how close they are.  All of this is a recipe for no fun.

Get Me Back To The Fray

Okay, so your boss is special.  He’s cool, he’s got some killer moves, he’s nearly impossible, you’re proud of him, and you want him to pound the player into pulp as frequently as possible, for as long as it takes for the player to reverse-engineer whatever half-baked puzzle you have in your head.  Fair enough.  But if you do, stop and ask yourself, does failing suck in your encounter?

It’s almost impossible to believe, but decades after the birth of the platformer, we still have games that insist on throwing you to a distant save point if you die.  Or those who don’t allow the player to replenish his health and energy before getting stomped a second time.  Or worse, save points inside of locked doors, where the player who stumbles into the boss fight at half health can no longer get back out to find a med pack and heal up.

And then there’s the absolute worst: the rare gem of a game that gives me a 3 minute load time between fight attempts.

Then let’s talk VO.  No matter how great the narrative or witty the dialogue, not even the sharpest one-liners can survive the repetition of being heard 30 times in a row – especially when compounded with the frustration and humiliation that comes with repeated failure.  If you can’t put the save point after the cinematic, at least be sure it can be skipped.

Teach me how to beat the boss

Players have a tendency to fall into a rut.  It’s all good and well for the designer to want to break the player out of predictable, rote patterns, but utterly changing the game on him will almost always backfire.  This is more common than you’d think: examples I’ve seen include introducing jumping elements in a game that previously had none, changing the timing of the player’s attacks, making him fight on a horse he’s never ridden before, giving him all-new skillsets to learn or even remapping the player’s controls.

One game I played threw an underwater boss at the player before the player had ever done any other swimming.   This left the player with no choice but to learn a difficult and confusing boss creature while simultaneously attempting to master moving and handling an avatar AND a camera in full 3D – a feat that alone can overwhelm many players.

The solution is simple – this game should have thrown a longer swimming level at the player with lots of easily squashable grunts, to allow the player to gain practice and confidence with the change.  This solution works for nearly any of the complex abilities I outlined above.  Want to throw a boss at the player that remaps his controls and makes him whistle Dixie into the microphone while balancing on a WiiFit board?  Fine, but first throw a miniboss with a similar ability, so the player can learn these skills in a relatively comfortable and non-threatening environment.

Make Interesting Fights

Good bossfights make the player feel smart or skilled, if not both.    They make the player feel like they’re applying what they’ve learned.  The puzzle inherent in the bossfight flows well from the natural mechanics of the game.  While players should feel like they can bring their existing skills to the task at hand, they also should feel like a suitable challenge has been placed in front of them.  Boss fights that wander too far from the core game mechanics often feel out of place or clunky – often because all-new mechanics are not as well thought-out, balanced or polished as the core mechanics.

On the other end of the spectrum, the most common way to make a boss is just to throw a gazillion hit points on it and see if the player can outlast it.  MMO zealots call these fights ‘tank and spanks’ – derisively.  To a designer, such simple fights have a use – to help introduce players to unique concepts of boss fights (in the case of MMOs, tanking and healing) – but for the most party, they are generally regarded as utterly lacking in design imagination.

Great boss fights aren’t just stat boosted bags of hit points.  They’re impressive in scale and scope, often with tailored intros, interesting VO, and unique special attacks.  They need to provide an interesting but somewhat transparent puzzle, which might require an investment in AI, in combat mechanics, and even possibly in user interface.  To feel like the emotional high point you want them to be in your game, they need to bring a little pomp and circumstance before offering a worthy challenge to the jaded player – and not just feel like a showy speedbump.

Make Beatable Fights

Is your big bad boss even remotely defeatable by mortal men?  Too many are just way to hard, depend too much on luck or serendipity or depend on the digital agility of a virtuoso.  If a decent FPS player can’t get past your boss, even after reading the Prima strat guide and watching Youtubes of the finer details, what hope does Mr. Casual have?

It takes discipline for designers to remember that they are uniquely skilled with their own game controls and GUIs, and uniquely aware of the depths of the mechanics.  They need to remember that not everyone has been playing the game non-stop for the full length of a 3 year development cycle.  One cannot let their own familiarity with the game skew what their idea of ‘hard’ is.

Even beyond this, the variables even a good player can bring to a match can often create wild variance in boss difficulty.  An encounter that is trivial on a PC might be nearly impossible on a console controller, due to the difference in precision and camera control.  A boss fight with a bigger light show than Dick Clark’s New Years Eve party might bring a laptop to a crawl, and kill the player with bad performance. And don’t make a boss fight require a tactical nuke if you can’t be reasonably sure the player’s got one in his pocket.

Having a lot of character creation options (such as class choices or weapon loadout restrictions) can be particularly problematic.  If you allow low- or non-combat roles such as stealth, healing or diplomacy, then you need to ensure that players who choose these classes feel like their choices are respected and validated – but you can’t weaken the boss to favor them so much as to invalidate the player who has chosen the path of the bloodthirsty savage.  This gets even trickier in free-form character building systems, when the designers can have no guaruntees as to what combination of powers the player will bring to the fight, but should, in theory allow the player to finish the game with any they have chosen.

One more thing: if you do have an open world game, and you let me go level up some more after your boss thrashes me, don’t level up that boss to match my new level.  Nothing will make adventuring in the open world experience feel more futile.

In Conclusion

Why are so many boss fights in games still miserable?  I suspect it’s heavily a process question – bosses tend to be complex, and late in the game.  They have to come later in the schedule than the basic combat and advancement system.  Most are integrated into the game late, and get fewer QA cycles the later in the game they can be found.

If you believe that boss fights are meant to be the emotional highs in your gaming experience, though, this is no excuse.  Designers need to do a better job in identifying these boss fights, understanding their importance to the game’s flow, and respecting their role inside the game experience.  They need to build them early, playtest them often, and react in ways that results in boss fights being interesting, challenging, worthy opponents for the player.