gamedesign.dev: Classes

The archetypal fantasy adventuring group, to most people, usually consists of a battle-hardened warrior, a crafty thief, a good-natured cleric, and an all-knowing wizard. Now, most games which have you play as a single character, or which emphasize a specific character in your party over others, will usually focus on the first of these, although that's not to say the second and fourth don't get playtime. So if people want to play characters in other roles, they have to look for games that allow for you to choose a "class" for your character, i.e. defining their abilities and skillsets. This is such a pervasive trend in RPGs that it's even bled over into other genres, particularly FPS. Problem is, not a lot of people are good at designing these classes with any kind of balance.

It's not an easy task, of course. Every element that significantly changes the playstyles available to a player has to be checked against all the existing ones to ensure that one isn't just outright better than another. There are a few games that have done this right (offhand, Team Fortress 2 and Golden Sun) but a lot of them are just not even pretending to be balanced. The most egregious offender of this is Final Fantasy III, perhaps more because it's looked at as one of the first game to really put this idea of choosing classes for your characters into play, but it is so loaded with false choices that they may as well have not bothered. If you can recall the Garuda boss who sends out an attack every few turns that only one of your sixteen or so classes can survive, you'll get the idea.

Another problem is that if you choose a class emphasizing one playstyle, they may suffer at the others so much that even offering them as an option is simply a mean move. Final Fantasy X, for example, with each character having a specialty that suits them, but a crippling inability to be useful against other enemies at all. They muck it up even further by giving the healer the ability to summon massive dudes that obsolete everyone else. Team Fortress 2 does this fairly right: even if a demoman's explosives can be used offensively like the soldier, he loses out on close combat thanks to having no sidearm. Which isn't to say he can't do it: his array of melee weapons is formidable and can be unexpected.

One last side effect is that certain playstyles just aren't very fun for some players. I've only ever seen one game make a dedicated healer role fun to play as, and that is, as before, Team Fortress 2. It emphasizes acting fast to save wounded players and prioritizing who can do and take the most damage in a fight before dying, as well as finding players who know enough to defend you and being able to protect yourself when you can't find such players. It's a surprisingly fast-paced playstyle which differs from the usual healer playstyle of "wait for their attack, use your strongest healing move, rinse and repeat until they get bored and go watch stupid shows". What makes this worse is that healers are practically required on most teams with class systems, even with consumable healing items.

Now, there's only one way to fix this one, and it's to question if having a class system at all is necessary or fun for your game. Hint: if you're designing encounters so that you can only clear it with a single specific setup out of who knows how many, you've done it wrong. If it becomes clear that one class is more overpowered or underpowered than the rest, then either you have to change it (without upsetting balance or making the class no fun at all to play) or remove it (I like playing as reavers in Dungeon Crawl: Stone Soup and I know they're trying to fix it, but if they can't then oh well). Lastly, make sure that the playstyle is at least fun for someone. If no one chooses a class because it's underpowered, that's one thing, because Dan is underpowered in Street Fighter and he's hilarious to play as, but if they don't choose it because it's no fun, you've done something wrong.