Player control over game avatars (of whatever sort, from the Galaga ship to Sora to a WoW avatar) is part of the User Interface. It’s not what we might typically think of when addressing UI design, but the mechanics of control are nevertheless crucial to making a game work. The GUI (Graphical User Interface) is usually what we see and think of when talking interface, but it’s only part of that user-game interaction. We need to be able to actually control the game with some sort of input.
One of the recurring pet peeves I have with games is where I’m expected to direct my avatar rather than drive them when the gameplay itself doesn’t work best that way.
A Kingdom for Keflings (by NinjaBee, the company I work for) brought this dichotomy into sharp focus for me. The game was originally released on the XBox as a Live game (downloadable), but we also made a PC port for it. The design goal for the PC port was to make it purely a mouse-driven game, and if possible, single-click. The XBox version uses a gamepad controller, and it has a lot more buttons to work with, though much of the actual gameplay tends to use just one button. (An “interact” button, effectively, though we also needed a “cancel” button for menus and assorted GUI. The PC port uses the keyboard for some of those other functions.)
Making the game mouse driven means that there is no parallel to the control stick of an XBox controller. The player avatar (a giant among Kefling villagers) simply follows the mouse cursor when prompted to (via a click or drag mouse function). The player directs the avatar’s actions. This is in stark contrast to the XBox version, where you drive the avatar with the controller, their motion controlled by the analog stick. For me, it is easier and more fun to play the XBox version because of this precision, and ultimately, it just feels better. It’s the exact same game with the same core game mechanics, but the sort of control I get with the XBox just works better for me. The core game is still fun, but the method of interaction on the PC isn’t something I like.
I also find that making the avatar control more abstract as it is in the PC version puts another conceptual layer between me and the player character. The XBox version uses the Microsoft player avatars, effectively putting the player’s persona into the game. The PC version doesn’t have that option, but even then, the pure mechanics of the controls sets the player further back into a role of a director of a giant who then directs Keflings, as opposed to the XBox role of a giant directing Keflings. To be fair, both really are you as the player directing the giant, as in any game, so the levels of abstraction are at least similar, but mechanically, when the giant (player avatar) in the PC game is following an element of the GUI to interact with the game world, that’s one more small distance between you and the game world, one more subtle push out of the suspension of disbelief.
On the other hand, our Band of Bugs also started on the XBox, and was also ported to the PC. That game works well with either control scheme (mouse/keyboard or controller), since you’re never actually driving your characters to start with. The XBox and PC controls are different, and have different pros and cons, but they feel pretty similar in the long run. That’s one nice thing about a tactics type of game where it’s all about direction in the first place.
I think this is also why some MMO players complain about the “click to move” control scheme found in some MMOs (usually Asian ones, like Atlantica Online) as opposed to the keyboard WASD movement scheme. The former has players directing their avatars in the game world by telling them where to go (and a pathfinding AI takes over), the latter has players driving their avatars around the world. It’s a more visceral level of control, and it seems to be more satisfying. (Tangentially, I am curious about the cultural implications of this difference, but have little data to examine.) This is also tied to the oft-repeated complaint about Guild Wars characters not being able to jump. Many players just want that control. They want to drive.
The difference between the two is also what I believe to be a major factor to why I think Amorphous and Recettear‘s combat have significantly different feels to them. Andrew rightly noted the similarities between the two over here (rightly complaining about some of Recettear’s warts), but I’ve tried to describe why I don’t see them as being all that similar in gameplay. The Amorphous avatar just follows my mouse cursor, but I get to drive Recette’s dungeon diving compatriots. It’s a subtle thing, but it makes a world of difference in how a game feels to me. (And tangentially, the default keyboard controls for Recettear aren’t good, remapping is silly… but with a gamepad it reportedly “just works” and works very well.)
There are also games where the whole point is to direct the character, intentionally abstracting the controls to allow for different functions, like Aquaria or Machinarium, and others where you’re not really meant to have a high level of individual control, like Lemmings. Still, the control interface really can have a pervasive if subtle effect over how a game is played and how it feels. Neither directing nor driving is the solution in all places, and indeed, applying one where the other would be more appropriate can be a problem. It may not be a gamebreaker, but it can be important to the tone and feel of a game.
Also, as these guys note (adeptly and humorously), when you’re in the driver’s seat, it changes a lot of things. The abstract director role lets you internalize things differently from the driver role. (And similarly, playing “yourself” in a holodeck would push things even further and have different psychological implications.) Matching game mechanics to storytelling intent is something that most games just don’t do well. Matching storytelling to UI can be even trickier because there’s less to work with, but it can be more important.
UI design is a tricky, sometimes subtle thing. Still, I believe it’s crucial to get right, or else a game just won’t work right. Giving players control is key to making games work, and if the controls don’t work well, all the pixel shaders and voiceovers in the world won’t make a difference. Games are interaction, and if that interaction is inept, a game can crash and burn, and it may not even be clear why.