As a designer, for everything you're trying to allow a player to control, you need an input. From there, it's just a choice of how much fidelity of control you want each input to have. We've come up with more and more choices over the years for creating inputs for digital devices, starting with buttons and joysticks as carry-overs from physical machines, eventually adding things like D-Pads, mice, trackballs, analog triggers, racing wheels, motion controls, touchscreens, voice controls and motion tracking. Today, we even have people like Emotiv
experimenting with direct brain control, EyeWriter
experimenting with eyeball-tracking, and Novint
making a unique device called the Novint Falcon.
They can all work. They can all work well. It's all down to a matter of how much control a game developer wants the player to have over a particular type of input.
There are amazingly fun games that can be controlled with exactly one digital button, like SFCave
These types of games often turn the player's lack
of input fidelity into the central skill and fun of the game.
As the intended interactions become more complex, so do a game's controls. You might think of a game like the original Super Mario Bros. as incredibly simple, but days upon days of work went into getting those controls perfectly right. It has no fewer than eight different types of input—go left, go right, down, pause, shoot fireball, run, jump and BIG jump.
A lot of thought went into how those controls were distributed, too, and much of the fun of Mario comes from how the control scheme makes certain types of interactions artificially difficult. Why not just make Mario run at full speed based on the arrow keys, or assign "shoot fireball" to the select key instead of the same key as the "run" key? Because it's more fun to have to balance your combat abilities balanced to constrain your mobility. Conversely, we have the "Up" key on the keypad unused, when it would make perfect logical sense for the "jump" command, but because we want jumping to be something Mario can do easily and at any time it gets its own button separate from the other movement controls so you can do both at once. We also have "jump" and "big jump" on the same key, because of the human instinct toward the force and/or duration of an input influencing the resultant output.
Nintendo didn't know that when they were first making Mario. They tried and experimented with those different combinations, until they arrived at the modern 2D Mario control scheme. Every game developer has to experiment with these things, to arrive at a control scheme that works for their game.
As games have increased their fidelity—as we've moved away from abstract concepts and two-dimensional interactions to three-dimensional worlds acting as escapist fantasies—gamers have come to expect more complex interactions in their games. Some games try to model this via a menu system. That detailed modeling is what got us pen-and-paper games like Dungeons and Dragons, a game people absolutely love, but one that requires patience and imagination on the part of the players to produce immersion. Many early computer games attempted to automate some of the more tedious parts of those pen and paper RPGs, resulting in turn-based and then eventually real-time games. The move toward real-time forced them to abstract, hide, or take care of more and more systems on the player's behalf, though. It's the balance any modern game developer has to strike with their control schemes: how much control do we give a player?
Some developers have tried to give gamers every possible control as a unique input. That resulted in the unwieldly behemoth commonly known as Steel Battalion.
Other, less crazy developers moved toward context-sensitive control schemes, allowing a single button to do one of many complex interactions based on the player's position in the world, or even taking control of the character and automatically performing an action required to maintain immersion without the player's input. Today, more than ever, developers have to struggle to figure out exactly what inputs are necessary for gamers to interact with their games and have fun in the process.
One of the biggest struggles is figuring out camera controls. Camera will always be a struggle to figure out when controlling something in a 3D space viewed through a fixed 2D plane. Even the fabled Star Trek Holodeck won't fix that, unless your idea of a fun time is to run around following whatever it is you're supposed to be paying attention to and being limited by your own physical capabilities when playing a game.
Joysticks have been used since before we even had video games, and have grown to be the fundamental input device of modern video game controls. That's no coincidence—more than any other control device we've invented so far, they replicate the full fidelity of anything with a two-axis range of motion while offering feedback of the input's limitations. That feedback is what makes the entire system not dissimilar to the motion of the human head. That's why the joystick has become the go-to camera control—it emulates the human head while working within the confines of viewing the action of a fixed 2D plane.
Over the past console generation, we've been sold the idea that motion controls are supposed to be intuitive. And they are, to an extent—you get pretty much instantly how a given control is supposed to work. No matter how intuitive motion controls are supposed to be, though, they're terrible at providing feedback when you've reached the limits of a given input, and that feedback is critical when you want to have a control system that feels tight and responsive. Nobody likes camera controls that feel loose and unresponsive, and there's a reason why every attempt to use motion controls for camera controls has been critically panned.
That's not to say that motion controls don't have their uses. They're the modern version of those old one-button controls where the limitations were the player's skill—though this time, it's the player's physical coordination that determines the limitations. Hell, I regularly pull up a version of that same Cave game on my phone that uses tilt controls rather than a single button to control the motion.
Back in the first paragraph, I mentioned a device called the Novint Falcon
. Novint advertises the Falcon as an incredibly immersive device, offering an incredible range of control and a high degree of feedback. I'll admit, the feedback it offers is impressive, but the range of control it offers makes it a member of a particular class of device—a 3D Joystick
So where do touchscreens fit into all this? They're in a weird position, to be sure. They can emulate pretty much any type of input, other than motion control, in whatever configuration a developer chooses—at the sacrifice of haptic feedback for those controls. In that sense, the 3DS circle pad is unnecessary (you can already use the touchscreen as a virtual second joystick) but one can easily understand why a gamer using the device on a long-term basis might prefer it.
The issue the circle pad brings to light with Nintendo isn't a matter of control scheme laziness, but rather falling into the same trap SEGA did back in the Genesis era that led to their downward spiral—an excess of addon devices for their systems, which have historically never been particularly successful when intended to be used for more than one game, rather than focusing on exploring the core hardware's abilities to their fullest extent (which is what Nintendo did and led them to some of their greatest successes).
As long as game developers want to use an input with a two-axis range of motion and minimal skill requirement on the user's part, the analog stick will be the go-to input device. It has been since the earliest video games, and no innovation in controller design has ever truly managed to supplant it. That's not a failing on the part of the hardware design industry, but a testament to the success of a well-tested design that's been in use since before the invention of the microchip. You may as well ask why nobody's managed to replace the wheel yet.