Is the basic argument in favor of digital joysticks that the state of the controller itself can be better determined by the user because there are only two states (per direction) and that the controller encounters a physical stop when the active state is reached? While I do agree that the absolute state of the controller is much more evident, that argument wouldn't be entirely
true. Based on some research I did when building a digital controller, I found that there is a specification (even on rubber dome switches common in gaming joysticks) that indicates how far the user can move the switch beyond the point where electrical contact is made. In other words, the game could respond to the active condition well before the user hits the physical limit on the switch. This physical phenomenon occurs with microswitches, leaf switches and I'd guess most types of mechanical "digital" push button. So, there's not technically
a complete correlation between physical stop and electrical contact in a lot of digital controllers.
There is a widely available joystick device in which electrical contact is practically synchronous with the physical limit of the switch mechanism. The construction of the Tac-2 affords such a feedback system as I think the OP is referring to. I don't personally care for that feel and would sell you my Tac-2 sticks if you wanted to buy them and I could locate them.
Our human joints aren't entirely compatible with the short-throw, hard-stop sort of physical man-machine interface. Typing on a membrane keyboard for more than a second or two would convince most people of that. I haven't done any formal study of ergonomics, but I want to think those who have would generally tend to support this position.
While the analog control itself is effectively working on a continuum of positions, it is possible for the user to ascertain the state of the device by how the game is responding. Thus, visual feedback gives the necessary information instead of strictly tactile feedback. I suppose it would have to take the human brain longer to process "is it moving as fast/far as I want" vs. "is it moving", but it seems to work well enough in many situations. There is also the sense of proprioception which allows us to know how far we've moved the input device. There's also in many cases a (slightly progressive) spring in the joystick which provides some additional tactile feedback regarding position.
In our analog world, there are many devices that I would not feel as comfortable controlling with a digital joystick. In the case of a piece of computer controlled construction equipment, all or nothing speed when swinging the boom would be pretty scary, not to mention hard on the mechanical systems. An advantage of analog sticks is that you can control the speed at which something happens and you can vary that speed according to conditions at hand. While the computer could handle the "startup ramp" to keep from tearing the machine apart by its own power, I might want to swing the boom slowly throughout its range. Since it's the same humans running the equipment and the games, I think the paradigm applies to video game control.
I really, really wouldn't want to drive my car with an all or nothing throttle, steering or brake control. (The analog/digital division gets fuzzy because a true digital encoder or similar device would likely be used to approximate analog control.) I feel the same way about digital controls in a driving game. I've never been in need of one, but I imagine an electric wheelchair is controlled by an analog joystick for similar reasons.
The position of an analog joystick can be used by the software to control the rate at which something happens or the physical position can determine the ultimate position to which the controlled object will move. Properly programmed, the rate of change of the input device could be detected and a single controller can control both where the target object is moved to and how fast it moves there. A digital controller doesn't really offer that level of user input. (A form of this is done with mouse acceleration in Windows, etc. so it obviously can apply to certain types
of digital devices excluded from this discussion)
Of course, you can build into the software a "ramp" that will allow an object to move progressively more rapidly, but that doesn't feel as intuitive to me. And tapping a digital joystick can provide a measure of fine control. Again, I have a harder time using that method.
Digital joysticks are simpler to implement, routinely made more compact and must be considerably cheaper to implement than an analog joystick. Those factors alone would tend to favor their use in video games (really, anywhere that they are practical). They've obviously been successfully deployed very widely. From this perspective, I would like digital joysticks better if I were a game manufacturer.
Clearly, move-at-a-fixed-rate-in-limited-number-of-directions control methodology is far better served by a digital joystick. Games like Pac-man that only support movement in very narrow, discrete directions and don't require variable speed are ideal candidates for a digital joystick. I agree with those who say that the design of the game dictates which control methodology works best.
All in all, if I were pressed to choose one technology over the other, I'd have to go with analog due to its flexibility. I can apply some mechanical stops to an analog joystick and set the software to trigger "all or nothing" at some point in the travel to act like an analog stick. I can imagine a possible way to implement an analog simulation with a conductive rubber dot type of digital joystick (assuming conductivity increases when more pressure is applied), but it'd be tricky if it even worked and would be somewhat limited.
NB: I am well aware that I haven't cited a single source of scientific evidence for any of this. I might be willing to read any scientific evidence presented to refute any of my opinions. However, the question in the poll does start with "Do you prefer..." so I choose to focus on preference and my reasons for said preference.
(BTW - did I miss the citation of scientific fact upon which the thread purports to be based? It's a long thread and I tend to miss things a lot.)
Edited by BigO, Mon Oct 25, 2010 7:10 PM.