Good comparative analysis. I've put my thoughts in reddit, but generally, I don't know how much faster of a spin I would want. Too much faster would feel unrealistic. I watch CoD gameplay and it just seems like everyone walks/sprints around like their gun is on a steady-cam and their head is on a swivel. It doesn't even seem like you are walking, but you are just a camera with a gun.

The way they describe input for the controls is that 'sensitivity' corresponds to a plot directly relates to the speed you will move your crosshairs along an axis relative to the distance the 'stick' is from 0. They have a lot of ways that they could plot this graph. If it is just a diagonal line with 1% of the distance away from 0 corresponding to 1% of max speed, then you'd have 50% correspond to 50% of max speed. I think they've described this plot as sub-optimal, because people don't seem to spend a lot of time in the 'speed range' of 20% and 100% max speed. Typically, players are more concerned with fine adjustments.

So what they become concerned with is how finely they plot that lower 25% of sensitivity. Do they put it mostly in ADS? Do they put it in the first 50% of the stick's movement zone, or in the first 75% of the zone. How do they tune it for each weapon that will have different maximum ranges and likely tracking speeds for different targets. If you can adjust the sensitivity for the controller, does that correspond to a flattening of the whole curve, or two a reduction of the max turn speed, and therefore a 'lowering' of the curve within the plot?

http://www.youtube.com/watch?v=2VHlBwFp3xM

Void Echo put this in another thread. As of yet, your DS3 examples haven't really included Aim assist examples. Still, your point remains about the fluidity and expectations of speed management, but as far as comparative target acquisition, I think you still have a burden to prove a major significant and strenuous difference.

You get that the input difference between mouse and DS3 is completely different though right? One is raw input the other is indirect. AA is to compensate for the fact that the raw input lets you be more precise. Trying to control turn speed with a DS3 would be like trying to drive a car with a mouse, the input does not match the style of output that is reality. I think it is completely fair to have a system that compensates for this lack of direct control. Any tuned curve that you can tailor could semantically be twisted to be considered 'aim assist' because sticks on contollers will never be 'direct input' as they represent a force to move, not a limited distance that the crosshairs should move.

Direct - move mouse 5 inches, and the distance your crosshairs move will be the same most of the time.

Indirect - the distance your crosshairr moves for the controller is related to how long it spends in each section of the 'acceleration curve'

Mouse is like driving a stick shift, you acceleration inside each gear is always the same.

DS3 is like driving an automatic transmission. Your acceleration is subject to the pre-programmed CPU's definition of when to shift.

The challenge with designing an input system has a lot to do with how we perceive acceleration (obviously the change in speed) vs. distance. With MKB, your strafe speed isn't even linear, it's binary. Hold D for one second and you always move the same distance. You are moving at max speed or you aren't. Then based on you mouse settings, the speed that you move your mouse at will always return the same output because mouse distance will be proportionally represented by cursor/crosshair move distance. With DS3 you cannot control distance nearly as directly. You only can control speed.

Think about how your brain is interpreting this then when tracking an object while using KBM. Your ground speed becomes MUCH more predictable since it is binary, then all you really have to try hard to track with it is the movement rate of your target. The acceleration of you cursor feels much more like an extension of your arm/wrist.

With DS3 you have an acceleration curve that is 'slippery' for both your ground acceleration, your target's speed, and your crosshair acceleration. I wouldn't be surprised if the CoD 'left thumb' deadzone, or reduced movement zone, is much smaller. This could be a major factor. If you reach max speed, (or near it) when you have moved your left stick to only 25% distance from center, then the strafing feeling much more closely approaches the ADADAD feeling of the keyboard. You still have to pass through the deadzone, however, causing highly variable ground speeds. Which, when combined with target movement, throws off aiming.

You COULD effectively say that with MKB your brain just has to 'predict' acceleration of the two axes that the target is moving in in the distance (left to right/up down). Direct crosshair input and binary speeds on the ground means that acceleration control is more easily managed.

With DS3, your brain must compensate for the acceleration of your body, crosshairs, and target in about six axes. This is not really normal for the mind. AA 'cohesion' exists to compensate for the acceleration 'disruption' caused by variable speeds in foot movement (because KBM has none). AA 'magnetism' exists to compensate for finer acceleration controls because your brain is registering 'distance' when you can only can control velocity and acceleration with DS3. You have more control over aiming 'distance' with mouse, and nobody can argue otherwise.