One of the more surprising things that I found is that pin-point accuracy actually is supported. The caveat is that it is far from instant. From what I can determine, MWO's convergence system is theoretically a passable interpretation of this. Don't stop reading, though, it's not that simple.
In the fiction, targeting accuracy is line-of-sight based. Detecting something on sensors doesn't automatically allow the pilot to hit it. In order for the 'Mech to ensure that its weapons will hit the target, the pilot must hold the targeting reticule on the body of the target for a period of time. The precision of the shot depends upon the amount of time that the reticule is held on-target before firing.
If the pilot fires immediately after placing the reticule on the target, it is unlikely that all of its weapons will hit. Intentionally striking the location directly under the reticule is almost impossible.
If the pilot holds the reticule on-target for a few seconds before firing, the vast majority of its weapons will hit the target 'Mech somewhere. The location directly under the reticule will probably be hit by at least 1 of the weapons, but never all of them.
If the pilot holds the reticule on-target for an extended period of time (10+ seconds), pin-point accuracy becomes possible against a stationary target. The best example of this was in the case of a hostile 'Mech taking a hostage in a city. An Atlas arrived to find the enemy 'Mech with its fist poised above a civilian vehicle with people inside. The pilots of the two 'Mechs exchanged several statements before the Atlas could fire. While they were talking, the Atlas had targeted the other 'Mech's cockpit, waiting for the Gauss Rifle in its right torso to achieve pin-point accuracy. When the Atlas fired, the Gauss slug destroyed the cockpit instantly, leaving the rest of the 'Mech undamaged. This demonstrates some important points: Torso weapons can converge; Pin-point targeting is possible; Accuracy increases as the reticule is held over the target.
If we want to emulate this system, there is one very important thing that can be done in MWO: slow down the speed of convergence. The speed suggested in the fiction is probably too slow for MWO, however. I would suggest having the crosshair on the target for maybe 5 seconds for convergence to reach pin-point accuracy. That would force snipers to spend more time exposed before firing if they want to hit their target precisely. It would make it more difficult to hit smaller, fast-moving targets at long range, improving the survival rate of light 'Mechs. It will also make snap-shots at targets that suddenly appear from behind cover much less effective, making the use of cover more important.
It would remain possible for players to fire immediately, before their weapons converge fully, but they would lose the ability to precisely hit a single armor section except by luck. On the other hand, it would actually help faster 'Mechs maintain accuracy at short range. Because convergence moves slower, holding the crosshair on-target to get an accurate shot and then rapidly moving off-target while maneuvering, then aiming on-target again would give less time for convergence to deviate from the desired aim-point.
Slower convergence should even help ballistic weapons, which require leading the target. With the current system of rapid or instant convergence, leading targets is difficult because your weapons will converge at whatever point is under the reticule immediately, even if that is a building 300m behind the 'Mech you are leading. With my suggested adjustment, a player could place the crosshair over their target while their weapons converge, then move the crosshair off-target to lead the shot. Since convergence is slower, it should be possible to fire before the aim point adjusts very far away from the intended target. This would have a similar effect to 'locking' convergence at the distance of your selected target with current convergence speed, without the need for any extra coding.
Another effect of slower convergence would be a change to the risk/reward for jump-sniping and Alpha-Strike builds. In the case of jump snipers, their weapons would converge on the cover they are hiding behind while they wait for their jump jets to recharge. Thus, in the brief time that they can actually see their target while jumping, convergence would not have time to adjust to a target more than a couple hundred meters away. The only way to get around this would be to expose the cockpit of the jump sniper while they wait for their jump jets to recharge, so they can put the reticule on-target early, which also means that they would be vulnerable to return fire during that time. Similarly, holding the reticule over the cover that a jump sniper is hiding behind would move convergence much closer to the jump sniper's 'Mech, so that they are easier to hit by players waiting for them to pop up again.
High-Alpha builds often function similarly to jump snipers, staying behind cover while their weapons recycle and their heat dissipates. As long as a team mate keeps their target locked, they know exactly where to aim the moment they emerge from cover. With slowed convergence, this is impossible. Instead, they would be forced to expose their cockpit to put the reticule on-target while they cool down, making them vulnerable and giving away their exact position. At the same time, this would allow their opponents to let their weapons converge either on the exposed cockpit or the cover they are hiding behind, shifting the targeting advantage away from the Alpha build. Instead, 'Mechs that have a variety of weapons that can fire while keeping the crosshair on the target should be competitive again.
If this were to be implemented, it might be wise to have some indication of weapon convergence on the crosshair. Maybe the crosshair could turn gray when convergence is aiming at the distance under the reticule, only showing gold when convergence is within 25-50m of the target point? Something to that effect, to make it clear when a shot will be accurate.
It's certainly not a perfect solution, and it won't solve all of MWO's problems by itself, but it might move gameplay in a good direction.
Taken from another thread for independent discussion.
(EDIT: Target leading effects clarified.)
Edited by Renthrak, 26 April 2013 - 12:50 PM.