Honestly, I think the
least OP approach to a TarComp is to override the current way convergence works.
OK, right now convergence works by making all your weapons hit at the distance your reticule is hovering over. Be that on the mech in front of you or the terrain behind the mech you are targeting. If you are leading a target at distance, that means that any projectile style weapon system will not converge on the target, but behind it, with a spread dependant on how far away the background in the lead was. This means you could do anything from an unfocused hit with your weapons, to hit with half of your weapons, to a total and complete miss in the case of some skinny fast mechs.
IE: Enemy mech is at 500 meters, your reticule is over terrain 750 meters away, your shots converge at 750 meters.
I would propose a TarComp set convergence to your
currently selected enemy. You lead a target, your shots are still converging as if you were aiming at the target, and if your lead was correct you hit with a focused attack as opposed to a spread one regardless of how far you had to lead or how far the background was relative to the target.
IE: Enemy
targeted mech is at 500 meters, your reticule is over terrain 750 meters away, your shots converge at 500 meters.
This offers no easy mode lead indicator. No easy mode hit box overlay. Hell, it does not even require any new graphics or icons to be added to the game. It punishes players who refuse to use the R key, but Clan pilots are supposed to be skilled enough to smash their face across the keyboard once and if the player cannot hit the damned button that is on them.
Edited by Pariah Devalis, 20 May 2014 - 02:00 PM.