I have only briefly expounded upon this idea in others' posts, never giving it the thorough analysis it needs - primarily because I lacked a model that elegantly and coherently resolved the issue without sucking.
Rather, I lacked an elegant and coherent model until now.
To preface, at present, our battlemechs, their limbs, and their associated weapons stay pointed exactly where the pilot intends, regardless of what maneuvers the battlemech is executing. Additionally, all of the battlemech's weapons converge to a single point when firing. I believe that this model is in desperate need of replacement.
Inaccuracy Mechanics:
Instead, I believe the devs should code the game so that a battlemech's torso sways while moving, in a predictable path. As a result, the pilot's view would move while the battlemech moves - but most importantly, the point where the battlemech's weapons are pointing would also move, in a predictable arc.
The torso targeting reticule would remain centered in the player's view, while the view deviates from the battlemech's path of movement. Thus, when the pilot fires their weapons when moving, the weapons will still remain perfectly aligned to the reticules - but firing will still yield a subtle deviation in accuracy.
Articulated limbs (arms) would cancel 50% of this movement-induced fire deviation. Since the torso reticule stays centered in the pilot's view, the arm reticule would drift in the opposite direction, cancelling half of the absolute angular deviation in the battlemech's firing pattern.
Since this motion is predictable, as player skill and hand-eye-coordination increases, this 'view drift' can be partially compensated for - at least for short periods of time. This also has the interesting side-effect that when extremely skilled players choose to manually cancel this movement to 'pinpoint fire', they can only do so with either the arms or the torso, not both simultaneously.
This can also be adjusted on a per-battlemech basis - for example, the Hunchback and Dragon have impressive arm movement range, and thus are 'more' articulated - it may be appropriate to buff the movement induced fire deviation reduction to 60% or so. Similarly, limbs such as the Cataphract's arms are 'less' articulated, and as such, it may be appropriate to reduce this accuracy bonus to 40%.
The question now is, how much should this view deviate? And what circumstances should be tied to the deviation? I believe that throttle percentage should be the primary factor - though tonnage and absolute maximum velocity should be secondary factors which subtly influence it.
This graph illustrates how throttle percentage state will influence movement-induced deviation. Keep in mind, this is the maximum deviation during the battlemech's gait - at 100% throttle, the battlemech's torso and the pilot's view will be shifted, at peak, by 2.5 degrees off-center.
I also feel that two other factors should be considered: Inertia (a property of mass and movement) and the top speed of the battlemech. If a mech is faster or heavier, it will be affected slightly more.
Suppose we have a Commando that moves at 125 KPH and weighs 25 tons. We divide tonnage by 18 and top speed by 22.5 to obtain two values; these are the number of percentage points that the throttle-movement-deviation curve is modified by. These are additive, not multiplicative values.
125 KPH / 22.5 = 5.55
25 Tons / 18 = 1.38
5.55 + 1.38 = 6.93%
Due to the Jenner's speed and tonnage, the speed-induced deviation is increased by 6.93% - meaning that the pilot's view and weapons will be shifted, at maximum, 2.673 degrees off-center.
A 65 KPH Cataphract suffers a 6.76% penalty to deviation based on mass and tonnage, meaning a maximum deviation of 2.669 degrees when at top speed.
A 45 KPH Atlas suffers a penalty of 7.55%, or a maximum deviation of 2.689 degrees when at top speed.
None of this needs to be calculated on the fly by the servers, and can all be handled in the Mechlab before ever assigning a battlemech to a match. The servers can calculate the maximum deviation when the mech is saved, and along with it, save the value by which to scale the mech's movement-induced accuracy.
Convergence Mechanics:
For accurately depicting how weapons on a battlemech actually WOULD behave in any semi-realistic and interesting way, we need a little bit of geometry. (unlike in most other cases, this is a case where implementing realism coincidentally creates GOOD game design, which is very rare, because usually the opposite is true). WEAPONS will never converge - only limbs will.
For convergence purposes, every battlemech will have an imaginary line extending from the cockpit, and 'forward' to infinity. This is the line that the pilot looks down. We will call this the torso vector.
Each arm (with weapons) will also have an arm vector. This arm vector will be roughly in the center of the polygon structure that comprises the arm, and may not necessarily pass through a fire point.
Weapons mounted on the torso will be fired PARALLEL to the torso vector, and will not converge. This means that the Hunchback's right torso AC-20, which is shifted a meter or two to the right, will aim a meter or two to the right of the torso reticule.
Weapons mounted on the arms will be fired parallel to the arm vectors. In the case of the Cataphract's right arm, which has one weapon a few meters higher than the other, it will be entirely possible (though rare) to have one beam strike the target, and for the second beam to miss.
If the limbs are horizontally articulated (capable of moving left and right) then the arm vectors will converge to the torso vector, either at the distance of the selected target, or if no target is selected, to the distance of the reticule.
If the limbs are not horizontally articulated, they will not converge at all, and the arm vectors will remain (horizontally) parallel to the torso vectors. This means that in order to pinpoint fire with non-horizontally articulated arms, the pilot must first fire the weapons on one arm, turn ever so slightly, and fire the weapons on another arm.
Yes, I have designed games before. This is what interesting game design looks like.
There are a number of other glaring issues with the game, which need addressing: http://mwomercs.com/...minimum-effort/
And, while we are at it, let's plug in some weapon values which actually work: http://mwomercs.com/...20#entry1477020
After that...the game is fixed! As in...playable and balanced! Meaning that the only thing the devs have to work on is the lagshield and new mechs! After that, the game is ready for launch.
Edited by Xandralkus, 12 March 2013 - 05:44 PM.