Posted 09 June 2013 - 01:22 AM
Basically, the correct solution to pin point high damage weapon loadouts is divergence. Heat is by far the worst (and most popular) solution given how many other issues it introduces while failing to completely (or even) solve the primary issue. Weapon hard point restrictions is a little better - but makes mech load outs a little too rigid. Which would be ok if we had hundreds of mechs to choose from, but less good when we don't. Also it doesn't stop 'natural boating' - so it'd really just reduce the number of mechs in the top tier.
Spatial divergence can be iffy - but I think this is probably the best of the solutions for introducing spatial divergence (it's actually a hybrid spatial/temporal divergence idea).
But the implementation of it requires some insightful forethought in order to pull it off correctly.
So the idea is that the weapons start off at parallel to the reticle. The longer you hold the reticle on target, the more the weapons converge - faster for fully articulated arm weapons, slower for no elbow arms and torso mounted weapons.
The trick though - is that this only applies if the player is firing weapons from multiple hard points at once. If firing from a single hard point, then it automatically converges on the reticle irrespective of shot location.
But what about players the circumvent it by setting 2 PPCs on the left arm to one group, and 2PPCs to right arm to another weapon group and then firing them in short succession (i.e. as fast as you can click on mouse 1 and 2)?
The first fired weapon group would set the convergence point, while the second group would be offset the distance from the left arm group minus the distance that has already converged.
The main problem with this solution is that it's not transparent - as in, if you don't know exactly how or why things are behaving as they are, it can be quite difficult to know why your shots are going off the way they are. But with adequate messaging from the developers, even this barrier can be overcome. And it's certainly a damn sight better than having to deal with massive pin point damage that is in a very real sense, game breaking.
So what about the implementation for convergence? How long should it take, and how on target do you need to keep things?
Obviously, it can't be the case where if there's any movement whatsoever, the convergence is thrown off entirely. If this was the case, the only time you'd be able to converge is when aiming at a shut down opponent. Otherwise convergence would only start when you're dead still AND the opponent isn't moving around.
Something more like the behaviour of streaks - where you need to keep the target under reticle for 2-3 seconds before been able to converge. Moreover, unlike streaks, you can only converge if pointing directly at the target; if your target moves behind cover, convergence is broken and the process needs to be started again. So, if you want to snipe with your AC20s, Gauss, PPCs, etc, you'd essentially have to spend a few seconds vulnerable out in the open. Or you could just pull the trigger, accept divergence and minimize your exposure risk.
If there is a problem with this implementation, it may be in the server load - i.e. each mech has a continuous invisible beam (not unlike TAG, except invisible) which would be traced between the player and the target in order to check for line of sight. But given the number of lasers and tags that can be loaded on mechs, I don't think 16-24 extra beams is going to impact that heavily on the server.
Anyway, the longer you hold the more on reticle your weapons converge. Once locked (golden reticle), like streaks, there's a cone that you can stray from before you break convergence. So if you're a sniper and you stand out in the open long enough, your subsequent shots should be more accurate; commensurate with the risk you're taking.
Finally, if you fire weapons in one body location - the convergence is only set for as long as the reticle is still (or within a small range of movement). As soon as you start moving your reticle, the convergence is reset for the new body location (but the convergence timer will continue ticking down as long as your reticle is on target/tracking the enemy)
So that means instead of shooting your AC/20 in your left arm of your Jagermech, then swinging your torso to track the enemy, shooting the right arm AC/20 - and having your shot go stray (because you're aiming with the reticle without accounting for how much convergence there is left), the first AC/20 shot goes where you point, you track, and the second AC/20 shot goes where you point.
On the UI front, the reticles now grow and shrink is size commensurate with the degree of convergence that the player has achieved. They turn gold once full lock is achieved.
So the end result here is a system that encourages group firing by weapon location and serves also to buff brawling - in a brawl, you're typically firing by weapon location anyway, keeping up a steady stream of weapon fire - moving your reticle to track the enemy as you dance around each other. You'll be completely unaffected by divergence. OTOH, if you're a sniper, you'll have to make yourself more vulnerable in order to have a pin point alpha, or you'll be forced to fire the weapons in appropriate groups and adjust aim manually to compensate.
With this sort of implementation in play, it introduces a subtle increase in the level of mastery required to play the game. New players will not be tremendously disadvantaged - while experienced players will adapt to the changes as required. It's got a nice balance between in universe logic, game logic and {Noble MechWarrior} logic, while actually doing a fair amount to solve the primary meta problem plaguing the game.