I was going to post this in the youtube comments but actually this is probably a better place.
I don't feel that Kanajashi's calculations go far enough to substantiate the claim that his numbers are any better balanced than PGI's. He doesn't actually factor in the number of heatsinks required to operate a laser, nor the range or duration of the laser, to produce a number which describes a lasers 'quality'.
What we want to do is to find a reasonable metric to describe how effective a weapon system is. And, yes, this is massively affected by your positioning and tactics and aim etc, but without a single raw figure describing each weapon, we're just throwing darts - and that's not a good system.
So, how about we _find_ a metric we can agree on? I'll start.
https://docs.google....vloCeo/pubhtml#
Note that Kanajashi's original DPS columns are not correct. MWO does not include the beam duration in the listed weapon cooldown times, so you have to add the duration to the cooldown to get the complete cycle time to find the correct DPS of a weapon (HPS too).
I've added some columns at the end of the sheet, they are:
- Damage Per Second / Tonne - adjusted for heatsinks required, assuming a really low number of sinks (i.e. a hot 'mech)
- DPS * Log₂Range * Log₂Damage Per Duration / Tonne - adjusted for heatsinks, assuming a really low number of sinks
And then I've copied them a couple of times with different values to represent how many heatsinks you might expect to require to operate a weapon system. dps/t vs heatsinks per system would be an interesting set of graphs if somebody feels like it.
I chose 1.5, 2 and 3 tonnes of heatsinks per heat-per-second because these seem reasonable values to actually field. Perhaps too high, but for approximating shorter fights they should be ok. It actually takes 1/0.14 = ~7.1 tonnes of heatsinks (assuming doubles) to completely sink 1 heat per second, which is only one medium laser - but all 'mechs have a baseline 2hps sink rate and so a normal-ish 6 medium laser loadout producing 6hps only needs about 2/3 of the heat produced by a laser to be sinked by additional heatsinks beyond the minimum.
This is obviously a fairly rough approximation of actual heatsink requirement because
- the proportion of additional heatsinks you need depends how much total heat you need to sink - since the engine mitigates a flat 2hps, not 1/3 at all times
- maps are different temperatures
- jumpjets make heat - quite a lot of it
- being completely heat neutral is insane or impossible anyway. But somewhere between 2 and 3 tonnes per laser is probably fine, certainly as a guide. It affects the absolute numbers, but not trends.
The other columns contain my metric for 'goodness' of a laser.
Clearly more range is better.
Clearly 600m isn't _twice_ as good as 300m. probably.
Ditto for damage per beam-second-duration.
So; logarithms!
Log₂Range and Log₂DPS use base 2 for their logarithm kinda arbitrary. Finding the actual correct bases to use would take some experimentation, but base 2 will do as an example. Feel free to try alternate bases, or add a cubic term to represent any advantage that extreme long range might provide.
After adjusting each laser to account for the heatsinks required to operate it, and the range and duration of the laser we have a single number for each laser's 'quality'. How well does this actually work?
Well... Take a look at the numbers. I ran them for both current and proposed values.
Would you say that (presuming you can get into range) the cSPL is a good weapon? Well that has the highest numbers using this formula. Do you often use IS small lasers? ER-larges? Well they have the lowest numbers. That's by no means an absolute proof this is perfectly accurate, but the approximations seem to hold true.
Given the relative values of the IS medium and MPL (MPLs score about 20% better) you can see I'm probably under-valuing range slightly. Maybe. Bear in mind that what we want is even numbers for everything - it's ok for a SPL and ER-large to have the same number but for the SPL to always win in a straight 1v1 fight provided the ER-large gives an advantage of equal value in some other way - the ability to peek or kite or whatever.
What would be useful here is a poll for people to rank how effective they think each laser is. Then we could compare that with what I've got to see how closely it matches.
Anyway what's clear is that there are distinct outliers in the current weapons, and Kanajashi's alterations aren't fixing all of them. Nerfing all the clan pulse lasers a bit is fine. But why nerf the ER-larges? Why not buff the IS small laser? Do LPLs really need a 15% nerf?
I've included a set of my own values for comparison - they're only baselines and they are only as accurate as the model I'm using, but it's not too hard to push things into what looks like a reasonable shape. I've even thrown some values in for the upcoming weapons - not predictions, I just wanted to see what sort of values would be needed to fit the trends. Unsuprisingly the required fixes seem to be increasing the range of IS lasers and reducing their heat - exactly what quirks have been doing for years now.
I've flattened the curve off for small/medium/large lasers and tried to drag them all towards a central value; they're within 30% or so instead of >200% of each other. That might be too flat if we want to actively encourage a more brawly playstyle where short range lasers are supposed to be _really_ notably better, but the IS/Clan balance should be more reasonable. Or it would be if IS 'mechs weren't mostly covered head-to-toe in armour quirks. Some of that is because hardpoints but some of it is probably because their lasers are just worse.
Anyway, thoughts, comments, flames?