MustrumRidcully, on 18 May 2013 - 02:15 AM, said:
That's not a spreadsheet.
Phaesphoros, on 17 May 2013 - 02:43 PM, said:
Well here's what my spreadsheet says about the PPC
MustrumRidcully, on 18 May 2013 - 02:15 AM, said:
This isn't either. It's a chart, a graphical representation of a spreadsheet (or at least some of its results)
[chart]
The chart shows how much damage you get for your weight out of a number of weapons of the given type if you want to deal a certain minimum amount of damage within a given time frame, if you consider the heat sink requirements to fire for that duration and the ammo cost for firing that duration for a given number of engagements.
You also mention the reason why I didn't insert more complex metrics in my post:
MustrumRidcully, on 18 May 2013 - 02:15 AM, said:
The generally considered underpowered AC/5 for example completely otperforms the PPC at a 12 second engagement time. Too bad that snipers don't try to alpha for 12 seconds, and the 4 second mark is more realistic for them...
I.e. more complex metrics rely on more complex interpretations. In your metrics, the interpretation requires scenario-dependent parameters. It is necessary then to compare those scenarios, which is also highly non-trivial. For example: a brawler configuration typically is optimized for a high sustained dps, whereas a sniper configuration relies on high peak dps (or high alpha). Even in a scenario limited just by duration where both opponents are at the optimum range of their weapons, the brawler doesn't necessary win, because the peak dps of the sniper might be high enough to take out the brawler before overheating.
Therefore, I refrained from using more complex metrics. My interpretation from those simple metrics is that the peak dps of PPCs and the heat efficiency is too high for a sniper weapon. It wouldn't be viable if the alpha damage would be reduced, or if you couldn't shoot three to four of them at the same time (heat). But you could, for example, raise the cool-down time (IMO it's ridiculous that's it's lower than LPL atm).
But like with all claims, there's only one way to find out: test it. There can be side-effects you have not considered in a (complex) metric.
jeffsw6, on 18 May 2013 - 03:12 AM, said:
Where are you getting those figures? At optimal range, the PPC is actually quite efficient, at 1.25 dmg/H. This compares to 1.00 for ML (and SPL), 1.29 for LL, and 1.37 for LPL. For the lost readers, more dmg/H is better.
ML: 5 dmg, 4 heat -> 1.25 dmg/heat
For the rest, I agree. Those are also the numbers I have in my spreadsheet.
jeffsw6, on 18 May 2013 - 03:12 AM, said:
Also keep in mind that the LL has the best heat efficiency across a broad range, from 0 to 450; and continues to be pretty decent dmg/heat further out. The PPC is much better dmg/H than the LPL as you exceed its 300m range.
LPL has the best heat efficiency (after SL), but it's only 10 % better than PPC: (1.37 / 1.25 - 1) * 100 % = 9.6 %
I agree about the statements on PPC range.
jeffsw6, on 18 May 2013 - 03:12 AM, said:
If we were only looking at "spreadsheet-warrior," this might seem to be a minor limitation which is offset by the < 90m drop-off of PPC; but unfortunately, most enemies don't wait until they are at 300m to begin shooting you. The maps we have encourage long-range combat and don't really promote short-range brawling; a 300m weapon is just inferior.
Also, 300 m are brawling range: ML, AC/20, SRM, SSRM. Of those, only the ML has a worse heat efficiency than the LPL.
The LPL as a brawler weapon IMO is not a good choice atm if you can use an AC/20 (compare 2 LPL vs 1 AC/20) or SRM instead.
Edited by Phaesphoros, 18 May 2013 - 11:21 AM.