Quote:
I forget most of my statistics. Isn't the chance of it firing at least once in SB equal to (1-(60%*60%))? Or 64%. Where as the probability of both SB's triggering the proc = 60%*60% = 36%.|
Originally Posted by Nihilii
CU: (25 / 1.59 + 2,376) * 4.5 / 60 = 1.3572, or 90%, as it is apparently the cap according to patch notes.
SB: (8 / 1.59 + 1.584) * 4.5 / 60 = 0.5975, or 60%. Making SB the better choice, given 2 SB uses for every CU. Right? |
So... let's assume 100 pts proc damage .... I can't quite calc expected value because I don't know how to account for the chance of both occuring ...Someone smarter than me probably can. The following expected values are probably wrong
EXPECTED VALUES
SB (2 SB's per attack chain):
Both procs: 36%*100*2 = 72 pts
At least one proc: 64%*100 = 64 pts
CU (1 CUper attack chain):
90%*100 = 90pts
I don't think you really need to delve into the math too much, as these are independent events. Just multiplying the average damage for one use by the number of uses is good enough.
So I'm doing
2 * 105 * 0.6 = 126 average damage from purple proc in SB
105 * 0.9 = 94.5 average damage from purple proc in CU
You can get to the same result with your way, just takes longer.
Chance to get exactly 2 procs (210 damage): 0.6 * 0.6 = 0.36
Chance to get exactly 1 proc (105 damage): 0.6 * 0.4 + 0.4 * 0.6 = 0.48
Chance to get no proc (0 damage): 0.4 * 0.4 = 0.16
0.36 * 210 + 0.48 * 105 + 0.16 * 0 = 126
PPM sure seems like quite the boost... Especially for stuff like Achilles' Heel. Keeping -20% res perma doesn't seem so out of reach anymore. 
Also going to make for some unintuitive build choices... For example, I just realised removing a second recharge slot in BU would make more sense in my planned build, as it'd raise the Gaussian proc chance and the extra 3 or 4 seconds of recharge I got from that recharge slot wouldn't allow me to use any more often with the particular chain I'm considering.
Quote:
Yep.|
Originally Posted by Nihilii
I don't think you really need to delve into the math too much, as these are independent events. Just multiplying the average damage for one use by the number of uses is good enough.
|
Looks like you've got things pretty much right with your edit, except the PPM formula uses just the animation time, it doesn't account for Arcanatime.
Thanks! I suppose a ~0.1s to 0.2s difference won't change much overall, but it's always better to have the right values.
Based on this, the simplified formula for ST attacks is
Proc Chance = (Enhanced Recharge + Activation Time) * PPM / 60
So let's say I have a Street Justice scrapper running a CU SB RC HB SB RC chain.
CU recharges in 7.5s and animates in 2.376s.
SB recharges in 2.5s and animates in 1.584s.
A purple proc (4.5 PPM as of this writing on I24 beta) would have a
(7.5 + 2.376) * 4.5 / 60 = 0.7407, or 74% chance to proc in CU.
(2.5 + 1.584) * 4.5 / 60 = 0.297, or 29.7% chance to proc in SB.
CU remaining a better choice than SB even with 2 SB uses for a single CU, in this particular example.
Am I doing this correctly?
Edit: guess I'm not, as it's supposed to be just the enhanced recharge.
...This is why thinking outloud is good, even if it makes you look like a fool. Likely, I would have considered it correct without looking at it anymore and moved on if I just kept to myself.
So looking at it again, it'd be
+59% rech (Heca set without the dam/rech)
CU: (25 / 1.59 + 2,376) * 4.5 / 60 = 1.3572, or 90%, as it is apparently the cap according to patch notes.
SB: (8 / 1.59 + 1.584) * 4.5 / 60 = 0.5975, or 60%.
Making SB the better choice, given 2 SB uses for every CU.
Right?
...Wow. Procs are becoming really, really strong, aren't they? Unless I'm missing something again. I hope I'm not.