Get the Desktop App for Battle.net Now
- All your games in 1 place
- Log in once
- Automatic game updates
While I was thinking that might be the case, the problem is we're not seeing a noticable difference between 10 and 20 APoC.
If rounding is the culprit and at 10 APoC we're getting 2 AP per crit, that could still explain why 10 APoC works. That would change my expected AP regen from 51.5 to 82.5, or 92.5 including the base regen rate.
Then, the difference between 10 and 20 APoC is pretty neglegible because you're already gaining more AP than you're spending.
Another way to test would be to compare say 8 APoC to 9 APoC. If the rounding is true there should be a noticable difference since 8 APoC would result in around half the APoC regen as 9 APoC. That's assuming you're just casting WW and nothing else to proc APoC.
If it is doing Ceiling rounding, then it would provide:
1 to 8 = 1 AP per crit
9 to 16 = 2 AP per crit
17-24 = 3 AP per crit
25-30 = 4 AP per crit
I am gonna try to not break the bank buying junk gear to get those breakpoints so I can test
here are my numbers, unfortunately I had to gimp myself a little to avoid spending money on crap.
8 apoc - 4.17x
9 apoc - 4.35x
Those don't really support the rounding theory, as it is completely linear through the breakpoint.
What's your 20 APoC multiplier Shaggy?
You really shouldn't use EA for farming unless you are running Ubers. The extra 5 CC from pinpoint is nothing compared to the effective dps from shocking aspect.
I didn't do any corner forcing or anything but ghom didn't move at all either.
here are the average runs I have done now.
email@example.com aps 43.5% crit - 4.17x
firstname.lastname@example.org aps 43.5% crit - 4.35x
10@2,76aps 43.5% crit - 4.77x
email@example.com 42.0% crit - 6.27x
firstname.lastname@example.org 42.0% crit - 6.46x
All done with bonechill sns no bubble.
Of course I am well below 50% crit, so that could be it.
Shaggy were you running out of AP on your tests at all (the ones with 4x multiplier)? When I did these tests at full windup, I never dipped below 70% of my maximum AP. The reason I'm asking is that I noticed your cc is lower than the rest of us that tried this test. So one potential hypothesis is that whatever "rounding" effect we're on AP returns from APoC might be a function of both crit chance and APOC.
Do you have a spare cc ring for your right finger? I can lend you a 9%ias and 4%cc ring. That should put you up to presumably the cc that Loroese had on his test.
EDIT: Hmmm... you and Loroese appear to have the same cc. So that potentially throws that theory out the window.
Edited by TekkZero#1963 on 2/17/2013 8:01 PM PST
I wasn't using my profile gear, but I listed the crit and aps I used for the run.
with the 8, 9, 10 ones I was out of AP a lot, it bounced erratically, leaving me unable to constantly cast twister, for sure.
with the 20 and 30 apoc ones, I just had a completely full globe constantly.
Edited by Shaggy#1317 on 2/17/2013 8:03 PM PST
Those are in line with the results we expected before, but I got completely different results. Mine are more like 6.32 for 20 APoC and 6.01 for 10 APoC, if I extrapolate my results to include what I'd expect with bone chill for about 66% uptime. To me, those multipliers are practically the same considering I only ran 2 tests for each so there is some varience present. That's a lot different than 6.27x vs 4.77x though.
When doing my tests, I did run him into the corner, so it's possible it makes a bit of a difference.
Instead of multiplier tests, what I was talking about was just doing tests with nothing but WW (and DS with crystal shell for mitigation). If your AP is continually out at 8 APoC but full at 9 APoC, it would stand your idea is right. I'm guessing that's not the case though because you didn't see much difference in effective dps for those values.
Ok, so I repeated the test, but this time dropped my cc by 7 (subbed in my FireWalkers to lose the 7%cc bonus from Nats):
*10APoC/106.6kdps/43%CC - 5.77x (82s kill time)
*18APoC/96.4kdps/44%CC - 6.70x (78s kill time)
This is what I remember from 1.0.6. The 10ApoC case behaves exactly like the cases in 1.0.6 where I couldn't keep a stable AP. My AP levels were all over the place and the DPS multiplier dropped substantially (finally...). So it seems like maybe cc is important in whatever APoC breakpoint we might be presumably looking at.
Edited by TekkZero#1963 on 2/17/2013 8:20 PM PST
@Ossian: Probably not since it seems like a breakpoint of some sort. Shaggy showed us values that fit what we expected from 1.0.6. And I was just able to replicate it by dropping my cc. I strongly suspect it's more to do with some change in how APoC returns are being calculated in the game and it seems to be nonlinearly impacted by cc (suggesting some sort of rounding).
10 APoC? interesting... might need to test it myself first after work :D
same here; already mentioned but lower APOC, higher damage wands, 1.55 black weapon/trium and S/C, etc.... quite interesting actually. Curious really about how much iAS vs. CC is needed. I kindof think it's in the 2.4 AS/55CC range...
RulerEric is all over this one ATM; he's thinking 2.38 is viable...
Threats of violence. We take these seriously and will alert the proper authorities.
Posts containing personal information about other players. This includes physical addresses, e-mail addresses, phone numbers, and inappropriate photos and/or videos.
Harassing or discriminatory language. This will not be tolerated.