Diablo® III

APoC why do we need so much?

Without Meteor, I am finding that I can get by pretty damn well with only 10 APoC.

I'm at 2.95 APS and 50% CC and I don't run out.

Am I just not paying enough attention or are we all grossly over speccing our APoC requirements now?

Standard SNS or CM/WW does not seem to need big APoC numbers in my opinion now.

P.S. If I run meteor I need a lot of APoC and Prism to keep it going without running out instantly
Reply Quote
Without Meteor, I am finding that I can get by pretty damn well with only 10 APoC.

I'm at 2.95 APS and 50% CC and I don't run out.

Am I just not paying enough attention or are we all grossly over speccing our APoC requirements now?

Standard SNS or CM/WW does not seem to need big APoC numbers in my opinion now.

P.S. If I run meteor I need a lot of APoC and Prism to keep it going without running out instantly


What about single target situations? ie. vs shaman and succubus?
Reply Quote
I'm at 2.76 aps and 59.5% CC, and I think I could drop to from 19 to 17, maybe even 16 APoC, but I definitely wouldn't even consider 10 APoC..
Reply Quote
It depends on your stats when I use my slorak's my energy twister crit chance is at 78% with only 10 apoc and 2.78 aps still works fine.
Reply Quote
I run out of AP with 16 APoC 50 CC vs two targets.

I will never consider 10 APoC with standard SNS. However, Prism + Conduit allows you to stimulate 16 APoC pretty well.
Reply Quote
I'm EXTREMELY confused... I came into the thread hoping to show quantitative proof that the OP was wrong and that 10APoC would impact our DPS multiplier, since that's what I measured in 1.0.6 Ghom Tests. In 1.0.5 and 1.0.6 this would have never worked, but have you guys done a Ghom Test in 1.0.7? I just did four MP7 Ghom tests with either 10APoC or 18APoC and I came to pretty close results (2.74aps, 110-130ms Latency):
*10APoC/111.4kdps/50%CC - 6.57x Multiplier (took 67, 69, 69s, and 71s respectively)
*18APoC/100.5kdps/51%CC - 6.70x Multiplier (took 73, 74, 77, and 79s respectively)
In both cases at full wind up, I didn't run out of AP on the tests. This isn't what I saw in 1.0.6. Can anyone else replicate these results? If it's true, 10APoC doesn't seem to be as outlandish in 1.0.7 as it was in previous patches. I dunno why it's different though and clearly something is also different with the uber Seigebreaker fight... has something actually been changed?
Edited by TekkZero#1963 on 2/17/2013 12:28 PM PST
Reply Quote
I'm EXTREMELY confused... I came into the thread hoping to show quantitative proof that the OP was wrong and that 10APoC would impact our DPS multiplier, since that's what I measured in 1.0.6 Ghom Tests. In 1.0.5 and 1.0.6 this would have never worked, but have you guys done a Ghom Test in 1.0.7? I just did three MP7 Ghom tests for with either 10APoC or 18APoC and I came to pretty close results (2.74aps, 110-130ms Latency):
*10APoC/111.4kdps/50%CC - 6.57x Multiplier (took 67, 69, 69s, and 71s respectively)
*18APoC/100.5kdps/51%CC - 6.70x Multiplier (took 73, 74, 77, and 79s respectively)
In both cases at full wind up, I didn't run out of AP on the tests. This isn't what I saw in 1.0.6. Can anyone else replicate these results? If it's true, 10APoC doesn't seem to be as outlandish in 1.0.7 as it was in previous patches. I dunno why it's different though and clearly something is also different with the uber Seigebreaker fight... has something actually been changed?


Ok, that really surprises me since it's single target. I doubt I'll have time to test it myself but now I'm super curious. That makes me wonder if the APoC behavior is acting in a strange way like the CM and SA procs. It would be consistant I suppose. The other issue is that you say you did similar test in 1.0.6 and got different results, which is odd, unless the previous APoC behavior used something more like the LoH mechanics while after the patch they changed it to match the CM/SA behavior.
Reply Quote
Yeah... I'm extremely surprised by my results. And these tests are super-quick to do. Back in 1.0.6, I remember trying 16APoC with Crystal Shell and I was occassionally get the out of Arcane Power message on the Ghom Test. I don't remember if I tried 10APoC exactly with SNS, but these results make me question a lot of what I know about APoC. I think I'll see what a Crystal Shell test does.
Reply Quote
I just did one test of Crystal Shell (instead of Diamond Shards) and got:
*10APoC/111.4kdps/50%CC - 6.03x Multiplier (took 75s respectively)
*18APoC/100.5kdps/51%CC - 6.05x Multiplier (took 83s respectively)
Edited by TekkZero#1963 on 2/17/2013 12:37 PM PST
Reply Quote
02/17/2013 12:26 PMPosted by TekkZero
Yeah... I'm extremely surprised by my results. And these tests are super-quick to do. Back in 1.0.6, I remember trying 16APoC with Crystal Shell and I was occassionally get the out of Arcane Power message on the Ghom Test. I don't remember if I tried 10APoC exactly with SNS, but these results make me question a lot of what I know about APoC. I think I'll see what a Crystal Shell test does.


I've Noticed that my Apoc generated from my meteors from 1.06>1.07 has become very Very spikey. I.e. instead of constantly gaining a little apoc bit by bit I get huge bursts. sometimes enough to fill my entire pool. Could it be that the amount of Apoc you get is no longer affected by the skills proc rate? I.e. before I could have sworn that Star pact only gave back 3 apoc if it had a crit now its giving back 30. Could this be why?

Was it always like this?
Reply Quote
don't use ghom as a good guage of how much AP you need. it's so fat, you have more twisters hitting him than other smaller targets.
Edited by Excalibur#1742 on 2/17/2013 12:50 PM PST
Reply Quote
I couldn't resist doing a couple tests myself.

I ran 2 tests with 20 and 10 APoC each. I didn't have a rune on FN, which is why my mutlipliers are a little low.

Average multiplier with 20 APoC = 5.79x
Average multiplier with 10 APoC = 5.51x

Conclusion: They're pretty close, and a lot closer than I expected. I expected something like a 4 or 4.5x multiplier, before reading Tekk's results.

To further satisfy my curiousity I did one more test with WW, crystal shell, and 10 APoC. Our previous expectations, with 2.5 APS giving 33 tics per cast, is that I should use about 87.5 AP per second and have 15 WW up at a time. My expected AP return rate per WW is 33*0.125*0.5/6 = 3.43 or 51.56 AP per second with 15 twisters up, using 50% CC. So I should be using 87.5 AP per second but only gaining 51.56, meaning I should run out of AP continually.

My results: Windup was a bit tricky (partly because I swapped to energy armor for extra defense, so I lost some AP), but I was easily able to sustain my AP and it stayed mostly at max with spikes when I cast WW but my AP filled VERY fast.

Overall conclusion is that APoC seems to be using a different mechanic than previously assumed, at least with respect to WW. My suspicions is that it is using the same mechanics as CM and SA, which we haven't figure out yet, mostly due to lack of time for testing for those of us looking into it. Rough estimates put the mechanics for high APS at about 2.5x the expected proc rates compared to the LoH model. It's also possible that APoC returns just aren't using the spell coefficients but without more results I can't say for sure.

EDIT: After doing a couple quick tests with AO and Blizzard, I'm pretty sure the spell coefficient still applies to APoC. I'm more inclined to believe it's just an issue with the crits for WW and spells like that.
Edited by Loroese#1415 on 2/17/2013 2:04 PM PST
Reply Quote
Thanks Loroese for corroborating my initial results. I feel a bit better. Like you, I had come into this thread expecting to show that the OP would have a much lower DPS multiplier and that AP regeneration at 10APoC would be a clear problem. But that doesn't seem to be true in 1.0.7! Though I'm still unsure what to think of the results. A preliminary quick-run on a single elite (ie I wouldn't consider this definitive), the wind up appears to take a bit longer at 10APoC, but once I'm in the steady-state I'm not seeing AP-deficits. I'm very befuddled right now, as this is not what I saw pre-1.0.7.
Reply Quote
So... mempos and GG non-apoc forces for the win now?

:O
Reply Quote
I think what's going here is people are forgetting how much more geared wizards are now as compared to before in previous patches, at first we needed about 30% cc and that was it, then we needed 40%, then we needed 2.5 aps now it's 2.73 aps and 50% cc, all of this was to achieve permafreeze, no one ever thought about the fact that our cc and aps has been going up so much that maybe we didn't need as much apoc, because we always needed at least 20apoc. I think with as high as our current cc and aps is, it has also effected or arcane regeneration to where we no longer need as much apoc.
Reply Quote
I believe they changed APoC completely in order to fix blizzard's APoC, but it is all theory right now, not enough controlled testing.

Before blizzard had .0125 proc. so if you had 30 APoC you would get .375 AP per crit, the problem is that is always rounded to 0.

So the old method was:
A single crit returns [proc%] x APoC (like LoH)

New method is
A single crit has a [proc%] chance to return your full APoC. (kind of like Critical mass)

The "average" should be about the same, but it doesn't work out that way when you approach 50% crit rate I am thinking.
Edited by Shaggy#1317 on 2/17/2013 3:25 PM PST
Reply Quote
I just tried a bunch of times with arcane orb. I SWEAR I am getting some crits with ZERO AP returned, which kind of confirms my theory
Reply Quote
I believe they changed APoC completely in order to fix blizzard's APoC, but it is all theory right now, not enough controlled testing.

Before blizzard had .0125 proc. so if you had 30 APoC you would get .375 AP per crit, the problem is that is always rounded to 0.

So the old method was:
A single crit returns [proc%] x APoC (like LoH)

New method is
A single crit has a [proc%] chance to return your full APoC. (kind of like Critical mass)

The "average" should be about the same, but it doesn't work out that way when you approach 50% crit rate I am thinking.


I disagree. If you cast blizzard you see a somewhat gradual increase to your AP regen. To me that indicates an AP gain of AP*Coefficient, though it might round it up to 1. Instead, if you apply your idea for the new method, you would get a spike every couple casts corresponding to the very small proc rate, but would mostly see no differences in AP regen.

It's actually pretty easy to test with blizzard. Just stand on Ghom, drain your AP pool, and cast one blizzard. Then time how long it takes to cap your AP. Then repeat a couple times. It should take about the same amount of time each attempt. Otherwise, if one attempt is much faster than the rest, that would confirm your idea.

I mentioned I messed around with AO a bit. I never noticed my AP shooting up really fast, which should occur if your theory is right. Instead all the AP returns were fairly gradual, which again indicates the APoC return is modified by the proc coefficient.

02/17/2013 03:05 PMPosted by slade
I think what's going here is people are forgetting how much more geared wizards are now as compared to before in previous patches, at first we needed about 30% cc and that was it, then we needed 40%, then we needed 2.5 aps now it's 2.73 aps and 50% cc, all of this was to achieve permafreeze, no one ever thought about the fact that our cc and aps has been going up so much that maybe we didn't need as much apoc, because we always needed at least 20apoc. I think with as high as our current cc and aps is, it has also effected or arcane regeneration to where we no longer need as much apoc.


Again, I disagree. We had to gear for high APS and CC for WW to gain AP per cast once the coefficient was reduced. If you see the math I mentioned above, in my test I used some 87 AP per second and should have only gained 50 AP per second, well 60 if you include the base 10 regen per sec, yet my AP was full the entire time after wind up. That means I was regenerating AP faster than I should have been based on our previous assumptions of how APoC worked with WW.

There are a lot of cases pointing to crits not working with WW in the same way that LoH does. These cases all indicate that something is going on that we don't completely understand, and the APoC now seems consistant with those results.
Reply Quote
I wonder if they just changed rounding and that is what we are seeing. Maybe any portion of an AP point is rounded to 1 now. It would certainly account for the findings on twister.

Imagine this:
16 apoc *.125 = 2 ap per crit
10 apoc * .125 = 1.25 ap per crit rounded down to 1.

so 10 apoc was HALF the energy twister AP regen from 16

But now...

16 apoc = 2 ap per crit
10 apoc = 1.25 ap per crit, rounded up to 2

Who knows. I am not sure there is a definitive way to test, and there is no way in inferno that blizzard would actually tell us.

If I am right, there should be no difference between 16 and 10, but a noticable difference between 17 and 10 (or 17 and 16 for that matter)
Edited by Shaggy#1317 on 2/17/2013 5:41 PM PST
Reply Quote
I was very interested after reading this thread and decided to do a few Ghom Runs of my own. Here are my findings:

Ghom HP on MP6: about 36,000,000

Firstly, my 18 APoC Build. DPS: 105,272, Crit Rate: 55%

Trial 1 - 48.41 seconds Ghom Kill time
Trial 2 - 45.67
Trial 3 - 45.55
Trial 4 - 49.19
Average - 47.205
EDPS: 762631.077
Multiplier: x7.244

Secondly, my 9 APoC Build. DPS: 105066, Crit Rate: 54.5%

Trial 1 - 53.45
Trial 2 - 49.89
Trial 3 - 49.95
Trial 4 - 52.59
Average - 51.47
EDPS: 699436.565
Multiplier: x6.657

I tried to keep most stats static, and hopefully the 0.5% crit rate difference does not account for too much. Windup for the second build took longer, but after establishing full windup it was fine. Arcane orb dips to about 30-40% at its lowest point (with rare 10-20%'s) in the low APoC build.

For the loss of about 0.6 of a multiplier, you can POTENTIALLY run lower APoC. Since we have only tested this on ghom (I would think ubers/fast mobs such as shamans and soul rippers would provide more difficulty for the low APoC build) we can't really draw any quick conclusions.

However, if this build works efficiently to some extent, it allows me to access a much greater amount of DPS. I can switch out my storm crow for a AR Andy's and switch out my amulet (Currently Vit/AR/Armor/Attack Speed/Crit Chance) to one a crafted amulet with High int, Armor, Life on Hit, Crit Chance and Crit Damage. This shift would probably yield me about 30k DPS, which greatly offsets the 0.6 multiplier loss. Would be great for budget wizards!
Edited by Clifton#1905 on 2/17/2013 5:51 PM PST
Reply Quote

Please report any Code of Conduct violations, including:

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.

Forums Code of Conduct

Report Post # written by

Reason
Explain (256 characters max)
Submit Cancel

Reported!

[Close]