Over the past two weeks I've been recording my clear times for T6 rifts with my Jade doctor. I've discovered a few things, but most interestingly I have repeatedly found IAS to be a very important contributor to my clear times. Trust me, I was puzzled when I discovered this too; after all, isn’t Jade’s damage output mostly a reflection of damage per hit, which IAS contributes nothing to at all?
So where's the evidence for this claim? Before you say the sample size is too small for each of these comparisons, look at how consistent and large the benefit is. These tests were done with Hwoj Wrap, Warzechian, TotD, and Horrify. Assume the weapon used was Sunkeeper unless otherwise stated.
5:58 avg for 1.57 IAS (Azurewrath); n = 8
7:00 avg for 1.45 IAS, -4.3% crit; n = 4
7:10 avg for 1.46 IAS, -10% CDR, n= 5
7:18 avg for 1.43 IAS, +50% crit (Witching hour), n = 3
7:19 avg for 1.34 IAS; n = 5
7:21 avg for 1.34 IAS, -50% area damage; n = 5
7:29 avg for 1.34 IAS, -10% CDR, n = 5
7:31 avg for 1.34 IAS, String of Ears instead of Hwoj; n = 5
Hereis my profile. I might not have my Sunkeeper up; but for reference it's 2050DPS with 24% elite damage and 700 int. My Azurewrath is max DPS with 9% CDR instead of Int, and 17% Cold.
I started thinking about why IAS might be more important than originally thought. My theory is that the ability to quickly group and CC enemies in a small area seems to be a key factor in clear times. IAS, which shortens animation times, contributes to this in a major way by decreasing the time to set up small patches of high-density mobs. By allowing you to more quickly cast your DOTS and Piranhado, which is probably one of the slowest animation spells, IAS decreases the window of time where mobs can run away from you after your CCs lose their effect. Like CDR, IAS also contributes to your survivability (by decreasing your animation time to let you move out of the way of attacks), as well as your movespeed. The importance of IAS diminishes at higher rift levels, where you can no longer one-shot even white mobs. However, IAS still plays an important role in higher rift levels by letting you proc Furnace more often, as well as making it easier to gain Taeguk or Gogok of Swiftness stacks.
What this means (does not apply for greater rifts):
Azurewrath, not sunkeeper, might be BIS for Jade
Paragon points into IAS is not a bad idea
You should keep IAS on Quetzacoatl unless you have severe mana issues (i'm looking at all you people who use PtV instead of Gruesome Feast)
IAS on Jade gloves is not a bad roll necessarily
The 2.1 Lacuni's (which give IAS as a 5th primary) are a good alternative to Warzechian/Strongarm
Gogok of Swiftness (20% IAS, 20% CDR) is likely going to be a viable gem choice
Witching Hour is better than originally thought since the IAS roll is not useless
Interesting topic. However, I really don't feel like you've documented all the variables thoroughly enough for complete analysis.
I've found that rift density and monster types play a much more significant role in my clear times. For instance, if I get Corpse Raisers and Maggot Broods, I'm more likely to finish faster than if I have Mallet Lords, Morlu Incinerators, Death Maidens, and things that like to run away. Hopefully multiple rifts covered for that somewhat, but even by your outline I'm not sure that you isolated IAS adequately.
What would be interesting is if you had 0% IAS from paragon and did 50 rifts like that and then added a full 50 points for 10% IAS and did another 50 rifts like that. No other item swaps. Just 10% IAS difference.
EDIT
I'd just like to echo a lot of the responses you've received on Reddit as well. There are a ton of variables just on the weapon swap. Azurewrath vs Sun Keeper isn't simply a matter of IAS. There are differing base attack speeds, different stats, etc. It's very possible, as someone suggested, that a rather high chance to freeze helps Jades out because it keeps monsters grouped up better for SH. Who knows?
There is no doubt that 10% IAS > 0% IAS. If you're not sacrificing anything, in the case of Paragon, then surely it's going to help you clear faster. But it's a matter of how much faster you clear. You really should be doing this test with the EXACT SAME gear setup each time and 0% Paragon IAS for many runs and then 10% Paragon IAS for many runs. Demonstrate what value 10% IAS has in terms of total clear times across MANY runs, not just 5-8 per item swap. It will give a significantly-clearer picture one way or the other.
You've simply jumbled so many variables together with gear swaps that it's impossible to have any reasonable conclusion. One of the basic principles of attempting to evaluate a hypothesis is to isolate your variable. Without doing that it adds considerable "noise" to the results of the testing.
Clear times are more dependent on mob types, mob HP, density of mobs, map layout, and Rift Guardian type. IAS doesn't say much there.
If you really want to test IAS vs. no IAS, go run a few Ghom tests with weapons with near identical min-max damage, just one faster than the other. For example, compare kill times between an Utar's Roar and Azurewrath, both with identical min-max and identical cold bonus. Or use two rare weapons, a mace (1.2 APS) vs. a dagger (1.5 APS) with identical min-max.
For solo, maybe you can argue it considering most blue packs will get one-shot even with subpar Jade gear.
But for group games, where most yellow packs can have upward of 2 billion HP, you're going to need to maximize the damage (which is why many run Harrington Waistguard).
And about maximizing total damage output, that's why I suggested doing a single target test. If you're not doing more eDPS over the span of one Soul Harvest against a single target with Jade, how are you going to be doing more damage against a group when area damage is factored in?
In addition, all Azurewrath is good for is spitting out Haunts and Locust Swarms (which you shouldn't need to spam anyway) faster. However, you don't need to spam that much against elites since they only come in packs of 3-5. You're not going to need extra damage against trash. You'll need extra damage against elites to come as close to one-shotting them as often as possible and that's where Sunkeeper comes in.
Bear in mind I talk about most of this while playing in a 4-man group (I act as Jade, teammates are Rimeheart-EP monk, cold-speed barb, shotgun crusader). We've all used Azurewrath before on our Jades during our gearing phases, but Sunkeeper wins out in terms of maximizing the potential to one-shot elite packs.
My Jade doctor for reference. - only change I'm thinking about making is swapping out my Serpent for Thing of the Deep considering my survivability isn't too bad in a group.
as others mentioned, clear times are dependend on other factors.
for jade, ias is not important, because you won't be able to use soul harvest faster (which is, how the build deals its damage). rather its the CDR, thats why you put a diamond in the helm and some also add CDR on shoulders,mojo etc. also the DoT ticks won't tick faster, you are simply able to cast haunt & locust more often, but how does that benefit you if these 3-5 mobs already have been haunted each one? not at all.
for jade build in first place, it's the most important how often you can nuke and this managed by the fact of how often you can 1-shot mobs. thus dealing more damage with single nukes with Sunkeeper (meaning on elites, trash is 1 shot anyway).
I dont think the faster cast rate of haunt from more ias matters much because you arent just spamming it against one target. You try to haunt as many targets as possible and with some ias from paragon/rorg the limiting factor becomes how fast you can click on the individual mobs, not your attack speed.
Well consider this guys, if majority of mobs can be killed with 1 soul harvest whether you're using Sunkeeper or Azurewrath, then perhaps having more ias in this instance will in fact benefit us. If using a slow, high-damaging weapon doesn't save you a round of Dotting->Harvest then casting faster would be more beneficial. Azurewrath also provides extra damage for Haunt which helps close the gap in damage range between it and Sunkeeper as well.
Well consider this guys, if majority of mobs can be killed with 1 soul harvest whether you're using Sunkeeper or Azurewrath, then perhaps having more ias in this instance will in fact benefit us. If using a slow, high-damaging weapon doesn't save you a round of Dotting->Harvest then casting faster would be more beneficial. Azurewrath also provides extra damage for Haunt which helps close the gap in damage range between it and Sunkeeper as well.
There are actual ways to test this, though.
The OP didn't test anything in a scientific fashion so his conclusions are clouded with heaps of ambiguity.
Just switching Sun Keeper for Azurewrath involves a lot of variables and even if you kill faster with Azurewrath it's logically wrong to conclude it's because of IAS (or APS). As someone from Reddit suggested, maybe it's just the high chance to freeze preventing monsters from escaping Soul Harvest range as much. As others pointed out only doing 5-8 rifts per setup isn't really enough to compensate for the hugely-random nature of the rifts either.
The simplest way to test this is to keep your gear static. Do not change any items. Remove all paragon points in int and all paragon points from the Offense tab. Then test each stat individually (int, CDR, IAS, CHC, CHD) with 50 points. Do this across roughly 50 rifts each, which should account for most of the randomness of rift tilesets, monster selection, and monster density. Then come back and give us information on each clear time (this is important so we can see outliers), as well as average clear time for the following:
Control group (0 points in int, CDR, IAS, CHC, CHD)
50 points in int (250 int), 0 points in everything else
50 points in CDR (10% CDR), 0 points in everything else
50 points in IAS (10% IAS), 0 points in everything else
50 points in CHC (5% CHC), 0 points in everything else
50 points in CHD (50% CHD), 0 points in everything else
With that information we could actually begin to evaluate how worthwhile each stat is. Now, much like WoW simming, some of the stats are going to be dependent on the player's current gear setup, so it would be helpful if someone who were to do this were using fairly-representative items. For instance, if a player had only 5% CHC on their gear, the value of CHC would be disproportionately higher than someone who had a "normal" amount of each stat.
Even with that shortcoming, that method will more closely isolate the variable (the stat which you're increasing) as opposed to just randomly swapping in items and doing a handful of rifts. There's simply too many variables which aren't controlled at all in the OPs data for it to be considered anything other than a hypothesis. Lump on top of that how most of the changes are statistically-insignificant given the outrageous margin of error from improper testing (his "base case" is 7:19, and everything except one scenario fell between +12 seconds and -19 seconds), and it's a wonder why this is even being discussed.
There are ways to gather information. What the OP did is not the correct way. When you don't gather information scientifically it really hurts your ability to draw conclusions from that data.
Ghom test would be 100% pointless for Jade clear since a lot of Jade's clear time is dependent on CD resets from killing mobs. OP's entire theory is based on faster trash cleanup through faster cast times, not higher damage output. Sunkeeper will poop on Azure in a Ghom test, and for at least OP, he is seeing faster clears with Azure.
I think op may be onto something in the context of solo jade runs - I do notice its hard to dot everything, then rush in and harvest, within the time frame of a piranado (before everything starts wandering off.) I sometimes find the haunt cast so slow that it seems like I unintentionally break it off mid cast. The Azure freeze + faster cast times may alleviate some of this problem, and in the context of soloing, end up providing a faster clear. (I personally suspect the freeze may be more of a factor in OPs experience than the asi, but who knows.)
Agree though that for optimizing damage in a group setting where you are not solely dependent on piranahdo for clustering / cc, Sunkeeper is hands down superior, as a slower, heavier weapon, with more + elite than Azure has + cold.
All that said, I'll be sticking with my Keeper for now, both solo and in group.
Also Jaetch you may want to consider the perma Fear build since you're not using your dogs for move speed. You can switch out the serpent for totd if you have one, the extra pickup radius is beautiful, and it's my understanding that 100% uptime on Horrify - Frightening Aspect provides more ehp than serpent + leaching dogs.
Havn't tried it myself as I'm loathe to give up the passive move speed from Fierce Loyalty, but I would love to be able to pick back up spirit vessel.
So where's the evidence for this claim? Before you say the sample size is too small for each of these comparisons, look at how consistent and large the benefit is. These tests were done with Hwoj Wrap, Warzechian, TotD, and Horrify. Assume the weapon used was Sunkeeper unless otherwise stated.
I started thinking about why IAS might be more important than originally thought. My theory is that the ability to quickly group and CC enemies in a small area seems to be a key factor in clear times. IAS, which shortens animation times, contributes to this in a major way by decreasing the time to set up small patches of high-density mobs. By allowing you to more quickly cast your DOTS and Piranhado, which is probably one of the slowest animation spells, IAS decreases the window of time where mobs can run away from you after your CCs lose their effect. Like CDR, IAS also contributes to your survivability (by decreasing your animation time to let you move out of the way of attacks), as well as your movespeed. The importance of IAS diminishes at higher rift levels, where you can no longer one-shot even white mobs. However, IAS still plays an important role in higher rift levels by letting you proc Furnace more often, as well as making it easier to gain Taeguk or Gogok of Swiftness stacks.
What this means (does not apply for greater rifts):
I've found that rift density and monster types play a much more significant role in my clear times. For instance, if I get Corpse Raisers and Maggot Broods, I'm more likely to finish faster than if I have Mallet Lords, Morlu Incinerators, Death Maidens, and things that like to run away. Hopefully multiple rifts covered for that somewhat, but even by your outline I'm not sure that you isolated IAS adequately.
What would be interesting is if you had 0% IAS from paragon and did 50 rifts like that and then added a full 50 points for 10% IAS and did another 50 rifts like that. No other item swaps. Just 10% IAS difference.
EDIT
I'd just like to echo a lot of the responses you've received on Reddit as well. There are a ton of variables just on the weapon swap. Azurewrath vs Sun Keeper isn't simply a matter of IAS. There are differing base attack speeds, different stats, etc. It's very possible, as someone suggested, that a rather high chance to freeze helps Jades out because it keeps monsters grouped up better for SH. Who knows?
There is no doubt that 10% IAS > 0% IAS. If you're not sacrificing anything, in the case of Paragon, then surely it's going to help you clear faster. But it's a matter of how much faster you clear. You really should be doing this test with the EXACT SAME gear setup each time and 0% Paragon IAS for many runs and then 10% Paragon IAS for many runs. Demonstrate what value 10% IAS has in terms of total clear times across MANY runs, not just 5-8 per item swap. It will give a significantly-clearer picture one way or the other.
You've simply jumbled so many variables together with gear swaps that it's impossible to have any reasonable conclusion. One of the basic principles of attempting to evaluate a hypothesis is to isolate your variable. Without doing that it adds considerable "noise" to the results of the testing.
Clear times are more dependent on mob types, mob HP, density of mobs, map layout, and Rift Guardian type. IAS doesn't say much there.
If you really want to test IAS vs. no IAS, go run a few Ghom tests with weapons with near identical min-max damage, just one faster than the other. For example, compare kill times between an Utar's Roar and Azurewrath, both with identical min-max and identical cold bonus. Or use two rare weapons, a mace (1.2 APS) vs. a dagger (1.5 APS) with identical min-max.
Armory | YouTube | Twitter | Clan Site
For solo, maybe you can argue it considering most blue packs will get one-shot even with subpar Jade gear.
But for group games, where most yellow packs can have upward of 2 billion HP, you're going to need to maximize the damage (which is why many run Harrington Waistguard).
And about maximizing total damage output, that's why I suggested doing a single target test. If you're not doing more eDPS over the span of one Soul Harvest against a single target with Jade, how are you going to be doing more damage against a group when area damage is factored in?
In addition, all Azurewrath is good for is spitting out Haunts and Locust Swarms (which you shouldn't need to spam anyway) faster. However, you don't need to spam that much against elites since they only come in packs of 3-5. You're not going to need extra damage against trash. You'll need extra damage against elites to come as close to one-shotting them as often as possible and that's where Sunkeeper comes in.
Bear in mind I talk about most of this while playing in a 4-man group (I act as Jade, teammates are Rimeheart-EP monk, cold-speed barb, shotgun crusader). We've all used Azurewrath before on our Jades during our gearing phases, but Sunkeeper wins out in terms of maximizing the potential to one-shot elite packs.
My Jade doctor for reference. - only change I'm thinking about making is swapping out my Serpent for Thing of the Deep considering my survivability isn't too bad in a group.
Armory | YouTube | Twitter | Clan Site
for jade, ias is not important, because you won't be able to use soul harvest faster (which is, how the build deals its damage). rather its the CDR, thats why you put a diamond in the helm and some also add CDR on shoulders,mojo etc. also the DoT ticks won't tick faster, you are simply able to cast haunt & locust more often, but how does that benefit you if these 3-5 mobs already have been haunted each one? not at all.
for jade build in first place, it's the most important how often you can nuke and this managed by the fact of how often you can 1-shot mobs. thus dealing more damage with single nukes with Sunkeeper (meaning on elites, trash is 1 shot anyway).
Sunkeeper wins.
The OP didn't test anything in a scientific fashion so his conclusions are clouded with heaps of ambiguity.
Just switching Sun Keeper for Azurewrath involves a lot of variables and even if you kill faster with Azurewrath it's logically wrong to conclude it's because of IAS (or APS). As someone from Reddit suggested, maybe it's just the high chance to freeze preventing monsters from escaping Soul Harvest range as much. As others pointed out only doing 5-8 rifts per setup isn't really enough to compensate for the hugely-random nature of the rifts either.
The simplest way to test this is to keep your gear static. Do not change any items. Remove all paragon points in int and all paragon points from the Offense tab. Then test each stat individually (int, CDR, IAS, CHC, CHD) with 50 points. Do this across roughly 50 rifts each, which should account for most of the randomness of rift tilesets, monster selection, and monster density. Then come back and give us information on each clear time (this is important so we can see outliers), as well as average clear time for the following:
Control group (0 points in int, CDR, IAS, CHC, CHD)
50 points in int (250 int), 0 points in everything else
50 points in CDR (10% CDR), 0 points in everything else
50 points in IAS (10% IAS), 0 points in everything else
50 points in CHC (5% CHC), 0 points in everything else
50 points in CHD (50% CHD), 0 points in everything else
With that information we could actually begin to evaluate how worthwhile each stat is. Now, much like WoW simming, some of the stats are going to be dependent on the player's current gear setup, so it would be helpful if someone who were to do this were using fairly-representative items. For instance, if a player had only 5% CHC on their gear, the value of CHC would be disproportionately higher than someone who had a "normal" amount of each stat.
Even with that shortcoming, that method will more closely isolate the variable (the stat which you're increasing) as opposed to just randomly swapping in items and doing a handful of rifts. There's simply too many variables which aren't controlled at all in the OPs data for it to be considered anything other than a hypothesis. Lump on top of that how most of the changes are statistically-insignificant given the outrageous margin of error from improper testing (his "base case" is 7:19, and everything except one scenario fell between +12 seconds and -19 seconds), and it's a wonder why this is even being discussed.
There are ways to gather information. What the OP did is not the correct way. When you don't gather information scientifically it really hurts your ability to draw conclusions from that data.
I think op may be onto something in the context of solo jade runs - I do notice its hard to dot everything, then rush in and harvest, within the time frame of a piranado (before everything starts wandering off.) I sometimes find the haunt cast so slow that it seems like I unintentionally break it off mid cast. The Azure freeze + faster cast times may alleviate some of this problem, and in the context of soloing, end up providing a faster clear. (I personally suspect the freeze may be more of a factor in OPs experience than the asi, but who knows.)
Agree though that for optimizing damage in a group setting where you are not solely dependent on piranahdo for clustering / cc, Sunkeeper is hands down superior, as a slower, heavier weapon, with more + elite than Azure has + cold.
All that said, I'll be sticking with my Keeper for now, both solo and in group.
Havn't tried it myself as I'm loathe to give up the passive move speed from Fierce Loyalty, but I would love to be able to pick back up spirit vessel.