The following 3.1 change removes all theoretical advantage of DW EL:
Ancestral Awakening: This talent now accounts for your ineffective healing, rather than effective.
The below post is preserved for historical reasons.
Dual Wield Chain Heal and You:
With the advent of WotLK, resto shamans gained a fun new toy in Riptide. Although excellent for 5 man heroics and 10 man raids, the oh, sh#@! covering capabilities of Riptide are usually not needed in a 25 man heroic raid setting.
The viability of EL DW is still not determined; however, it is obvious that it is more valuable in a raid-damage heavy 25 man setting over a fight like Patchwerk or 10 man raids.
Dual Wielding does remove the possibility of grabbing Earth Shield and Riptide.
Dual Wielding offers up to 124 (187 with enchant) additional SP over the highest SP offhand currently available to shamans - [Voice of Reason], and increases the chance for an Earthliving proc on CH targets - up to 43.75% with the [Glyph of Earthliving Weapon]. Because of the nature of the build, DW EL emphasizes heavy throughput more than mana efficiency.
Because of the spammy nature of this spec, 2 Pc T6 may be viable for the mana cost reduction in chain heal. As stated in the TTT Article, "[s]haman who want to reduce the cost of Chain Heal with two pieces of Tier 6 will want to use [Skyshatter Belt] and [Skyshatter Bracers]."
Earthliving Weapon:
It is theoretically possible to consistently keep EL up on 7 targets with 0 haste, using a DW build.
It is theoretically possible to consistently keep EL up on 4 targets with 0 haste, using a standard build.
Missing information (to be added ASAP):
Proposed HEP values for a chain heal spec.
How EL affects this spec.
...
Corrections and Additions:
2/16:
Comprehensive rewrite (in progress).
2/04:
Updated information.
1/14:
Added [Blade of the Unrequited] as a possible offhand. Thanks to Ranyon for pointing this one out.
1/13:
Added EL Proc Rate (theoretical maximum) graph.
1/11:
Thanks to MatsT for correcting the proc rate of dual earthliving at 43.75% instead of 50% as assumed. Theoretical maximum targets with DW EL is 7 targets, not 10.
What is the equations or modifiers that are added cause a 1600 hps increase at 300 haste. Does that number include the EL procs? The reason I am interested in this is that even including the 300 hps increase of EL, that still is a 1300 hps increase from 184 spell power (and there is a 4% crit loss).
That seems abnormally large for a ~200 spell power increase.
Or am I missing something?
Can you tell me what music is? It's completely intangible. It's something you can't grasp. You see art, watch people dancing, but you have to give part of your life to hear music. It grips you, gets into your soul, the most sublime of all the arts. Why should a minor chord sound sad, a major chord sound happy? There's no logic to it at all. ---George Martin
It's a side effect of the way EL weapon procs on targets.
[Total HPS] = [Base HPS] + [EL Single Target HPS]*[EL Proc Rate]*([12 Seconds]/[Spell Cast Time]) is a close approximation, I will update the original post with the verified correct formula once I manage to track down one odd little behavior.
Here's a chart if you're interested:
R = EL Proc Rate, H = Heal Cast Time, E = EL healing amount
Time:
0
H
2H
3H
4H
5H
EL Total HPS:
R*E
2*R*E
3*R*E
4*R*E
5*R*E
6*R*E
So for a non-hasted HW, at 2100 SP, 20% proc rate:
Time:
0s
2.5s
5s
7.5s
10s
12.5s
EL Total HPS:
38.3
76.5
114.8
153.1
191.4
191.4
and the hasted version:
Time:
0s
2.1s
4.2s
6.8s
8.9s
11.1s
13.2s
EL Total HPS:
38.3
76.5
114.8
153.1
191.4
229.7
229.7
Since chain heal with DW has a 200% (50%*4 targets) chance to proc on each CH, it scales ridiculously well with haste. (compared with the (20% * 4 targets)=80% chance of the standard build). However, there is a caveat: it's nearly impossible to ensure your EL procs always occur on different targets.
The weird behavior I've mentioned is how EL falls off after 12 seconds. I'm not quite sure how to model that algebraically.
Since chain heal with DW has a 200% (50%*4 targets) chance to proc on each CH, it scales ridiculously well with haste. (compared with the (20% * 4 targets)=80% chance of the standard build). However, there is a caveat: it's nearly impossible to ensure your EL procs always occur on different targets.
I am having a hard time trusting stats posted when statements like this are mixed in.
[Dezzimal] The first iteration of Algalon looted your raid after a kill and posted screenshots on his website.
I am having a hard time trusting stats posted when statements like this are mixed in.
Chain heal DW - 50% chance to proc off each target. 4 total targets. .50*4 = 2 total targets up per cast.
Chain Heal RT - 20% chance to proc off each target. 4 total targets. .20*4 = .8 total targets up per cast.
[edit] Misunderstood. Unfortunately, until there are some more WWS parses available, there's no readily available way to state your EL weapon effect will have X% overall efficiency. So, for the purposes of this document at this time, I've listed the theoretical maximum.
If you have 4 separate 20% chances, the overall chance to proc is calculated as follows:
Chance to proc on any target = 1 - (chance to proc on no target) = 1 - (0.8*0.8*0.8*0.8) = 1 - 0.4 = 0.6
Same for 50% proc chance:
Chance to proc on any target = 1 - (chance to proc on no target) = 1 - (0.5*0.5*0.5*0.5) = 1 - 0.06 = 0.94
So with DW you have a 94% chance that EL procs at all on a given chain heal. This includes all multi-procs.
/edit: I guess this was just a phrasing problem. The expected number of procs is indeed 2, but that doesn't mean EL has a 200% proc chance.
If considered in an infinite series, it becomes the .5*4 and .2*4 chances to proc, if I am correct. For the purposes of the HPS calculations, infinite spam-ability is assumed.
I've only done light testing so far, and have found EL to do approximately 10% of my total healing, depending on the encounter. WWS parses incoming soon.
I have not yet tested the setup without the glyph.
edit: Seems I've misplaced all my combat logs that had this particular setup; however, I wasn't saved to 10M VoA. not exactly the most compelling of tests, but it's a start until I can either find my old logs or create new ones from next week's resets.
I've only done light testing so far, and have found EL to do approximately 10% of my total healing, depending on the encounter. WWS parses incoming soon.
I have not yet tested the setup without the glyph.
edit: Seems I've misplaced all my combat logs that had this particular setup; however, I wasn't saved to 10M VoA. not exactly the most compelling of tests, but it's a start until I can either find my old logs or create new ones from next week's resets.
That shows EL doing 4%. That's a long way from 10%.
A very quick blast through my 25 man raids using the EL from myself (50/50 CH/LHW) and another healing shaman (100% CH) shows that from 24 data points there was an average of 6.35% done from ELW.
There is only 1 fight in the data sets with numbers ~=10% and that's Sapph (due to the constant raid damage). Most of the numbers are in the range 3-6% (just eyeballing).
Our numbers are likely to be higher than average, because of a couple of reasons.
1. We run healer light (5-6 healers in our 25 mans)
2. That dataset contains our Heroic Undermanning (5 healers) so there are more people taking raid damage for longer without being topped up.
3. We have traditionally been more melee heavy, which makes CH rather good as it's hitting much closer to the higher end of the targets.
Pewsey has heard about tact and discretion, but tends to regard them much as children view vegetables.
There are only two kinds of MMOs: the ones people complain about and the ones nobody plays. (inspired by Bjarne Stroustrup)
I was under the impression that two imbues with earthliving didn't stack additively, but just gave you two chances to procc it. This would mean a 43.75% chance and not 50, thought i'm not sure about this. The important thing however, is that assuming 0% overhealing on earthliving is so far from reality that it makes the entire calculation irrelevant.
Isn't it foolish to assume 0 haste when you're always going to have haste from the dual-wielded weapon? Well bigger point is you mentioned +184 sp, but did you also take in to account the extra haste/crit?
I also imagine that any WWS pre-3.08 might not be as accurate of the situation post-3.08. Have to take into account CoH will snipe a lot of HoTs... I suspect Earthliving Weapon will do better performance with the CoH cooldown.
Last edited by Starfire : 01/12/09 at 12:31 AM.
Originally Posted by arison
Everyone should start from the same place and rise based on their abilities, desires, and schedule. No one plays MMOs to *be* powerful, they play MMOs to *become* powerful. It's the journey, stupid. The rarer loot is, the more cherished it is when you get it, but only so long as there is a reasonable expectation to get it. The rarer loot is, the better it feels when you kill a boss or when $AWESOME_TRINKET drops.
That's actually a rather good point. I'm going to go through combat parses to confirm the proc rate on EL weapon first, though, before I update the tables again. I also want to get more real-world values so I can compile an expected healing values table in addition to the theoretical maximums table.
I'm starting to suspect that the value of the EL procs scales better with haste than with raw SP, but more data is needed before any conclusion is drawn.
[edit]:
Working on some more data here. It appears that over the course of a 5 minute fight, (VoA 25 man linked earlier), I had procced EL 89 times, 30 of which were refreshes on a target with EL already present.
Chain heal hit 222 times, making the overall proc rate of EL 40.2% for this fight. (std error ~ 3.3%). It does appear that procrate is 43.75% instead of an additive 50% as assumed earlier. I will continue to test this with additional parses that are made available - get me your parses!
For comparison, Ancestral Fortitude procced 48 times, resulting in a 21.6% proc rate (std error ~2.7%). My current crit is 22.8% with this setup, so it appears that the numbers are spot on, which is somewhat dissapointing.
At the moment the theory seems to assume that
1)Each new proc of EL lands on a new person & thus runs the full course of the heal
2)There is an large enough number of people to choose from that it will never need to "overwrite" the buff
While "perfect" surroundings is alright for generating a basic initial assumption, I think for these results to be of practical use a few more considerations need to be made. This distinction is important, if we take for example the statement that "It is theoretically possible to consistently keep EL up on 10 targets with 0 haste, using the dual wield build and EL glyph." While this may be true if your healing the perfect number of people in a perfect distribution, it looses all validity if 90% of your healing is focused on 5 people.
If 90% of healing is focused on 5 people then the maximum number of people at that time getting the EL buff is slightly more than 5 and 5 is pretty darn close to the theoretical maximum of 4 people that a standard build offers.
These styles of numbers are very practical due to the "focused" nature of AOE damage. Yes global AOE damage does occur, but it is the less common situation.
I don't know how you could model this but I think that stating the difference is important. A purely theoretical study is interesting, but it really needs to have a practical element & to achieve this it must be tested in a realistic numbers which means refreshing hots etc.
Well bigger point is you mentioned +184 sp, but did you also take in to account the extra haste/crit?
This was neglected because the shield that would otherwise be taking that spot has similar (though slightly different) distribution of stats. The stats offered by an offhand are of no gain except being able to put an EL enchant on them & spell power enchant.
Can you tell me what music is? It's completely intangible. It's something you can't grasp. You see art, watch people dancing, but you have to give part of your life to hear music. It grips you, gets into your soul, the most sublime of all the arts. Why should a minor chord sound sad, a major chord sound happy? There's no logic to it at all. ---George Martin
With the current numbers I have from the (admittedly limited) parses on hand, it appears that we can expect a ~33% refresh rate on EL weapon. I'll soon add theoretical maximums for 5 melee and 10 melee raids, once I am able to get my brain screwed on straight again.
Nicadimos, do you have any recent WWS parses we could peruse for more data?
Another issue with the theory is that 4 people have to be in range and have to have taken damage. This makes pre-emptive hotting not possible because a chain heal won't jump to someone that has 100% hp. I'd be curious if there was a way to track what % of your chain heals actually hit 4 people and so on down for 3 2 and 1. You could then calculate at what point it would be more effective to do dual wielding vs standard resto build.
With things like dual speccing in the future, this would be an interesting option to switch to for fights like Sapph that any additional help to chain heal will be welcomed.
A problem with using HOTs to mitigate raid damage in the first place is the massive overhealing that results; people who are raid healing are typically topped off by direct heals before the HOT can run its course and it's wasted. You don't have Druids using Rejuv to heal raid damage because HOTs just aren't ideal for that situation in practice; healers see raid members below max HP and heal him. They don't look for a HOT and wait for it to run its course. Many healers won't even be able to see the HOTs on their targets in a raid if those HOTs are from another class. So sacrificing a ton of healing power and versatility for a narrow niche spec basically for the sole purpose of getting more Earthliving Weapon procs seems silly to me. You can probably find specific fights/raid makeups where it outperforms the alternatives but that doesn't make it a great idea.
It's very possible that we're looking at this the wrong way, and the only real advantage of the dual wield spec is the SP bonus that can be granted to Chain Heal. If this is the case, the advantage of dual wielding will diminish with gear; if not, dual wield will gain viability with increases to your haste.
Another thing I realized is your spec. You're not taking the talent that increases the chance to proc Earthliving and increases your critical strike. If you put 2 points there instead and had 1 point to make 5% of your int go to spell power, I think you'd be pushing out more healing. Someone better with numbers than I could do the comparison better than I, but considering 1000 int, that's a loss of 100 spell power and a gain of 4% crit and 100% earthliving proc on someone below 35%. Seems like an easy choice for me, but I'm a crit fiend.
This is the build I'd go with if I were to try this:
It's got some enhancement changes too to focus on improving your totems across the board which could save some points for other shaman / DKs in the raid.
I find it somewhat ironic that one of the cornerstones of a DW Resto build is "more spellpower", yet you're advocating dropping one talent point that is at least a third of the spellpower you'd be gaining.
4% crit is not worth the loss of 100SP for this build; you can actually loose up to ~9% crit if I remember correctly and still have equal healing with CH without counting EL procs just from the spellpower bonus.
All of the spellpower weapons in the game are Main-hand, not one-hand. The item everyone mentions, with its combination of haste and crit, is the closest we can get.
Because you need an off-hand or one-hand weapon; not a main hand weapon.
Originally Posted by arison
Everyone should start from the same place and rise based on their abilities, desires, and schedule. No one plays MMOs to *be* powerful, they play MMOs to *become* powerful. It's the journey, stupid. The rarer loot is, the more cherished it is when you get it, but only so long as there is a reasonable expectation to get it. The rarer loot is, the better it feels when you kill a boss or when $AWESOME_TRINKET drops.
01/10/09, 3:22 AM
Similar Threads
