Background: Between the announcement of Raids and the unveiling of the Revenant/Herald class, some of us hit upon the notion of perhaps using it to provide healing in raids. Initial testing quickly quashed this, however, as Druid proved to be superior at healing in every way, while Herald was more useful focusing strictly on DPS and providing passive support.
(This is going to be wordy, so if you don’t like reading things, you can just not read it, without being a diva.)
Observations: The developers seem to desire that no profession provide very good healing and very good DPS at the same time.
- Ranger always had poor sustained DPS, so coming into the expansion, they decided to make its elite spec capable of healing while still putting out its maximum (poor) damage.
- This idea was explicitly stated in the patch notes, re: Elementalist changes.
- Druid healing scales poorly with the Healing Power stat, encouraging Druids to heal in full damage gear. Revenant healing, on the other hand, scales very well with Healing Power, suggesting that perhaps one could sacrifice DPS for healing.
Experiment: To determine if it is possible that Revenant output raid healing while providing damage worthy of a raid spot (and, if so, where the cutoff is).
Materials: I plotted this build as a starting point. Note that I was not able to cobble together precisely this build. I used Exotic armor, and an Exotic offhand (I didn’t have any ascended weapon that a rev could use and that wouldn’t throw off the experiment), costing me some stats.
It relies on maximizing healing output. Although I did not test it myself, I have read that healing output bonus percentages stack additively. Thus, the healing bonus is:
10% (Rune) + 10% (Sigil) + 10% (Food) + 20% (Tranquil Balance) + 20% (Invoking Harmony) + ~12% (Selfless Amplification)* = 82% bonus healing
With these numbers, the ability Natural Harmony should output 4,139 healing to each target each cast. You can do 3 of these in a very short time.
(* The Warrior minor Tactics trait Inspiring Presence gives 10 Healing per Might stack from the Warrior, adding 250 Healing, assuming you get all your Might from the PS. This is why I took the Protection trait in the Herald line.)
I then experimented with adding different elements from the standard raid build, to determine what this setup would cost in terms of damage output.
All tests were done using 5 boons (25 Might, Fury, Regen, Protection, and Quickness), as well as Warrior buffs (Strength, Discipline, EA). The target was a 4 million HP golem with 25 stacks of Vuln. For initial testing I did not use Ranger buffs, as the idea was to see what would happen with the Rev in the sole healing role.
Throughout this, keep in mind that although I had been thinking of this for a while, I did not get around to doing it until last night when I got in between 2 and 3am. While I was sober, I cannot promise my play was at its sharpest. Also, I used the benchmarks posted by the fine folks at [qT] for comparisons of expected DPS.
Control: First (actually, it was the last thing I did), I decided to establish a baseline to which to compare damage loss. I used (almost) the proper build, with Strength runes instead of Scholar (it’s from my PS; my Scholar set is only 4/6 asc — look, I don’t pve much), and only a 5%/3% stone (it was all I had in my inventory and I was kinda winding down). I should also note that I have logged many hours on Revenant since the pre-release Betas, roughly zero of which have been in group PvE content.
In any case, my “standard” turned out to be: 18,850 dps
This is about 20% lower than the benchmark posted by the top level folks, so figure that between my imperfect gearing and my sloppy play, we lost that much damage. We will keep this figure in mind to suggest adjustments to other data.
The Tests: I had initially planned to run 10 samples of each setup, but after the first few proved so awful, I let it go at one. Each experiment will be titled by the elements used to fill in the blank parts in the build posted. After running the first tests, I noticed that the [qT] runs had used Alacrity, I guess on the assumption that being in party with the Chrono, you’ll be a priority for it, and also because Alacrity does very little for a Rev. For all tests with Invocation, I pushed Energy below 50 during Ventari phases, to simulate healing (thus no Equilibrium damage from it). In all tests, I swapped Legends on cd.
Devastation/Assassin’s Earrings/Glint: 9,746 dps
- same, with Alacrity: 9,952
Devastation/Assassin’s Earrings/Jalis: 10,052
– same, with Alacrity: 10,160
Invocation/Berserker’s Earrings/Glint: 9,085
Invocation/Berserker’s Earrings/Jalis: 8,817
While the numbers were not expected, the relative performance was. I figured that giving up all the bonuses in Devastation would be worse than losing Equilibrium every other swap and the 20% crit. Also, the low Invoc/Jalis is probably a result of variance more than poor performance. Over multiple tests it would likely pull ahead. That said, Glint, like Devastation, is higher raid dps, since you’re probably the only good source of Fury.
I then went ahead and turned on Ranger buffs: Grace of the Land x2 and Spotter (no Spirit), figuring maybe you’re both healing.
Devastation/Assassin’s/Jalis w/Druid and Alacrity: 11,296
– same, second test: 11,314
At this point, I started looking to shave healing buffs to see what could pull ahead. First, I dropped the Salvation line entirely. You lose the larger part of your bonus healing with this, so, for example, Natural Harmony will now output 2,978 per target. You are almost certainly not providing the main healing. I pushed Energy below 50 on each Ventari phase to simulate the healing (so that Equilibrium damage would only proc after Glint phases).
Devastation & Invocation/Berserker’s/Glint w/Ranger buffs: 13,488
This is still pretty weak. Even if we assume that a top level player could manage 25% more damage, this would only be 16,860 dps — well below that of a Druid in the same role.
While I understand that, given the absence of the gear treadmill, raids must be tuned in such a fashion that suboptimal execution can still succeed, you don’t set out planning to bring a suboptimal performer with you.
Of note, in wild desperation, I did the unthinkable and dropped the Herald trait line altogether. My food was running out, so I only used a 1 million HP golem for this. I even threw on Jalis for the biggest numbers.
Devastation, Invocation, Salvation/Berserker’s/Jalis w/Druid buffs: 12,079
And at that point, you’re giving nothing to the raid but the healing. I stopped the experiment there. A second round of testing is possible, but one could probably figure out the damage contribution of each element and do the math without spending as much time on it.
Conclusions: Not much. While I am a passable Rev in competitive modes, I probably made a lot of mistakes in the pve rotation. Still, even being very generous with that consideration, I could only see making a meager concession toward healing — perhaps nothing more than slotting Ventari over Jalis and eating a Rice Ball — without wrecking your damage contribution, at which point, you’re probably not capable of solo-healing, at least as far as I estimate. Some raid groups may find the occasional tablet helpful, but I wouldn’t waste gold gearing for it.
Comments: I’m not really sure what ANet’s intentions were with the whole Ventari aspect. The traitline has some nice bits, but nothing for which you’d give up something better. You can technically make a Ventari stance Rev work in sPvP, but you have to have the right team around you, and the right opponents against you. You can get a bit of mileage out of it in a coordinated roaming group, for whatever that’s worth, but only foolish enemies won’t just focus you out of it. In any case, the tablet mechanic is so clunky that it’s often as much Ventari’s millstone as it is a boon. So it’s something you can pull out in PvE as a safety net to assist imperfect play, but not part of any ideal setup.
Maybe I’m wrong. There could be people using it very well, and I just haven’t encountered them. Any questions/thoughts/plans to pursue the experiment?
(edited by Ocosh.5843)