== Basics of Healing Aggro ==
Healing aggro is based on HP healed, and appears to be linear. With the Subtlety talent, healing generates 50-60% of the aggro per HP that damage dealing generates.
Since aggro is generated based on HP healed, overhealing has no effect on aggro.
Conventional wisdom is that healing aggro applies to all mobs in a fight. This is incorrect. Aggro for a heal is only generated on mobs that are “aware” of the player being healed. We’re not 100% sure what constitutes awareness, but we know that mobs are aware of any player they have targeted and conversely that being in the group or being near the mob does not create awareness. The most reasonable guess is that the mob’s awareness list is the same as it’s aggro list.
This means:
- as the puller *probably* has aggro on all mobs pulled (links may be different), healing the puller generates aggro on all those mobs
- healing a damage dealer who is hitting one mob only generates aggro on that one mob
Aggro (generated by damage or healing) does not decay over time. (tested up to 3 minutes)
== Aggro by Healing Spell ==
Greater Heal, Flash Heal and Prayer of Healing all generate the same amount of aggro per HP healed, subject to the application rules above. This means that using a Greater Heal or a Flash Heal to heal a player for the same amount of HP generates exactly the same amount of aggro. Using Prayer of Healing to effectively heal several players generates about the same aggro as Flash Healing each of them individually.
PW:S generates roughly half the aggro of a heal for the amount of damage it absorbs.
Since Renew heals over time, the aggro is generated on each pulse, not at cast time. We did not run a test on Renew for aggro numbers but assume it follows the other healing spells.
== Mana Usage ==
Fully spec’d out Renew is our 2nd most mana efficient healing spell, behind only fully spec’d out Greater Heal.
(related) For other people doing analysis, note that Thottbot numbers include some talents and not others. They will screw up your analysis. You have to either use allakhazam or take the time to verify each figure.
== More About PW:S ==
PW:S takes AC into account.
Warriors generate a very small amount of rage from taking hits. Switching to defensive stance does not *appear* to significantly change this. With PW:S on them, they still generate rage by taking hits. It may be reduced, but not apparently (significantly) so.
== Fade ==
A simple model for Fade is that it subtracts off an amount of aggro equal to around two flash heals (~1000 HP for a 41 priest w/ Subtlety) and then returns it after 10 seconds is up.
Note that a single Greater Heal can generate more aggro than Fade can get rid of. If you are near even aggro with your tank and use it, Fade will not save you from being attacked.
Healing done before and during the Fade have the same impact.
After the fade expires, the amount of aggro is the same as if you never Faded.
Fade appears to have several unintuitive behaviors which belie the simple model posed above:
- We have seen several situations where Fade expired and aggro did not return to us until we acted, at which point it behaved as expected. Other people have attempted to replicate this behvaior and not seen it. Chances are that this is a bug, something you might want to play with but not depend on.
- When Fade fails to remove a mob from you, it does not make that mob easier to remove. It’s all or nothing.
- It is possible that when Fade fails it actually makes it harder to remove the mob for the duration of the Fade.
== Holy Nova ==
Holy Nova has two separate effects, an AoE damage component and a threat reduction component. The AoE damage is 100 HP per mob for Rank 1 with a radius of 10 yards. The threat reduction operates the same way as Fade, except with a 5 second cooldown and less than half the aggro reduction.
We concluded that Rank 1 Holy Nova reduces around 250 damage aggro. Since it does around 100 damage, it only effectively reduces 150. In healing terms, for a Subtlety spec’d Priest, that’s less than 300 HP healed.
Surprisingly, the threat reduction effect is not dependent on doing damage. It either has a larger radius than the blast, or simply impacts all the mobs who are aware of the caster (ala Fade). Perversely, this means that as an aggro management tool you are better off standing far away so that when you use Holy Nova it doesn’t hit the mobs.
Holy Nova also displays the same behavior as Fade where aggro does not immediately return after the effect expires. Aggro returns only when an action is taken that the mob is aware of. This makes the spell marginally more useful, but strikes me as a bug. Other people have not been able to reproduce this behavior (with Fade), reinforcing the suspiscion that it is not intended.
Alternative hypotheses specifically tested for (and sadly disproved):
- Holy Nova does not eliminate the threat of spells cast under it’s influence. It reduces aggro and gives it back, just like Fade.
- Holy Nova does not reduce more threat than Fade, it actually reduces significantly less.
== Threat Miscellany ==
HoT and DoT spells generate aggro on each tick, not on cast.
Hunter traps generate aggro like any other DoT.
Bandaging generates aggro like any other HoT. I did not test in detail for the amount of aggro.
Health potions generate threat, although at a reduced rate per HP from Priest healing spells.
Mana potions generate threat per MP returned, which can be quite a lot of threat.
Surprisingly, engineering Explosive Sheep do not generate aggro. Most likely it is treated like a pet and thus generates aggro for itself, lost instantaneously upon it’s tragic death by explosion.
== Mana Efficiency Numbers ==
These numbers were taken off allakhazam for a level 40 priest and verified with the in game stats of my level 40 priest. Talents are included in parens with abbreviations:
Improved Renew (IR)
Spiritual Healing (SH)
Improved Healing (IH)
Mental Agility (MA)
HP per MP, lowest to highest (best efficiency for each spell in bold):
PW:S 1.52
PW:S (MA) 1.69
Renew 1.95
Flash Heal 2.04
Renew (SH) 2.15
Renew (MA) 2.17
Renew (IR) 2.24
Flash Heal (SH) 2.25
Greater Heal 2.34
Renew (SH + MA) 2.38
Renew (IR + SH) 2.44
Renew (IR + MA) 2.49
Greater Heal (SH) 2.58
Renew (IR + SH + MA) 2.71
Greater Heal (SH + IH) 3.03
The take aways?
Renew can smash Flash Heal for efficiency and you should really consider the IR and MA talents.
Spiritual Healing is a win if you have the talent points for it.
Doing PW:S + Renew is a fairly significant drag on your efficiency, but good from an aggro point of view as discussed above.
Doing PW:S + Greater Heal has similar efficiency to doing Flash Heal unless you go up to Improved Healing, so it may make more sense at that point to just Flash.
== HP per second ==
Without Master Healer, Greater Heal is only about 10% higher HPS than Flash Heal. Probably not as important a factor as mana efficiency and the problem of critical spikes.
With Master Healer, Greater Heal does about 25% better HPS than Flash Heal. Perhaps the best way to evaluate Master Healer is being able to pump out that extra 15% HPS. Useful? Someone else will have to answer that.
== Prayer of Healing ==
Obviously a different beast. You have to be sure that you are effectively healing (not too much overhealing) multiple people or it is a waste.
Improved Prayer of healing is scaled to match Flash Heal in terms of HPM. Effectively heal two people and it’s no different than using 2 Flash Heals. So clearly you want to be healing more than 2 people. Without the Improved talent, you need to be healing 2.5 party members just to match Flash Heal for mana efficiency.
The useful HPS of Prayer of Healing of course depends on how many people are taking damage. On any one person, the HPS is around half that of Flash Heal, which obviously stinks. If multiple people need to be healed, then the HPS is incredible.
== PW:S and AC ==
This is a common question because if PW:S doesn’t take AC into account, then its efficiency as a heal plummets.
Factors: AC only impacts physical damage, while PW:S absorbs all damage. Any sort of non-physical damage involved would skew the results.
To test this, I found a mob that only did physical damage, the Whirlwind Shredders in Desolace. I put up PW:S and let it hit me. Since you don’t see the amount of damage the shield is absorbing, you can only count the number of absorbtions and the leftover amount absorbed. I tried with and without Inner Fire on, three trials each to account for any unusual crit patterns. The numbers were pretty conclusive.
Hits absorbed without Inner Fire:
7 hits, 3 rends, 34 extra absorption
9 hits, 3 rends, 0 extra absorption
8 hits, 4 rends, 13 extra absorption
Hits absorbed with Inner Fire:
10 hits, 5 rends, 1 extra
9 hits, 5 rends, 1 extra
9 hits, 5 rends, 8 extra
Clearly the shield absorbs more hits with Inner Fire on, thus we conclude that AC is taken into account before the shield absorbs. Plugging in rough numbers shows a good 20% increase, which seems about right for Inner Fire.
== PW:S and Rage ==
I teamed with a Warrior and asked him to proximity pull without attacking and take some hits. He stood there and took it and I noted that little rage was generated (compared to attacking). I then put PW:S on him and noted that the increments of rage increasing did not appear to change. They were admittedly so small that I’m not confident that they weren’t slightly different. Either way, he was still gaining rage, and PW:S does not seem to make a significant difference.
Two trials done, 37 warrior against 40/41 raptors.
Additional trial done in Defensive Stance, 38 warrior against 39 gorilla.
== Basic Threat ==
Our method was to have the tank pull without attacking and take hits until around 400 health. I would then heal him to full using Greater Heal or multiple Flash Heals, pulling the mob to me. We then tabulated how much damage it took for him to regain aggro.
We ran three trials of each. The tank had to do 7-10 hits of around 60 dmg to regain aggro. We feel that this is sufficiently granular to smooth out the 60 dmg coarseness. Consistently we found that the mob would turn to attack him when he had done around 55-60% (in damage) of the HP I healed him for.
Obviously there is some initial aggro on him for the pull which we can’t account for. We therefore choose the more conservative 60% number. What is conclusive is that it is the same for Greater Heal as for multiple Flash Heals.
We repeated a similar experiement with PW:S, obviously not worrying about waiting for his health to drop to 400. Because it only took him 2 shots to regain aggro, there are granularity issues, so we also did trials where I shielded us both. We were able to narrow in to an upper bound of 33% and a lower bound of 25% damage to healing. Again avoiding absolute numbers, we conclude that PW:S generates half the aggro of a comparable heal.
== Advanced Threat ==
1) Tank 1 and Tank 2 separately proximity pulled (no attacking) two different mobs. I used PW:S on myself and got no aggro response. I then shielded Tank 1 and only the mob attacking him was pulled. We repeated this test with different distances and found that proximity didn’t seem to matter. We did some multiple shielding to verify that awareness has no random chance. Not conclusive, but unlikely.
From this we concluded that targeting clearly constitutes awareness, while being in a group together or being close together does not.
2) I proximity pulled a mob, then Tank 1 pulled it off me with a single shot, previously shown to generate less aggro than PW:S. Putting PW:S on myself or Tank 1 caused the mob to return to me. Putting PW:S on Tank 2 did nothing.
From this we concluded that aggro is applied based on awareness of the target of the heal, not awareness of the caster.
3) We reduced both Tank 1 and Tank 2′s health to allow healing. Tank 2 proximity pulled a mob, then Tank 1 pulled it off him with a single shot. I used a Prayer of Healing to effectively heal both tanks and pull the mob. We then tabulated the amount of damage needed to pull the mob off me. The result was consistent with the 50-60% ratio for the total amount healed.
From this we concluded that the mob was aware of the healing done to both Tank 1 and Tank 2, which is consistent with out findings. That leads us to conclude that Prayer of Healing does not have an aggro reduction, it simply exposes the more complicated parts of the aggro system.
The take away?
As stated at the top, we can only conclusively say that targeting constitutes awareness and that group membership and proximity do not. The very obvious conclusion, although not proven, is that mobs are only aware of the players on their aggro list.
This strongly argues against having the healer pull (including with MC), as self-healing would then aggro all mobs. It also suggests that everyone but the puller should be judicious about which mobs they touch. The healer in particular should count the cost of throwing out damage and making more mobs aware of him or her. With regard to Prayer of Healing, it means that it is no better or worse than Flash Healing multiple people. If you are of the “only heal the tank” school, then you shouldn’t use it. If you’re going to heal multiple people anyway, then Prayer of Healing is the best way to do so. There is no hidden aggro cost or savings in using it.
I think this explains some of the dynamics of large fight healing. The sense that healing aggro gets worse as the fight goes on is validated as more people touch more mobs.
== Subtlety ==
I am spec’d with Subtlety, and did not respec to test. However, I believe we can draw a few reasonable conclusions about it.
1) Since Flash and Greater Heal ended up with the same threat ratio, it’s probably reasonable to assume that Subtlety is appled per HP healed and does not help one spell more than another. It is *possible* though that without Subtlety you would see a difference in threat ratio.
2) Not having Subtlety would serve to exaggerate the advantage of PW:S with regards to aggro.
== Aggro Decay ==
It has been well pointed out in this thread that it would be very significant if aggro decays over time. To test this, we ran two separate experiments. We had previously shown that PW:S is equivalent to between 100 and 125 damage in aggro terms.
Run 1:
My partner shot the mob twice, doing 137 damage. We then waited 3 minutes while it attacked him. I then put up a PW:S on him. The mob did not pull, indicating that his damage based aggro had not significantly decayed over those 3 minutes.
Run 2:
I proximity pulled a mob and PW:S myself. We then waited 3 minutes while it attacked me. He then shot the mob once for 67 damage. The mob did not pull. He shot again for 70 damage and it did. We see that my healing based aggro did not decay either. 67 is certainly a bit coarse, being possibly 50% of the aggro I generated, but if it does not decay by half in 3 minutes, it is not enough to concern you in anything but the longest fights.
When considering HoT you can clearly be confident that the aggro build up is not decaying over the duration.
Related posts:
You must log in to post a comment.
{ 1 trackback }