My personal opinion is that two trees is not optimal. One set of hots can cover most of the burn healing except for a few high tick situations. In those situations, the ticks are high enough that you really want a direct healer.

We do usually run two resto druids and we assign one to burn (me) and the other to the tank. The backup Burn healer is normally a priest. Based on log parses, I usually devote roughly 80% of my healing output to Burn and the priest usually gives 35-45% of his healing to Burn, the rest to the tanks. I stack +healing to make my lifebloom stacks as powerful as possible. We made this chart to help split up our healing:



The columns are number of slashes (0, 1, 2, 3). The rows are time remaining on the Burn debuff. The numbers in the cells are damage taken per second. Italicized cells can be covered by a single lifebloom stack, or less. The priest doesn't help with these guys at all. I put a rejuvenation or regrowth (or both) on the people that are taking small ticks, and lifebloom stacks once the tick size becomes large enough to warrant that. Underlined cells are a little more serious, and will have lifebloom stacks, a rejuv, and either a POM or renew or something from the priest. Bold cells the priest will pull off the tank and begin casting direct heals. Red cells the person is pretty much dead because they screwed up and got slashed late. Note that the 2 slash, 30 second cell is probably miscategorized and should be bold instead of underlined.

Basically this means that the priest will help with direct heals during the final 10 seconds of Burn and also around the 30 second mark for people with high slashes. All other times it will be my hots plus maybe a hot from the priest.

Thank you very much! This was exactly what I was looking for. I'll definitely be sharing this information with my RL. Thanks again.

Our guild has two consistent Restoration Druids for Brut. I personally am assigned to focus on Burns, and the other Resto only does it for the last ten or so seconds, the rest of the time he's on tanks. A second stack of Lifebloom and Rejuvenation at the end of Burn works out almost perfectly, allowing for alternating Swiftmends towards the end if deemed necessary.

We have 2 resto druids for burns. Our system works in that one druid cycles LB on the tank tanking and 3 burn targets or in the case there are less burn targets, he simply fills in the gcd gaps with rejuves. I concentrate my focus 100 pct on burns in the case there are several, slowly filling up hots till the last 10 seconds when I spam 2 regrows then a swiftmend and move straight on to the next person. In the case that we get time between burns we spend quite a bit of time trying to beat each other out on wrath dps.

I think the *best* reason to use 2 Trees on Burns is quite simply flexibility. Burn is one of those fight mechanics that can be extremely easy or extremely difficult depending on the luck of the day. Druids are able to cover far more ground than a direct healer can and in the case that you suddenly find yourself deep into 3 burns and multiple people have slashes for whatever reason, it is very difficult for a direct heal healer or even one tree to keep up. This is also a fight where battle resses are worth their weight in gold. That isn't to say that a direct healer can't do the job, but when things aren't going well, I'd much rather have another Tree helping me.

So the main reason I think two druids isn't the best idea is because I think it is overall better to have the second tree dedicated to the tank. Tank death, at least in our guild, is a big problem and Lifebloom/Swiftmend/NS help very significantly. When we don't have a druid dedicated to the tank, from time to time a tank death would occur in a very small window by a small margin (a few hundred to a thousand hit points, in a roughly 1.5 second window). When this happened the death report would reveal that the tank received zero heals, presumably because the direct healers' heals happened to be timed not to land then. Lifebloom ticks, a swiftmend, an NS, or just simply more hots with more chances to tick would probably have made the difference.

In my view the purpose of the druid on the MT is to make sure he doesn't spike down in less than two seconds, and the purpose of the direct healers is to get him back to 100% after a spike that leaves him at a few hundred hit points. To me that means that when burns are running wild, it's better to pull a direct healer (which you have many of) off the tank than it is to abandon a dedicated MT-healing druid by either pulling him off the MT completely or locking him into a cycle where all he can do is lifebloom the MT (meaning no swiftmend, NS, or extra hots).

I guess my argument comes down to: a tree plus a direct healer can do the job of Burn healing, and I feel that the MT should have a dedicated tree, so I am not in favor of a 2-tree Burn healing setup.

I would think it's slightly more efficient to have the second tree get a stack of HoTs on the victim for the last ten seconds while maintaining them on the tank then to have a single target healer completely switch off for it.

Seems to be the most effective for our guild, anyhow.

So you said you wanted great detail eh? Becareful what you wish for...

My guild use 2 trees on Brut and I believe this is a strong, if not optimal, choice for 2 reasons: both druids can keep triple stacks on the tank and ToL aura.

You really have to take a step back and look at how healing is assigned for this fight. You bring 7 healers to the fight and you want to put as many of them as possible on the tank. Outside of his melee/stomp attacks, which are focused on the tank, there are only 2 other sources of damage to worry about: slash and burn. The thing about these two abilities, and I can't stress this enough, is that you only need to heal the raid enough to survive the ability, not top them off.



Slash can be handled by a single raid healer (be that shaman or coh priest) under the right conditions. Spriests put out a respective amount of raid healing to aid the shaman, especially since after the 3rd slash when the raid healer switches focus to raid members on the other side. In addition, you place any remaining shamans on tank healing using chain heal. As long as the raid is positioned correctly, the 2 extra bounces will make life a lot easier on the raid healer.



The final ticks of burn requires two healers. You have a choice: use 2 druids or use a single druid and a single target healer (priest or pali). I realize burn healing can be done with a pair of single target healers, but I think we can all agree that is sub optimal and not worth discussing.

Anyways, for a single druid and single target healer lineup, the single target healer has to shift attention to burn for roughly the last 10sec of the effect. In a worst case scenario there will be 3 burns up, thankfully that doesn't happen often thanks to resists, ice blocks, cloaks, and bubbles, but it can happen and you should plan for it. In this situation the single target healer would spend 30s on the tank, 30s on burn healing. So, if this healer is spamming the tank hard, their hps is halfed, bring a pali from 2.9K hps down to 1.45K or a priest from 2.2K hps down to 1.1K on average.

If you bring 2 druids, you can have both of them on tank healing as well. There is no longer a druid specifically devoted to just burn healing, you share the workload between the two of you. A burn victim only has to live through the final ticks of burn, not be topped off during them. By placing a pair of triple stack lifeblooms around 25s left, the burn victim will be topped off until the final 6 ticks of burn (3.2K a piece). As long as those triple stacks are ticking for 900 a pop (which ours do), that means the person is taking 8,400 damage. Anyone in sunwell should have that amount of hp. Just in case though, we place a rejuv on them as well, which not only places them at 3,600 damage, but also opens up the option of swiftmend just in case. Not only does this make burn healing smooth, but allows both tree druids to maintain at the very least a triple stack lifebloom on the tank. With a pair of ToL auras (which I'll get into more detail in a moment) this comes out to 1.1K hps per druid, or 2.2K hps on average.



Before I get into ToL aura benefits and my own rotation, I want to touch on the differences between the two. The strength of a single target healer is their ability to "throw down" when the time is needed (stomp). A druid cannot uprank, but can provide regrowth or swiftmend to help out. Please take special note, that a pair of druids put out the same amount of hps as a priest in their best case scenario, and aren't all that far off a pali (700 hps behind). In a worst case scenario, a pair of druids blow that additional pali/priest out of the water, and I strongly recommend you always plan for the worst case scenario (3 burns up). A pair of druids outperforms a priest and is competitive with a pali, but wait, it gets better...



ToL aura really, really shines on Brut since there are so many tank healers. With roughly 650 spirit and the [Idol of the Raven Goddess] the aura adds 200 healing. Different classes gain different benefits (this is assuming no downrank penalty)...

Take into account the spell cast times and you see the following hps boosts...

I'm not going to get into heal roster comparisons, but ToL aura is nothing to sneeze at. By using ToL aura with a pair of trees, you gain an additional 1,720 hps with my lineup. Thats almost like having an additional healer on the tank and more than makes up for any short comings a pair of druids have vs a pali (not to mention beat the pants off a priest).



So, I hope at this point, I've sold people that taking a pair of druids is not only a strong choice for Brut, but maybe even the best choice. Now to touch on my rotations, basically have four options...

lb, lb, lb, lb
lb, lb, lb, rj
lb, lb, lb, rg
lb, lb, lb, sm

It all comes down to planning ahead and looking at boss timers. It just so happens that in a worst case scenario I have four things to worry about and 4 gcd to play with. If my rotation is going to be off, I drop the lowest priority heal to refresh triple stacks, therefore buying me 1.5s. Most of the time you are not in a worst case scenario with 3 burns, and have a lot more room to play, but in any case, this is how I deal with my healing priorities...

End of a Burn...
Maintain the triple stack. Sacrifice the start of a burn gcd to throw up rejuv. If the person is going to die, throw away the rotation and swiftmend

Tank Healing...
Maintain the triple stack. Sacrifice the start of a burn gcd to throw down a regrowth during stomp. Attempt to have a swiftmend open to be used during stomp.

Middle of a Burn...
Maintain a triple stack. If the person is slashed your partner in crime, the other resto druid will need to step in and help you out with an additional triple stack. Depending on the time period, you may even need to have rejuv/swiftmend used.

Start of a Burn..
This is easily handled. Depending on where the boss's cool downs are you can either use a rejuv, or even drop a regrowth to buy yourself some additional time before starting up a triple stack. This is of low priority unless the person is slashed, in which case you need to start up a triple stack asap.


Here's the pair of kills that we've had so far. The first was a bit rough and we botched burns pretty bad, the second was flawless if memory serves

Kill 1
Kill 2

With all that being said, there are many ways to skin a cat.

You do bring up some good points, however you are slightly mistaken here. Yes the person will be toped by the lifeblooms until there are 6 seconds left and the Burn starts ticking for 3200. However your calculations consider the lifeblooms as damage reduction in reaching the 8,400 HP requirement, and not as a heal after the damage is taken.

Consider someone with 10,00 health exactly. With six seconds left on Burn they are topped. The lifebloom ticks each second between the damage.

T-6 3200 damage (6,800 HP left)
T-5.5 LBs heal 1800 (8,600 HP left)
T-5 3200 damage (5,400 HP left)
T-4.5 LBs heal 1800 (7,200 HP left)
T-4 3200 damage (4,000 HP left)
T-3.5 LBs heal 1800 (5,800 HP left)
T-3 3200 damage (2,600 HP left)
T-2.5 LBs heal 1800 (4,400 HP left)
T-2 3200 damage (1,200 HP left)
T-1.5 LBs heal 1800 (3,000 HP left)
T-1 3200 damage (-200 HP left) Target dies
T-.5 LBs heal 1800 (1,600 HP left) But target was dead and could not recieve this healing.

The real amount of HPs someone needs to survive on only 2 900/second lifeblooms is actually 1800 more than 8,400, or 10,200. Actually, since that is the exact amount of damage they take, they need 10,200 +1 = 10,201.

It might be important to know that Burn's damage doubles every 10 seconds, not every 6.

Every 11 seconds. So the last 6 ticks are around 3200. Here a log extract:

Imho the values in the seconds column of this table should be 60,50,39,28,17,6:

Huh, guess you learn something new every day. You're right.

Hmm yeah I knew the last bit was 6 seconds and not 10 (mentioned that on our internal forums but forgot to here) but I didn't make the connection that it was just 11 second stages throughout. Thanks.

Anyway it's first post updating time!

I recently made some changes to the first post to attempt to modernize it, since some of it was still living in the past. Let me know if anyone thinks any of the changes were a bad idea, or if you think there should be more. In short they were: removed the suggestion to drop 3 points in nature's focus, increased haste recommendation for a 5-GCD cycle from "180-220" to "250-280", added language about capping your mana on innervate to the Stats and Innervate section, removed references to trinketed lifeblooms from the Lifebloom section, added distilled wisdom and chromatic wonder flasks, removed Bangle, more strongly discouraged spellsurge, more strongly encouraged Boar's Speed, replace Royal with Purified, added a profession section.

Second, I would really appreciate it if a few of you all could download this attached spreadsheet and check it for usefulness and correctness. It was originally something small I made just to calculate some numbers for the first post, but things got a lot more complicated once Innervate mana capping became a somewhat common phenomenon. It's mostly a regeneration sheet, but it does have a short healing output section. The idea is that you put in your stats from the character screen, the estimated fight duration, what buffs you expect to have, and the manner in which you innervate yourself (if you do). The sheet should then tell you, for example, the mp5 equivalence values of various flasks and elixirs, the value of spirit compared to intellect given your setup, how much various heals should hit for, whether you are capped on Innervate and if so how badly, and so on.

The specific areas I'm looking for feedback on are the "Regen Stat Values" section and the innervate cap math. Those are probably the most important part of the sheet.

If you are kind enough to take the time to help out then I think it's somewhat self explanatory, but if you want some details then here are some brief comments. First about entering your input:

• Start by putting in your completely unbuffed Spirit, Intellect, Healing, and Haste values in the "unbuffed" column in the top left. For the "Pure MP5" value you can either add up the mp5 on your gear manually, use a mod to do it, or you should be able to play around with it until the "OFSR SpiInt + MP5" and "IFSR SpiInt + MP5" cells in the "Mana" section match your "Mana Regen" character screen tooltip (they're supposed to match).
• For the talents section, you want "LS" to be 1 if you don't have it and 1.15 if you do. "Intensity" should be 0 if you don't have it and 0.3 if you do. Naturalist should be 0 if you don't have it and 1 if you do. Sorry for the massive inconsistencies, but fortunately if you have a standard spec you shouldn't have to change the defaults.
• For all the buffs and consumables, you want them to be 0 if you don't have it and 1 if you do. Certain buffs have improved versions, and you want them to be the appropriate multiplier (for example 1.35 is imp mark). These should be written on the sheet for easy reference, and there should be pop up menus to pick from. The buffs and consumables should generally be easier to make sense of than the talents section.
• The "Non-MP5 Mana Gains" section is used only to determine what your Innervate mana cap is. Put how much DPS your spriest does, or 0 if you don't have one, and then put whether or not you have mana spring.
• The "Innervate" section is also mostly concerned with determining the mana cap. Put in how much mana you will have when you Innervate yourself, and estimate how much mana you will go through while the buff is active (just for reference, an LB in tree form every GCD will be 2350 mana). If you're in a rotation, you can look at the first post's Healing Strategies section which lists mana consumption for many common ones.

And for reading output:

• Like mentioned earlier, the "Mana Regen" tooltip on your character screen should match the "OFSR SpiInt + MP5" and "IFSR SpiInt + MP5" cells in the "Mana" section.
• In the "mana" section, the "+10 spirit" and "+10 int" numbers are how much raw MP5 you will get from 10 points of either stat on gear. This means it will factor in Kings and Living Spirit.
• In the "innervate" section, "Net mana (-30%)" is how much an innervate restores to you on top of what you would get otherwise. "Mana cap" is the maximum amount of benefit you can get from that Innervate mana. This is calculated by taking your total mana deficit and adding the amount you expect to spend during the Innervate, and then subtracting the amount of mana you would get from other sources during the Innervate.
• Directly below that, also in the "Innervate" section, is the amount of extra mana you will get per innervate by adding +10 spirit and +10 intellect. If you're capped, these numbers will be 0 and 150 (no kings) or 165 (with kings).
• The "Regen Stat Value" section attempts to put together the various things that spirit and intellect do and tell you how much each is worth in terms of MP5. For spirit this will add some combination of its OFSR and IFSR value (based on your FSR time input), and its benefit to innervate. For intellect you will also get its OFSR and IFSR value, its benefit to innervate, plus its contribution to your mana pool (over a 6 minute fight, the 165 mana added by 10 points of intellect can be viewed as restoring the same amount of mana as 2.3 MP5... and the sheet does view it that way).
• In the same section you can see MP5 values of various flasks and elixirs.

Eventually I hope for it to be something linked from the first post that people can use to help determine what gear to get and what consumables to use. If you have some time, please play around with it, let me know if you find anything interesting, have any suggestions, or just think the sheet is a piece of garbage. If you have some feedback, and it is something that other people might have an opinion on then just post it in the thread; for purely technical questions you can either post in the thread or send me a PM.

First of all thank you for this awesome spreadsheet.

What I'm missing is the spirit scroll in the consumables section. I added it for me since I'm using your first version. Our priests are usually specced 20/41/0 so we don't have divine spirit buff available.

It would be only a small improvement but for the sake of completeness....

Do you happen to know what it stacks with (Mark, DS, Draenic)? I seem to remember it not necessarily stacking with everything.

Good question.

It gets definitely owerwritten by the divine spirit. I'll figure it out when we are raiding the next time. Or probably someone has the right answer already.

good updates for the first post,what I miss:

[Leather Gauntlets of the Sun]
[Kailee's Rose]

in the proffessions-section.

The gloves are BoE, so they're not LW only.

ah yeah sry,the pattern was bop...

Could you also add in [Brilliant Mana Oil] to the database? I think its not that uncommon for people to use it as it often edges out to be one of the best to use.

I've been having a bit of a discussion with Sheshonk (who first identified this opportunity) and was wondering if anyone point me in the write direction to code a little tracker. Basically, there's a program called autokey that will spam click whatever button you're holding down every 1ms (since you still select the spell and target, I don't think it would be an issue because you aren't automating your toon). I'm thinking this should essentially remove the human error component in keeping up a 5-cast cycle (obviously target selection could still be an issue). Here's my theory:

At 113 spell haste, there are two factors that hurt 5-cast cycle ability. Human error and latency variance. Latency is as likely to improve as it is to worsen. This means that if you remove human error, at 113 spell haste you should be able to make the 5-cast cycle 50% of the time (assuming a normal distribution for latency variation).

I am sitting just under the 50% success mark right now, and think that I should hit it with around 235-240 spell haste. That means around 125 spell haste is offsetting the human error element. So in theory, autokey could reduce the spell haste requirements by as much as 125 haste.

I want to run a test where I create 5 macros to cast lifeblooms on 5 different targets, and use autokey to try to maintain a 5-cast cycle (thus I've also removed the target selection issue). I'm then going to vary my haste down to the point where I have a 50% success rate with refreshing LBs. If my theory is right, that point should be close to 113. Even if I'm wrong, the test should give me a good idea of the human error element (for me at least) and thus the potential value of autokey.

Now, I've never written any of these tracking or data analysis files. Is this a feasible thing to track? Could someone point me in the right direction for how I might go about this? I have a bit of programming experience (couple courses in Java) but not a whole lot.

There is a blue post confirming the use of autokey will grant you a ban, do not use it. Blue Post

Please excuse the copy/paste from my guild forum post, I believe a majority of it is relevant...

Laginfo is the best way I've found to track. Obviously, instead of nuking dr boom you would attempt to keep 5 lifeblooms up. As Kalaghan said, would be best to use a cast sequence macro with /target person instead of manually choosing people with raid frames to push this to the limit.

You could potentially look through combat logs, but those are made client side and have the opportunity to not only be incorrect, but miss information.

This sounds interesting to me, but before I would even consider playing with it, I would really want to see the blue post confirming they will not ban people for it.