Note: This post contains outdated numbers and mechanics for several raid cooldowns that were changed in Patch 5.4. However, the philosophical discussions of what cooldowns to use when, and the table at the bottom listing all raid cooldowns and their details, are still accurate and relevant.
|Raid Cooldowns 101|
|Healer Throughput CDs|
|Non-Healer Throughput CDs|
|Damage Reduction CDs|
|The More You Know: Spirit Link|
|Relative Strength of CDs|
|Available Raid CDs|
In the words of a dear friend, one facet of progression raiding is “raid cooldown choreography”, and in recent expansions – with the proliferation of raid cooldowns amongst a variety of specs, intended to lend utility and encourage raid leaders to bring all the classes – this is more prevalent, and more micromanagey, than ever. A lot of threatening mechanics are dealt with by throwing more cooldowns at them, until the raid survives, and then repeating this until the boss is dead or the raid runs out of CDs.
Recently, some of the more iconic raid CDs (Tranquility and Divine Hymn, but also Vampiric Embrace and Revival) were buffed to make them more competitive in 25-player content. “Finally!” said everyone ever, “Now we can compete with Power Word: Barrier and Devotion Aura in 25-mans!”
You see, some cooldowns scale with player power; others scale with incoming damage, and thus indirectly, with raid size. And the ongoing perception was that even though PW:B and Devo Aura don’t show up on the meters, there was no way that raw healing cooldowns could heal for as much as those cooldowns mitigated. That Holy Paladins and Discipline Priests were ahead on the meters already, plus their incredibly powerful raid cooldowns don’t even show up there, seemed somewhat imbalanced to a lot of players, who welcomed this change in the hopes that it would even up the score.
Well, this all got me curious. How close are the new 25-player healing cooldowns to the damage reduction cooldowns? How close are they in 10-player formats? What about the value of Rallying Cry, or Spirit Link – can they be quantified? And the cooldowns that didn’t get buffed – how are they doing by comparison? Prithee, Lightwell, how farest thee?
When I get curious about something, I tend to get a little spreadsheety about it too. In this article, I’ll be presenting a lot of information – some of which you already know, but hopefully some that you don’t – about all of the raid cooldowns we have available to us in this new dawn of cooldown choreography, and how they compare to one another in effectiveness.
Remember, this is not indicative of class balance – each class has its own toolkit, its own strengths and weaknesses, and this is just shining a spotlight on a single compartment of that toolbox, measuring the relative shininess of each class’s socket wrench, whilst ignoring the fact that some classes have a full wrench set but no drill bits and others have super fancy diamond-tipped drill bits and Monks are all, “Pfft I don’t use tools, man, I fix the plumbing with my mind and nail pictures to the wall with my bare hands” and wow this analogy has gone a little crazy so I’m gonna stop it right here and move on.
Though you probably already know what they all do, just in case you don’t, I’ve created a table of existent raid cooldowns. (It’s a long one, so I’ve shoved it down at the bottom of this post, where you can safely ignore it if you don’t need to refer to it.)
There are three different mechanisms by which cooldowns assist a raid in coping with heavy incoming damage:
- Healing. These cooldowns deliver large amounts of healing, rapidly. Note that damage must first be dealt – and survived through – in order for these cooldowns to be effective.
- Effective health. These cooldowns increase the player’s effective health, or the amount of damage they can take before dying. Effective health cooldowns raise both the maximum and current health pool by the same amount, and when these effects fade, that amount is typically subtracted from both the current and maximum health. This cannot kill you, but it can leave you on dangerously low health and susceptible to death immediately after the effect fades.
- Damage absorption. These cooldowns prevent some portion of the incoming damage, thereby increasing the effective health of the target, but do not subtract any additional health from the player upon fading. This class of cooldowns includes percentage-based damage reduction such as Power Word: Barrier and Devotion Aura, as well as strict absorption effects like Spirit Shell. Some of these cooldowns prevent variable damage depending on the type and amount of damage being dealt during the encounter.
To really illustrate and clarify the differences here, let’s consider an example. Consider a one-off attack that will deal a fatal blow – a version of Interrupting Jolt, for example, that deals 110 damage, when players only have 100 health. And consider three cooldowns on the raid: Revival, which heals for a burst of 20; Rallying Cry, which increases health by 20%; and Spirit Shell, which absorbs the next 20 incoming damage.
- With Revival: Players start at 100 health. 110 damage is dealt. They die. Revival goes unused.
- With Rallying Cry: Players start at 100 health. Rallying Cry increases that to 120. 110 damage is dealt. Players are on 10 health. When Rallying Cry fades, players are set to 1 health and are in extreme danger of dying. But, they lived!
- With Rallying Cry and Revival: Players start at 100 health. Rallying Cry increases that to 120. 110 damage is dealt. Players are on 10 health. Revival heals players for 20 health, bringing them to 30 health. When Rallying Cry fades, players are on 10 health.
- With Spirit Shell: Players have 100 health plus a 20-HP shield. 110 damage is dealt, eating through the shield first, and leaving the player on 10 health.
You can see immediately why absorption cooldowns and mechanics are so powerful: they increase the effective health of the player and leave them in a higher health position as if they had been healed as well. A Spirit Shell that absorbs X damage provides the equivalent of X effective health and X healing!
That doesn’t mean that these damage-absorption cooldowns are twice as strong as their effective health and healing counterparts. Healing and effective health can’t really be directly compared like that, in a one-to-one ratio. In some situations, healing will be more beneficial than effective health – e.g., for pulsing AoE damage wherein no single hit alone will kill you, healing will really shine, and effective health is not terribly impressive. In other situations, effective health is the only way that you can survive.
In most common situations (periodic damage abilities like Rampage, non-fatal one-off abilities like real Interrupting Jolt), healing and effective health will both be useful, but to varying degrees. However, in all situations, damage-absorption cooldowns will be useful, because they cover both angles, and damage you don’t take is damage that can’t kill you.
To demonstrate this, let’s consider a second situation, a DoT effect that deals 105 damage over 15 seconds (7 ticks of 15 damage) to a player with only 100 health. Again, we have the Revival that heals for 20, the Spirit Shell which absorbs 20 damage, and the Rallying Cry that increases health by 20.
- With Revival: The heal will be fully effective after the second tick, so we hold off ’til then. The player’s health looks like 100 – 85 – 70 – 90 (Revival) – 75 – 60 – 45 – 30 – 15. The player survives.
- With Rallying Cry: The player’s health is increased to 120 when Rallying Cry is used. Since Rallying Cry lasts only 10 seconds, but the damage is being dealt over 15 seconds, we have two options – using it to cover the first 10 seconds, and using it to cover the final 10 seconds. Let’s compare:
- Use at the start: Players’ health begins at 120. The effect will wear off between the 4th and 5th tick of the ability. The player’s health looks like 120 – 105 – 90 – 75 – 60 – 40 (loss of RC) – 25 – 10 – 0 (5 overkill). The player does not survive.
- Use to cover the end: Players’ health begins at 100, and Rallying Cry will kick in after the 3rd tick. The player’s health looks like 100 – 85 – 70 – 55 – 75 (RC kicks in) – 60 – 45 – 30 – 15. The player survives the mechanic, but shortly thereafter, RC fades, and their health is set to 1.
- With Rallying Cry and Revival: Players begin on 100 health. The player’s health looks like 100 – 85 – 70 – 90 (Revival heal) – 75 – 95 (RC kicks in) – 80 – 65 – 50 – 35. The player survives the mechanic with a large buffer, and when RC fades, their health is at 15.
- With Spirit Shell: The shield prevents the first 20 damage from occuring. Player’s health looks like 100 – 100 – 90 – 75 – 60 – 45 – 30 – 15. The player survives.
As you can see, in the periodic-damage-that-will-kill-you category, effective health is the only type of cooldown that will not necessarily save the raid. However, if an effective health cooldown is used in such a manner that it will expire after the completion of the periodic-damage, then for the duration of the boss mechanic, at least, effective health cooldowns act like a heal. (A temporary heal that then expires, leaving you very susceptible to death, yes. But a heal nonetheless.) Since all we really care about in this analysis is making it through the damaging event without losing raiders, we can approximate Rallying Cry and other effective health cooldowns as healing so long as they are timed appropriately.
You can also see that absorption effects and healing effects on this type of damage pattern end up identical, rather than – as with the Deadly Interrupting Jolt example above – absorption effects being superior. This is because temporary effective health for extended periods of periodic damage is simply not very valuable. Fortunately, absorption effects still have the “act like a heal” property as well as the “act like effective health” property, and so they maintain their usefulness across all damage patterns.
To analyse cooldowns that are encounter-specific, I first chose three different types of mechanics: a one-off, large attack (Heroic Dark Animus’s Interrupting Jolt), and two stacked-up DoT effects (the magic-damage Rampage in Heroic Megaera, and the largely physical-damage Fist Smash/Burning Blast confluence in Heroic Iron Qon). Using logs of my own guild’s kills of these bosses, and the encounter journal, I calculated the amount of damage dealt by each mechanic in each raid format:
|Damage Dealt by Heroic Encounter Mechanics|
|Megaera||Dark Animus||Iron Qon|
|Format||Rampage #5||Interrupting Jolt||Fist Smash #8|
I then used the tooltip values and Wowhead’s additional data (e.g. duration, radius, etc.) to determine how much damage each cooldown would mitigate for each of these attacks.
Cooldowns with a limited radius – Power Word: Barrier, Spirit Link Totem, and Anti-Magic Zone – needed a further step of analysis: Determining how many players could be inside them for Interrupting Jolt whilst maintaining a 5-yard spread to avoid Anima Font deaths.
I searched for ages for a pre-defined formula that would help me answer this question, but had to settle on completely fudging it with a diagram and a little bit of extrapolation and guessing. The damage reduction provided by these cooldowns was adjusted to reflect their inability to cover the entire party.
For the cooldowns that are solely gear-dependent, I wanted to get comparable gear sets – similar iLvl and budget of primary/secondary stats – for each spec so that all of the cooldowns could be compared on an equal footing. Now, since I don’t know everything about every spec, and especially not the DPS specs I wanted to calculate cooldowns for, I chose to pawn my efforts off to an external party: Ask Mr. Robot.
I selected non-Heroic-mode gear, since the intended audience for my blog is not really players who are already killing these bosses on Heroic. For each class I chose random professions, stuck to Alliance races to eliminate Horde racials and whether it’s reasonable to expect they’d be up for every use of the cooldown in question, applied full Valor upgrades to 12 items (which was the maximum you could have had at the time I began writing this post), and included the Legendary cloak. I selected (for the most part) the Default PvE profile and Asked Mr. Robot to do all the gemming, enchanting, and reforging for me, and then recorded the Intellect, Spell Power, Haste, Critical Strike, and Mastery ratings for every class.
Now, I know, I know. I know. It’s not cool to gear your character blindly by what a website recommends, without thinking about the demands of your raid and your own playstyle’s mana-efficiency and I know I know I know. I’m not saying these are the best possible gear sets, I am just saying they are all iLvl 536-538, have similar stat budgeting, and take into account raid buffs, profession buffs, and consumables. Every healer profile has way too much Spirit, for example, which could be reforged out of to gain additional throughput stats, but since this is a flaw across all the specs, I chose to leave it so I didn’t spend the rest of my life working on this analysis 😛
For those cooldowns which are dependent upon damage done (Ancestral Guidance and Vampiric Embrace), I exported the Ask Mr. Robot profiles to SimulationCraft and ran a quick sim to determine how much DPS the gear profile/class combination would produce. Whatever SimCraft settings exist were all left on default, because seriously you guys this is the first time I’ve ever SimCrafted anything and wow it looks scary! ><
You may well disagree with the choices I’ve made here, or with the numbers they produce, and that’s awesome, honestly. I just know that the more I fiddle with these details, because I know some specs better than I know others, the more bias I’d be introducing towards those specs, and I’m trying to avoid that to the best of my abilities. I’m totally open to evidence-based critique, so please post if you have any questions or corrections!
Finally, in order not to tear my hair out entirely, I’ve made the assumption that all the healing generated by these abilities is fully effective (i.e. no overhealing), just for the purposes of the calculations. Some raid cooldowns are more prone to providing overheal than others; Tranquility, for example, because it has a strong HoT component, is probably going to overheal more often than Revival would, assuming that both cooldowns are used at appropriate times. I’m just flat-out ignoring that for the calculations I’m presenting, but I’ll keep reminding you of it throughout the post.
The gear sets I used to calculate the value of these healing cooldowns are presented in the table below. You can click the links to see the gear set in question if you’re curious. This table includes Earthliving Weapon’s contribution to Resto Shaman Spell Power, and assumes that Priests are running the Inner Fire buff rather than Inner Will. When observing the apparent Spell Power imbalances, keep in mind that the classes have other mechanics which tend to balance this out – e.g. Holy Paladins have Seal of Insight increasing all healing by 5%, Resto Druids have Naturalist, and Mistweaver Monks have the healing bonus of Stance of the Wise Serpent.
|Healer Stats in iLvl 537-538 “BiS Pre-Heroics” Gear|
|Resto Druid||Mistweaver Monk||Holy Paladin||Disc Priest||Holy Priest||Resto Shaman|
Using these statistics, I’ve calculated the total throughput of each of the healing cooldowns that the various healing specs bring to a raid, as shown in the table below. I’ve noted all relevant buffs and assumptions in the “Comment” column of this table.
|Healers’ Throughput CD Comparisons|
|Tranquility (10-player)||2,842,807||+10% w/NV, +15% w/ToL, +26.5% w/both|
|Tranquility (25-player)||6,822,736||+10% w/NV, +15% w/ToL, +26.5% w/both|
|Revival (10-player)||2,094,458||Also removes Disease, Magic, Poison debuffs|
|Revival (25-player)||5,236,145||Also removes Disease, Magic, Poison debuffs|
|Spirit Shell: PoH||3,217,296||Assumes Archangel & BT|
|Spirit Shell: PoH w/PI||3,676,910||Assumes Power Infusion, Archangel & BT|
|Divine Hymn (10-player)||2,835,073||Worst-case scenario of 50% of ticks self-buffed|
|Divine Hymn (25-player)||6,791,215||Worst-case scenario of 48% of ticks self-buffed|
|Lightwell/Lightspring||3,663,447||Glyphed for 2 extra “charges”|
|Healing Tide Totem||2,572,883||Average raid health of 40%|
|Ancestral Guidance||7,295,946||Average raid health of 40%: UE-HR, 4x CH(4)-RT|
|Ascendance||6,420,664||Average raid health of 40%: pre-cast UE-HR, 4x CH(4)-RT, HR, CH(4)-RT|
Tranquility and Revival are both difficult to maximise, for opposite reasons. Tranquility deals a lot of its healing by leaving a HoT effect on the players it has healed. Sometimes this is great, because you don’t need all that nearly-7-million healing at once, and more damage is coming. But sometimes it’s not so great – like, for example, right after an Interrupting Jolt, where all the damage has been dealt, and the other healers in your raid are also working hard to top everyone up quickly. The HoT effect can be largely overheal in situations like this, so for responding to one-off large attacks, Tranquility may look a lot less powerful than its potential.
And for Revival, all that healing is dealt in one giant burst. At 5.2 million healing with a single GCD (for 25-player Monks, at least), that can do some pretty insane things for your average HPS. However, it also means that each player needs to have a health deficit of approximately 208k before the healing that Revival will do becomes fully effective. In practice, you’d want to see an even larger health deficit than that, because other healers will be doing things too.
When considering how much healing Revival provides, you really ought to take into account its ability to dispel harmful debuffs – how much damage would those debuffs have dealt had Revival not cleared them? How much mana would other healers have spent using single-target dispels (or the even more expensive Mass Dispel) to clear them up, and how many GCDs, and how much healing could they have done in that time? It’s fairly complex and the simplistic analysis I’ve done here really isn’t sufficient. It does mean that sometimes, Revival may be used primarily for its dispel capacities, rather than its burst healing capacity, and in these circumstances, don’t let the relatively high overheal of the ability – and its low representation on your meters – get you down. It provided an incredibly valuable service (*cough*Venomous Effusions*cough*).
I know, Light’s Hammer isn’t really a “raid cooldown” the way, say, Tranquility or Healing Tide Totem are. It’s roughly analogous to Spirit Shell in that it can be active every minute, and for the sake of evenness I’ve added it in. Plus, I was getting kind of sad not seeing any pink in the table. I have opted to leave out Avenging Wrath, Divine Favor, Guardian of the Ancient Kings, Holy Avenger, and Sanctified Wrath, for a number of reasons. I do not think that most raid leaders classify these as “raid CDs”, even though they may well classify Ascendance as such, and Ascendance is pretty similar to GotAK in function. The wider range and longer duration on Ascendance, plus its synergy with AoE healing spells in a way that GotAK just doesn’t have, makes a large difference here.
These Paladin CDs are all more like “personal throughput CDs”, and they are incredibly flexible, allowing them to function as mini-raid CDs or as single-target healing CDs. Since there’s so many ways to use them, and which particular option you choose depends heavily on your Talent tree choices and your role in the raid, I am leaving them out for the time being. (I do have some calculations of this sort in the spreadsheet linked to on my Paladin Changes page, but it’s not really an exhaustive analysis.)
Divine Hymn numbers above do not count the 10% healing boost from other healers’ spells. In order to estimate the maximum potential throughput this buff can create, I have calculated the average raid HPS, excluding the Holy Priest, for a 25-player raid with one of each healing spec in attendance, and assuming that no other throughput cooldowns are being used during Divine Hymn’s uptime (because I didn’t want to do endless permutations).
Assuming that for 8 seconds all healing done by the raid is buffed by 10%, this works out to 1.29 million healing generated by the Divine Hymn incoming healing buff. If you count this in to Divine Hymn’s throughput, it reaches 8.09 million healing, making it approximately 10% better than its nearest competitor, Ancestral Guidance. Of course, all of this healing is rather unlikely to be effective, but that is true of each of these cooldowns.
Finally, Shamans. I want to note that the incredible disparity between HTT and AG/Ascendance is not really fair. HTT costs only a single, one-second GCD to deliver its healing, and the Shaman can continue casting whatever healing spells he or she likes while it’s up. On the other hand, the AG and Ascendance cooldowns are inextricably linked to the amount of ordinary healing you’re doing, and so those numbers include that healing in them. You can probably do 2-3 million healing during HTT’s uptime, so if you want to make a more informed choice between HTT and AG you should factor that in.
That said, wherever you can get away with using AG, you really should. It’s phenomenal. Any time you need a lot of healing whilst stacked, it will really shine for you.
I don’t recommend stacking Shaman cooldowns. Well, I mean stacking AG and Ascendance, primarily. Since patch 5.2, they do not work together anymore, so while you will do a lot of healing while both cooldowns are up, you would have done the same amount with those cooldowns if you used them separately. And the more dangerous, spiky damage situations you can cover with your cooldowns, the better off the raid is. Using Ascendance or AG with Spirit Link Totem is way more defensible, for reasons which I will elaborate on in my Spirit Link dissertation below 😉
Once again, I’ll start with presenting the gear sets I used and the stats each class would have while wearing them. For those classes with DPS-dependent healing abilities, I’ve listed the SimulationCraft DPS I obtained from running a default-setting sim; you can repeat this by following the link to the Ask Mr. Robot gear profile, clicking ‘Export’, and following the instructions to generate the SimCraft data.
|Non-Healer Stats in iLvl 537-538 “BiS Pre-Heroics” Gear|
|Balance Druid||Feral Druid||Guardian Druid||Shadow Priest||Ele Shaman||Enh Shaman|
Note that I have chosen not to include a weapon swap for Feral and Guardian Druids – best practice would be for the Feral or Guardian Druid to carry around an additional weapon, with Spell Power and ideally Critical Strike and Haste, and to use it whenever using Heart of the Wild for healing or caster DPS. However, I didn’t want to end up reporting an entire table of results just on Tranquility alone, so I left it out. 🙂
The results of my calculations for the non-healers’ healing CDs are displayed in the table below, with all assumptions/circumstances laid out in the “Comment” field.
|Non-Healers’ Throughput CD Comparisons|
|Feral||Tranq (10)||2,544,693||HotW, no weapon swap|
|Tranq (25)||6,107,262||HotW, no weapon swap|
|Guardian||Tranq (10)||2,280,008||HotW, no weapon swap|
|Tranq (25)||5,472,018||HotW, no weapon swap|
|Shadow||Symbiosis Tranq (10)||417,568|
|Symbiosis Tranq (25)||851,816|
|Vamp. Embrace (10)||1,719,890||Unglyphed|
|Vamp. Embrace (25)||4,299,725||Unglyphed|
|Ele||Healing Tide Totem||975,505||5.3 version|
|1,463,258||Proposed 5.4 version|
|Ancestral Guidance||2,577,600||Just damage-to-healing conversion|
|Enh||Healing Tide Totem||850,128||5.3 version|
|1,275,192||Proposed 5.4 version|
|Ancestral Guidance||2,585,103||Just damage-to-healing conversion|
|5,571,425||Pre-cast HR with Glyph of Healing
Storm & 5 MW
|All Warriors||Rallying Cry (10)||1,240,000||2 tanks @900k HP, 8 others @550k HP|
|Rallying Cry (25)||2,890,000||2 tanks @900k HP, 23 others @550k HP|
A few points that I want to make here:
First, Shadow Priests’ Symbiosis: Tranquility is really bad. I’ve said it before, and I said it again, and I’m saying it again now. Seriously, do not bother with using this ability. Even when buffed by 2.5x in 25-player mode, it cannot break 1 million healing.
You’ll do almost as well with a Halo as with Symbiosis: Tranquility, and Halo doesn’t have an 8-minute cooldown and an 8-second channel, and it deals damage. You’re giving up ~1.5 million damage to do less than a million healing, and I’m just not sure that’s ever worth it.
Heart of the Wild/Tranquility, on the other hand, is very strong. This kind of makes sense – not only are you blowing an 8-minute cooldown on the Tranq, you’re also sacrificing 8 seconds of a 6-minute cooldown in order to deliver that healing. Sacrificing that much DPS has got to produce a lot of healing in order to be worth it, and in this case, it definitely is. There are a lot of situations where a short burst of 5-6 million healing from a DPS or tank is going to make a huge difference in your raid’s survivability.
While the Balance numbers are technically better than those for a Restoration Druid’s Tranquility, remember that Resto Druid Tranq has a 3-minute cooldown, and that I didn’t count having Tree of Life/Nature’s Vigil buffs up in those assessments, because Soul of the Forest/Heart of the Wild are incredibly popular rival Talents. A Resto Tranquility with either one of those cooldowns will surpass what a non-Resto Tranq+HotW will do, so there’s very little reason to feel cheated by these numbers.
Finally, the comparison between Ancestral Guidance and Healing Tide Totem for the DPS Shaman specs shows Ancestral Guidance as being alarmingly ahead – even in light of the intended Patch 5.4 changes to the Totem that will improve its throughput for these specs by 50%.
In the calculations above, I am using only the average DPS value, not a buffed-by-cooldowns value, so you could get a lot more out of this cooldown by pairing it with Ascendance when it makes sense to do so. It doesn’t always make sense to do so, though, so I’ve presented a more middle-of-the-road option. (If you use AG while you have no cooldowns active, of course, you’re likely to see a little less healing out of it than I’ve shown above, because SimCraft has rolled your cooldowns into that average DPS value, but cooldowns do not work that way in ‘real life’.) I’ve also removed a GCD’s worth of damage from the calculations, to mimic the fact that you may be activating AG right at the end of a spell cast, while the spell’s in flight, and flight time may waste a little bit of AG’s uptime.
The difference between AG and HTT is particularly overwhelming if we include a pre-cast Healing Rain in the calculations. Ancestral Guidance will happily convert your damage to healing while also converting your healing to more healing, and honestly if you’re in a position where you can afford to drop the Healing Rain, I think you should. At the start of Megaera’s Rampage, for example, there’s several seconds where there simply isn’t anything to target or DPS; an Elemental Shaman would do well to drop a Healing Rain during this time, cast Flame Shock on the new head, and then activate AG to take advantage of this.
For Enhancement Shamans, this is even easier – save up 5 stacks of Maelstrom Weapon and sacrifice a Lightning Bolt’s worth of damage for an additional 3 million healing. With the option of an instant, double-strength Healing Rain, at such a low cost to DPS, I don’t see an excuse for an Enhancement Shaman to use AG any other way.
These cooldowns mitigate damage based upon the amount of damage being dealt to the raid, and therefore you have to know what abilities you are mitigating in order to determine how much damage will be prevented. If you’ll recall, I’m using three mechanics from Heroic Throne of Thunder encounters: Megaera’s Rampage #5, a heavy magical damage DoT; Dark Animus’ Interrupting Jolt, a one-off, strong magical attack; and Iron Qon’s Fist Smash #8, a heavy physical damage DoT with a small magical damage component.
To display the relative power of each cooldown vs. each mechanic, I’ve done up a little bar chart. Since some of these mechanics are more powerful in 25-player than in 10-player, I’ve also separated the two raid formats into their own graphs.
Before moving on to the 25-player chart, I want to specify a few things:
- Zen Meditation doesn’t work on a lot of mechanics, but it does work on Megaera’s Rampage, so I have added it in only for that.
- Anti-Magic Zone does so well on Interrupting Jolt and Rampage #5 because of its unique mechanics. So long as all of the damage is delivered at once, Anti-Magic Zone doesn’t really notice that its cap has been exceeded, and it will just straight-up mitigate 75% of the damage. In my guild’s logs, the first tick of Rampage hit all players at the exact same millisecond, so in this analysis, I’ve calculated Anti-Magic Zone’s effect as reducing that entire first tick by 75%. This should be true so long as everyone is stacked tightly when the first Rampage attack goes out – the damage may be delayed by up to 0.5 seconds if some players are straggling behind, and thus the Zone may “pop” early.
- I haven’t included Demoralizing Banner in the Rampage #5 analyses because it only reduces damage dealt by enemies within 30 yards, and I have frankly no idea where the damage during Rampage actually comes from. Graphically, it appears to come from all the heads including those in the back – some of which would be out of range of the Banner – but I’m not sure that’s really true. If you’ve got evidence of Demoralizing Banner being effective, please post it here, and I’ll add it to my analysis!
These caveats hold true for the 25-player version of the graph as well:
With this analysis, you can see just how ridiculous Power Word: Barrier can be. The Fist Smash vs. Rampage disparity in its power, however, exists simply because Fist Smash deals more damage on 25-player than it does on 10-player mode.
You can see in these graphs that there are a few limitations to some cooldowns’ abilities to mitigate damage. For example, Power Word: Barrier performs very poorly on Interrupting Jolt due to the fact that it has a limited area of effect and the entire raid can’t pile into it without incurring Anima Font deaths. By comparison, Devotion Aura does much better here, thanks to its 40-yard range from the casting Paladin, but it does poorly on Fist Smash where the proportion of magical damage to total damage is so low.
While Spirit Link performs somewhat poorly in these comparisons, the graphs above are only taking into account the 10% damage reduction that the totem provides. There is very little mathematical way to analyze – in a general effectiveness sense – the value or benefit of the health redistribution, which can be a very powerful effect.
I’ll start off with the more general, difficult-to-evaluate benefits as I see them. Spirit Link’s health redistribution has several nigh unquantifiable effects:
- It evens out the healing received by the raid during its duration.
Most healers have certain spells that heal only a select few targets – for example, Chain Heal, Prayer of Healing, Wild Growth, Uplift. In a vacuum, when a Priest in a 25-player raid casts Prayer of Healing on one party, that party is “saved” from death against the next attack, but the remaining 20 raid members are not benefited by this spell. Under Spirit Link Totem, all healing is redistributed amongst the raid more or less evenly, making it less likely that any one player will be unlucky and miss out on smart/target-limited heals and later die.
- It reduces the possibility of overhealing, thus increasing efficiency.
By “stealing” health away from high-health players and redistributing it to low-health players, Spirit Link provides wiggle room for non-target-capped AoE heals to be fully effective. This affects Healing Rain, Holy Radiance, Halo or Divine Star, Spinning Crane Kick, Revival, Chi Burst or Zen Sphere’s Detonation, etc. It can also affect targeted heals that can sometimes be sniped by other healing before you’re finished casting the spell.
Just as an example, if you are in a 10-player raid and your Holy Priest co-healer has just Prayer of Healinged party 1 to full health, while the other party is at 80% health, 50% of your next Healing Rain ticks are going to be fully overheal (on party 1) and at 20% Mastery benefit on party 2. With the Restoration Shaman gear profile above, that next tick of Healing Rain would deal about 67k effective healing. However, if Spirit Link is down, you get 10% Mastery benefit on all 10 players, and you end up dealing about 124k effective healing – an 86% increase in throughput.
In practice you’re likely to see less of an effect than this – I mean, who is going to use Spirit Link Totem when the raid is on average at 90% health, right? – but the principle remains the same. And this helps more than just you and your Deep Healing/Healing Rain interactions. It helps any healer with spells that tend to get ‘sniped’ – existing Rejuvenations and Wild Growths, Eternal Flame HoTs, Echoes of Light, Renewing Mists, Gift of the Serpent, and many other small/splash healing effects will invariably benefit.
- It “smooths” individual variations in damage reduction out across the entire raid.
In other words, personal survivability CDs become raid survivability CDs. That Holy Priest from the previous point is a very conscientious raider and uses Glyphed Fade during every Rampage. He takes 10% reduced damage. Either way, the damage taken by the raid as a whole is reduced by 1% (remember this is a 10-player raid). But with Spirit Link, damage on each player is reduced by 1%, instead of one player getting a 10% reduction and the rest getting zip.
Now, I know that’s not exactly an exciting development. No raid leader is going to be like “Hey, pop your 1% damage reduction cooldown so we survive this mechanic!” But when you factor in that there are a number of passive or active damage reduction mechanics that players will probably be using already – Fade, Glyph of Inner Sanctum, Shadowform and Moonkin form (for now, at least), all the passive damage reductions that tanks have as well as active mitigation abilities, Feint, Anti-Magic Shell, Divine Protection, Diffuse Magic, Shamanistic Rage, Barkskin, Ironskin, Hand of Sacrifice, etc., etc., etc. – it can add up to quite a bit of a full-raid buffer when these occur whilst Spirit Link Totem is active.
This protects those players who don’t have cooldowns active – either because they are on CD from mitigating other dangerous mechanics, like Cinders or Matter Swap, or because their class just doesn’t have a lot of options – and those are the players who need the most protection.
Now for the more quantifiable part of Spirit Link Totem. Bear with me here, this is going to be a bit of math (I promise it’ll only hurt a little *cracks whip, adjusts glasses*) and a fair bit of talking, but I want to be explicit about all of this because I think a lot of people simply do not appreciate the power of the health redistribution effect that Spirit Link provides.
For starters, I’ll explain exactly what Spirit Link does. Every second, Spirit Link Totem whips out its little calculator and adds up two values: the current health total of everyone in the raid, and the maximum health total of everyone in the raid. It then divides total current health by maximum current health to work out the aggregate health percentage of the entire raid. It then sets every player’s current health such to that percentage of their average health. Just as a quick illustration, I’ll look at an example of two players, one on 535k/535k, and one on 193k/735k. Spirit Link sez:
Total current health: 535k+193k = 728k
Total maximum health: 535k+735k = 1270k
Aggregate health percentage: 728k/1270k = 57.3%
Player 1’s new health: 57.3% of 535k = 306.7k
Player 2’s new health: 57.3% of 735k = 421.3k
What Player 1 lost, Player 2 gained. It’s pretty straightforward, but note that the presence of high-health-pool players will tend to drive the overall raid’s aggregate health percentage down, and thus result in more dramatic shifts in health. This is a good thing, because those high-health-pool players have more health to spare before they are in danger of dying!
Now on to a somewhat more specific illustration. I’m going to use, as an example, Heroic Megaera’s fifth Rampage on 25-player mode. It is a very dangerous mechanic, delivering over 50 million damage in 20 seconds’ time, with the damage being dealt in calculable chunks every second. I’ve made some assumptions to make the analysis easier to handle without having to actually reconstruct every second of 25 players’ combat logs:
- The raid is 6-healing, with one healer of each spec present;
- The raid is on full health leading into the Rampage (or close enough so that residual healing + the immediate redistribution tick from dropping Spirit Link puts them on full);
- No other raid CDs are being used (this is wildly inappropriate for reasons you will soon see, but I am analysing Spirit Link only, and don’t want to get bogged down);
- There are 23 players with 550k health and 2 tanks with 900k health;
- I am ignoring all passive damage-reduction effects, again, in an attempt not to have to construct 25 combat logs.
- Each healer is delivering a maximum throughput AoE healing rotation, including some pre-charged abilities (e.g. Wild Mushroom: Bloom and Spirit Shell).
With all these assumptions – and yes, again, this is not a realistic scenario of how you would actually play the game, it is for illustrative purposes only – each player in the raid is receiving 85,118 healing per second. Spirit Link is dropped to cover the first six ticks of Rampage, and its immediate tick is used to stabilise all players’ health percentages just before the Rampage begins.
In the table below, I’ve outlined the raid’s current health following each damage-healing-redistribution event. The “No SLT” column shows what would happen if no cooldowns were used during this Rampage, and players are relying on the background AoE healing to keep them alive. The “SLT 10%” column shows only the effects of the damage reduction ability that Spirit Link provides, without taking into effect the health redistrubtion. The “Full SLT” column considers both facets of Spirit Link’s functionality.
Cells highlighted in lime green indicate the ticks of Rampage that are being mitigated by Spirit Link Totem. Cells highlighted in orange indicate that players would die to the next tick of Rampage if the background AoE healing was late or for some reason stopped, or if no personal/other raid cooldowns were used. I’ll call this the “danger zone” later on. A ” — ” entry indicates that the player has died.
|Spirit Link Totem’s Effect on Raid Health, vs. H Megaera Rampage 5|
|Non-Tank Player Health||Tank Player Health|
|Tick #||Damage||No SLT||SLT 10%||Full SLT||No SLT||SLT 10%||Full SLT|
Some of the more obvious effects:
Without any raid cooldowns, non-tank players (e.g. the DPS and healers) would be in the “danger zone” after the 11th tick of Rampage, and would inevitably die on the 13th tick. The 10% damage reduction component of Spirit Link’s functionality pushes the “danger zone” back a full second, to the 12th tick, and definitive dying is pushed all the way back to the 16th tick. Once the health redistribution is factored in, the “danger zone” is pushed back a further second, to the 13th tick. This is just showing that DPS and healers leeching health from the tanks gives the entire raid a little while longer to organise its next cooldown, or for players to use their own cooldowns, and ensure their survival. (It also shows that – as hinted at above – a single cooldown really isn’t going to be enough to cover this fatal encounter mechanic.)
Tank players will never outright die to Rampage so long as the healing keeps up. However, under the full effect of Spirit Link, at every point in Rampage’s duration, tanks are on much lower health than they would have been if no cooldown had been used at all. “Holy cats, Dayani,” you might be thinking to yourself right now. “How can that possibly be a good thing?” And the answer is multi-faceted:
- Many tanks have damage reduction effects (passive or active) that mean that they’d rarely take a full-power Rampage tick to begin with, so this is really a “worst-case scenario” view of what happens to them;
- Tanks have far more personal survivability cooldowns than most non-tank classes or specs, and a much higher base health, so they can “afford” to sit on lower health;
- Regardless of their health level, the tanks aren’t going to survive if the healing stops. So sacrificing a bit of their health in order to keep the healers alive, so the healers can keep them alive longer, is still a net positive for them.
Obviously this analysis is simplistic, but frankly, I didn’t feel like writing a Raid Combat Simulator just to prove a point. I don’t even have the right tools for that! But overall, my point here is that Spirit Link has the ability to save those people in the most danger of dying, by forcing those players in the least danger of dying to sacrifice a little health. Spirit Link Totem pushes back the “danger zone” a bit longer, gives you breathing room to pump out more healing or start another cooldown before deaths occur, and also prolongs the raid’s average survivability.
It’d be more drastic yet if Rampage dealt uneven damage (which, in practice, it would, because some players have damage reduction effects or cooldowns and others don’t), or if players entered the Rampage not on full health – you’d see a much more dramatic effect on the time-to-“danger-zone” and time-to-inevitable-death metrics.
I also want to note that this works the other way, too. If the whole raid is taking damage, but the tanks are taking a significantly higher proportion of their health in damage than the non-tanks are (let’s say, um, the end of Heroic Warmaster Blackhorn, for example), Spirit Link sacrifices non-tank health in order to keep tanks alive. In these situations, the non-tanks aren’t going to survive anyway unless the tanks survive, so again, sacrificing your health to keep the tanks alive under these circumstances still works out to a net benefit to everyone. So long as you view Spirit Link not selfishly, but holistically, it’s plain to see that it can be a very powerful cooldown when used properly.
I did an analysis similar to the Spirit Link table above to determine how much survivability the other damage reduction raid CDs would lend to the DPS players in the raid. (The analysis was not very informative for tanks since in almost any case the tanks are in no danger of dying.)
Rather than print out a huge table with a bunch of numbers, like I did above to illustrate the point with Spirit Link, I’ll just summarise the results in list form:
- No CD: Players are in the “danger zone” for tick 11, die at tick 13
- AMZ (5.3): Players are in the “danger zone” for tick 13, die at tick 17
- AMZ (5.4): Players are in the “danger zone” for tick 14, die at tick 17
- Devotion Aura: Players are in the “danger zone” for tick 15, die at tick 17
- Power Word: Barrier: Players are in the “danger zone” for tick 19, never die
- Smoke Bomb: Players are in the “danger zone” for tick 13, die at tick 17
- Spirit Link Totem: Players are in the “danger zone” for tick 13, die at tick 16
As before, I didn’t analyse Demoralizing Banner for Megaera since I am not clear that it actually provides damage reduction from all attacks.
It’s no surprise that Power Word: Barrier is the best – I mean, you can look at the graphs above to determine that. The combination of strong damage reduction potential plus the extremely long duration of the cooldown makes it so powerful that players will not die to Heroic Megaera’s fifth Rampage even if no other raid cooldowns are used! (Although, players will need to use a personal CD or a Healthstone to survive the 19th tick.)
It’s similarly unsurprising that Spirit Link Totem brings an earlier death than the other effects. In this particular analysis, Spirit Link is not necessarily at its best performance – it would do better if personal damage reduction abilities were taken into account for the entire raid, and if the damage were less even. Regardless, my screed about Spirit Link above wasn’t intending to make it sound like “the best” cooldown – it certainly has its weaknesses and you can see one of them illustrated here – but simply to reinforce that the health redistribution does not have a zero value.
The newly rebalanced healing cooldowns are actually pretty well matched to their damage reduction counterparts. For example, Power Word: Barrier mitigates 7.35 million damage during 25-player Heroic Iron Qon’s Fist Smash #8, while Divine Hymn would deliver up to 8.09 million healing. Devotion Aura prevents 1.2 million damage during 10-player Heroic Megaera’s Rampage #5, while Tranquility would deliver 2.8 million healing, a fair portion of which (but surely less than half? I hope?) would be overheal.
It makes sense for healing cooldowns to produce larger numbers than damage reduction cooldowns do. Damage prevention is simply more powerful, so for the cooldowns to be on remotely equal footing, the healing ones need to deal more. Exactly how much more is impossible to quantify, although it’s something I could look into a little more closely in terms of making sure the outcomes from using a healing cooldown mirror those of using a damage-reduction cooldown (in terms of pushing back the “danger zone” and “inevitable death” events).
If I had to give my opinion on how cooldowns could be more balanced – and let me be clear, I do not think that they necessarily do need to be balanced, just based on the results of this analysis or even in general – I would suggest that Power Word: Barrier could be toned down a little. I might lower its duration a bit, to 6-7 seconds like most other cooldowns, and perhaps to compensate increase its radius a few yards to make it a little easier to use in spread-out fights. It is ridiculously weak in spread situations (as a Shaman, I feel a little bit of sisterly solidarity with poor Barrier), although I’m not sure how sympathetic I should be given that with a little bit of group-member-Tetris, Spirit Shell can still perform optimally in these situations.
I might also give DPS specs’ Healing Tide Totem the Divine Hymn/Tranquility treatment for 25-player mode, simply as a way to make it more attractive than Ancestral Guidance for DPS specs. With a longer CD, being on the GCD, and having a much lower healing throughput than Ancestral Guidance has, there are very few reasons for a DPS player to choose Healing Tide. (The best reasons I have are: for a fight where a lot of damage is dealt while enemies are not targetable; a fight where a lot of healing is needed whilst players are stunned; a fight where lots of healing is needed while the enemy has a damage reduction buff up.) I don’t think this is warranted for Restoration Shamans, though, since we can keep casting our healing spells while Healing Tide Totem does its thing.
Monks have very powerful personal utility abilities – Leg Sweep, Ring of Peace, Diffuse Magic or Dampen Harm, Paralysis, etc. – but their raid utility is a bit lacking. I’d like to see Zen Meditation reworked in some way, so that it can be used as more than just a personal survivability cooldown. I’m not sure exactly how – I’ll leave that up to the devs and the more creatively-inclined. Revival seems a little weak by comparison to the rest of the cooldowns, but I don’t think I’d touch it – the Monk toolkit is ridiculously high-throughput, and the bursty nature of this cooldown certainly slots well into that toolbox. If the healing were increased by too much, or if it were reworked to be a HoT effect, I think that – combined with its ability to dispel harmful debuffs – it would be hard to balance so that it wasn’t too strong.
As a final thought, I’d just like to expand a little on my don’t-stack-cooldowns philosophy. All of the dangerous situations in a raid occur when the damage pattern gets spiky and unpredictable. Raid cooldowns help to smooth the incoming damage out (either by reducing it outright, or by healing back up in mere moments after the damage is dealt, thereby stabilising the raid’s health), directly counteracting the spikiness and making the damage pattern more predictable. By chaining cooldowns back-to-back, players may be taking more damage at any one time than they would have taken with a pair of stacked cooldowns at the start, but the spikiness is vastly reduced. The damage is more predictable. Everyone feels more comfortable. It’s a good thing.
|Summary of Available Raid Cooldowns|
|Ability||Class/Spec||CD (min)||Duration (sec)||Effect|
|Tranquility||Druid (resto)||3||8||Channeled direct heal + HoT|
|Revival||Monk (MW)||3||–||Instant burst heal + dispels Magic, Poison, Disease|
|Light’s Hammer||Paladin||1||16||Ground-based AoE HoT (and damage)|
|Spirit Shell||Priest (Disc)||1||10||Converts targeted healing spells to absorption shields|
|Divine Hymn||Priest (Holy)||3||8||Channeled direct heal, +10% healing taken|
|Lightwell||Priest (Holy)||3||–||HoT on injured raid members; charge-based|
|Tranq (Symbiosis)||Priest (Shadow)||8||8||Weak channeled direct heal + HoT|
|Vampiric Embrace||Priest (Shadow)||3||15||Converts 50% of single-target damage to healing|
|3||10||Converts 100% of single-target damage to healing|
|Healing Tide Totem||Shaman||3||10||Delivers several pulses of direct healing|
|Ancestral Guidance||Shaman||2||10||Converts damage to healing & multiplies healing done|
|Ascendance||Shaman (Resto)||3||15||Copies all healing and divides evenly amongst nearby allies|
|Rallying Cry||Warrior||3||10||+20% current and maximum health; health removed when effect fades|
|Damage Reduction Cooldowns|
|Anti-Magic Zone||DK||2||3||40% magic reduction|
|Zen Meditation||Monk||3||8||Personal 90% magic damage reduction; redirects up to 5 harmful spells from party members|
|Devotion Aura||Paladin||3||6||20% magic damage reduction, 40-yd radius|
|Power Word: Barrier||Priest (Disc)||3||10||25% damage reduction, limited to small area of effect|
|Smoke Bomb||Rogue||3||5||20% damage reduction; very small radius; glyphable to 7 sec duration|
|Spirit Link Totem||Shaman (Resto)||3||6||10% damage reduction; redistributes raid health every sec|
|Demoralizing Banner||Warrior||3||15||10% reduction of damage dealt by enemies within 30 yds|