Duct Tape and a Prayer

There is a !meme going around the healing blogs. I have been instructed by Amber that I am to fill it out and post it here. Since I value the life of my Sinister Squashling, I obey:

• What is the name, class, and spec of your primary healer? Theande, Discipline Priest.
• What is your primary group healing environment? 25-player raids, though those are thin on the ground these days.
• What is your favorite healing spell for your class and why? Penance. It’s fast and has three chances to crit! Its only drawback is the extraordinarily long cooldown.
• What healing spell do you use least for your class and why? Desperate Prayer. The only reason I have it talented is because I had nowhere else I wanted to put the point.
• What do you feel is the biggest strength of your healing class and why? Mitigation. Discipline priests use a lot of shields! Preventing damage > healing it after the fact.
• What do you feel is the biggest weakness of your healing class and why? We don’t have any strong AOE heals, which limits our utility.
• In a 25 man raiding environment, what do you feel, in general, is the best healing assignment for you? Flex healing. Discipline priests do best when we’re going where we’re needed – we’re fast and light on our feet for a reason.
• What healing class do you enjoy healing with most and why? Probably holy priests, because we complement each other.
• What healing class do you enjoy healing with least and why? Resto druids. Their HOTs account for a lot of my overhealing.
• What is your worst habit as a healer? Using spare GCDs for DOTs instead of shielding…
• What is your biggest pet peeve in a group environment while healing? It’s a toss-up between tanks who drag whirlwinding mobs back into the casters and DPS who think the healer isn’t paying attention to the health meters.
• Do you feel that your class/spec is well balanced with other healers for PvE healing? Reasonably. It would feel more balanced if Blizzard included mitigation in the healing reporting.
• What tools do you use to evaluate your own performance as a healer? The logs posted by the raid leader afterward, mostly.
• What do you think is the biggest misconception people have about your healing class? Another toss-up, between “Disc is just for PVP” and “Disc is only good for tank healing”.
• What do you feel is the most difficult thing for new healers of your class to learn? That preventing someone’s health bar from going down is just as valuable as making it go back up.
• If someone were to try to evaluate your performance as a healer via recount, what sort of patterns would they see (i.e. lots of overhealing, low healing output, etc)? Moderate overhealing, lower healing than other healers because of all the time spent shielding.
• Haste or Crit and why? Haste. Discipline priests should be fast, getting heals to their targets almost as soon as the damage is taken – not slow and clunky and relying on big numbers and Divine Aegis.
• What healing class do you feel you understand least? Resto shamans.
• What add-ons or macros do you use, if any, to aid you in healing? I use Grid to keep track of health bars (although it’s falling out of favor with me) and FortExorcist to keep track of cooldowns.
• Do you strive primarily for balance between your healing stats, or do you stack some much higher than others, and why? I tend to stack Haste and Intellect; faster heals mean that I need more mana (and more regen). After that, spellpower, then crit, then anything else.

There, now my Squashling is safe.

I will not threaten their Sinister Squashlings (or any other pets), but those who must complete this or suffer the (kind of plush, actually) consequences are Shy and Tart.

The following is the current (datamined) Healing Priest Tier 10 4-piece set bonus:

• Your Flash Heal spell has a 15% chance to reset the cooldown on your Circle of Healing and Penance Spells.

With my current haste rating, Penance has a 1.64-second channel, leaving me 6.36 seconds to cast Flash Heals before Penance’s cooldown is up. (I’m glyphed and have Aspiration.) Since my Flash Heal’s cast time is 1.31 seconds, I can cast it 4 times before Penance cools down, with 1.12s left over.

The math for determining how likely a percentage-based proc will go off is 1-[(1-proc chance)^(number of opportunities)]. (It seems unnecessarily complicated, but that’s how probability works.) In this case our proc chance is 0.15 (15%), and our number of opportunities is 4.

If I’m just spamming Flash Heal until Penance cools down – which I rarely am, but let’s assume it for the sake of a best possible scenario – then my chance to reduce Penance’s cooldown by at least 1 second is 1-[(1-0.15)^4], or 47.8%. Given that, roughly every other Penance cooldown will be shortened by at least 1 second.

(Without going into the math, the increased number of casts from dropping Glyph of Penance and Aspiration don’t justify the extension of the cooldown.)

For the record, if you’re just spamming Flash Heal after you use Penance (which I hope you’re not) and assuming 0 Haste (I hope you have some), you have 15% chance to reduce Penance’s cooldown by 4.5 seconds, 27.8% to reduce the cooldown by at least 3 seconds, and 38.5% to reduce the cooldown by at least 1.5 seconds. (Remember, the proc doesn’t go off until the Flash Heal spellcast does, so you have to let Penance cool down by at least 3.5 seconds – 2 seconds channeling Penance and then 1.5 seconds casting Flash Heal – before you can trigger the bonus at all.)

Incidentally, due to the nature of the problem, there is no point at which you’re guaranteed a proc. You get to 99.9% chance of having triggered the bonus at some point at about 27 Flash Heal casts, but it’s possible to beat probability and go forever without having triggered the bonus proc. That said, with no Haste, on average you’ll reduce the cooldown of your Penances by about 0.6 seconds. (If you have Haste, the average reduction is [0.6 * (1 + Haste %)].)

At this point, it’s generally accepted that Discipline is an effective healing spec in PVE. It’s also generally – although less so – known that Discipline relies on shields and other mitigation to get much of its healing done. What doesn’t seem to be well-known is why Discipline priests love them some shields, and here I’m going to try to correct that. As always, correct me if you think my math is wrong or my conclusions aren’t sound.

Shields mitigate damage.

“Well, duh.” Sure, this one’s a gimme, but it’s also the most important aspect of shielding. Discipline priests aren’t the only ones who gain a benefit to their shields from spell power – the damage absorbed by your Power Word: Shield is increased by 80.68% of your spell power. But we also have a lot of ways to buff our shields. We get Improved Inner Fire in our second tier, which increases the contribution to spell power from Inner Fire by 45%; we have Improved Power Word: Shield as a tier-3 talent, which increases the damage absorbed by 15% after spell power is applied; we get Mental Agility in the fourth tier, which reduces the mana cost of our instant-cast spells by 10%; we get Soul Warding in the fifth tier, which reduces the mana cost of Power Word: Shield by a further 30%; and we get Borrowed Time in the 10th tier, which adds another 30% of our spell power to the amount absorbed. (Sadly, Focused Power doesn’t improve PW:S – thanks to Rilgon for clarifying.)

A Discipline priest with all of these talents, and 2000 spell power – which is about average for a beginning raiding priest, in my experience – can cast shields that absorb about 4580 damage for 560 mana, or about 9.25 damage absorbed per mana spent. (For reference, our Flash Heal is about 5.46 healing per mana.) That’s a powerful, efficient mitigation tool, especially with the Glyph of Power Word: Shield, which heals the target for 10% of the potential damage absorbed (that is, the full 4580, not the damage that’s actually absorbed), and is affected by Focused Power.

And that’s just PW:S. Divine Aegis, our other major shield, costs us nothing at all (and scales nicely with spell power; we get 1 damage absorbed for every 10 healing on every critical heal we cast); the only drawback is that we can’t control when it goes up. Our third mitigation ability, Pain Suppression, is less often used, as it has a long cooldown and is reduction in damage rather than damage blocking like the shields.

Our shields don’t just mitigate for the target.

Every time a Discipline priest casts Power Word: Shield, the entire raid takes 3% less damage for 20 seconds. 3% damage mitigation to everybody in the raid, on top of the 4580 absorption for the actual target, can be the difference between a wipe and a successful boss kill. Sadly, Renewed Hope doesn’t stack, but at the least it means that PW:S should be used once every 20 seconds. However…

Our shields are fast.

The other key ingredient of Soul Warding, mentioned above, is that it reduces the cooldown on our Power Word: Shield from 4 seconds to 1 second – essentially, we can cast it as often as we want, since the global cooldown (normally 1.5 seconds) is always going to be at least as long as the Power Word: Shield cooldown. We don’t have to worry about whether we’re wasting a cooldown on the DPS who’s in the Slag Pot just in case the tank needs it in three seconds, because we can toss out shields as fast as the GCD will let us. Speaking of which…

Our shields make us faster.

Borrowed Time also increases our spell haste by 25% for 6 seconds or until the next non-instant spellcast. (Channeled spells, like Divine Hymn and Penance, count as instant-cast.) That improves not only the speed at which we can cast our spells, but our global cooldown as well. With Enlightenment (another 6% spell haste) and some minimal +haste gearing, we can get off a PW:S – Prayer of Mending – Penance – Flash Heal combination in 4 seconds, a burst of healing that can bring a tank from 1% very nearly back to viable health. (With proper macros or add-ons, all four spells can also go to different people, making us decent raid healers as well.)

Our shields restore your energy.

Speaking of restoring tanks, Rapture, in tier 8 of our talent tree, restores 2.5% of our total mana when one of our shields is dispelled or completely spent (but not when it wears off without being used up), and restores 2% total mana, 8 rage, 16 energy, or 32 runic power to the shield’s target. That restoration alone is a reason for us to want shields up on as many players as possible (even though the effect has a hidden cooldown), especially in AOE fights like Loatheb, XT-002, and Ignis.

Our shields, in short, are awesome.

Any questions or comments? Have I missed anything? Let me know in the comments!

Seriously, you’re going to hurt yourself.

I have three feeds from WOW Insider going: The Queue, The Daily Quest, and anything tagged “priest” (that last one can get really annoying when the authors/editors of, say, All The World’s A Stage decide to spam every possible tag they can on So You Want To Be A Herbalist). Today, in the Priest feed, the Forum Post of the Day came through. It was about Discipline priests in raids, and how a raid leader was singling out a Discipline priest because his healing on the meters was under par. Amanda’s final comment really rang true: “It’s not about being a star, or a prima dona [sic], raiding is about working together as a team to accomplish a common goal.”

This is going to come as news to a lot of raid leaders (go on, raise your hand if you are one), but really and truly, Ulduar is not Galaga. Galaga is about getting the highest possible score; that’s the whole point of the game. Raid leaders tend to treat raiding like it works the same way – they see the healing and DPS meters as high score charts, and penalize raiders who don’t get onto the charts. The problem with that is that unlike Galaga, World of Warcraft raiding actually has a very concrete win condition: did you defeat the boss? If you defeated the boss, the entire raid wins. If not, the entire raid fails.

If the raid isn’t failing – if you’re successfully downing the bosses you want to take down – then there’s no reason at all to single anyone out on the meters. You met the win condition! Clearly everyone is doing at least as well as they need to be, or else you wouldn’t be successful.

If the raid is failing, everybody needs to step up a little more. It’s really tempting to go through the high scores – sorry, the meters – and identify the people who aren’t scoring as high – sorry, doing as much damage or healing – and try to “fix” those people. (In the example in Amanda’s post, the RL asked the priest to go Holy to increase his HPS.) Unfortunately, that simply won’t solve the problem. Unless you’re being actively sabotaged or someone’s doing something incredibly dumb, fixing a single person isn’t going to make the raid go from not killing the boss to killing the boss. When a raid fails, nobody is exempt. If the raid isn’t doing enough DPS but the healing is okay, the healers can look at ways to keep the raid alive a little longer after the enrage timer hits, or to damage the boss between heals. If the raid is doing enough DPS but important players are dying, the DPS can work on staying out of the fire or taking the adds down a little faster. I don’t mean that the DPS is exempt from improving in the first scenario or the healers are exempt in the second; I mean that everybody shares the responsibility of a failed attempt.

(If you are being actively sabotaged or someone’s doing something incredibly dumb, the meters aren’t going to help you anyway.)

I’m not saying that meters are useless. The best – perhaps the only valid – use of meters is to check your own performance in the raid (and even then it’s not necessarily accurate, like the Discipline priest example above, which is why WOW Web Stats is handy), and to look at other raid members to see what you could be doing differently. Raid leaders, though, should be paying attention not at individual raiders – because I will guarantee that there isn’t a single player in the raid who can’t improve something – but at the statistics of the raid as a whole. It’s obvious, but damaging to the egos of the other raiders, that if you think a single person is holding you back enough that you can identify them as the cause of your wipes, it’s actually the whole rest of the raid that needs work.

Click the image for full-sized heuristics!

(The font used is Titillium, an open-source Futurist sans-serif typeface.)

Khaeli has a couple of good posts up on Shadow Weaving: Gearing your Discipline Priest in 3.1, where she talks about the theory behind gear choices, and Khaeli’s (Ulduar) Gear Wishlist, where she lists the actual gear she’s looking at. Along the way, she pointed out the Tier 8 healing priest set bonus (I hadn’t even looked at T8 yet); the 2-set is nice for Discipline priests, but the 4-set is fabulous for this style of healing, since not only will PW:S give us haste, it’ll give us extra oomph behind our healing spells as well.

Some comments on comments:

tom jones: Glyph of PW:S is definitely useful, but it’s actually not as geared to this play style as it looks like it ought to be. The trick is that we’re using PW:S largely for mitigation and for the speed bonus; a lot of the time, we’re going to be shielding people who aren’t actually taking damage, so the glyph isn’t going to be useful. However, when we do shield someone who’s taking damage, that 20% healing is going to be handy. As for PW:S and DA being rage sinks, they’re not anymore; at least according to some early patch notes, damage mitigated by magical shields now counts for the purposes of rage generation. (Original comment)

Tsark: I’d actually forgotten about Enlightenment when I started writing that post! The low amount of Haste you actually need, between Borrowed Time and Enlightenment, to reach the 33% soft cap means that you can easily get it from enchantments and consumables (more rigorous math over on Khaeli’s blog, linked above, leads to a figure of 77 Haste Rating), leaving you free to make gear choices based on other criteria like Crit Rating and Intellect. I’m not sure that I’d recommend getting haste gear at all – certainly not more than one piece – given how easy it is to hit 33% with this play style. (Original comment)

caladein: I haven’t been able to make it past Flame Leviathan so far (my guild raids on Wednesday, Sunday, and Monday, and on Wednesday night we were plagued by server crashes), but from what I’m hearing, you’re right on the money with your comments about Mimiron and Kologarn. I’m hoping that the pattern stays the same throughout Ulduar, in terms of making Discipline priests useful like this. (Original comment)

Let’s start from the beginning. (I’m told that it’s a very good place to start.)

Right now, the conventional wisdom is that Discipline priests are for single-target healing and should stack crit (for Divine Aegis and big heals), and that Holy priests are for multiple-target healing (and should stack Spirit for the spellpower bonus and mana regen). I’ll say up front that this is a perfectly valid way to play the class. I don’t think that either of these assessments are untrue; Discipline priests are strong single-target healers, and Holy priests are strong multiple-target healers.

However.

Just because Discipline priests can be strong single-target healers doesn’t mean they can only be strong single-target healers. WOW itself has an excellent example of this kind of dichotomy: just because Feral druids can tank doesn’t mean that’s all they’re good for. I see another approach to Discipline healing, which focuses less on talents like Divine Aegis and more on talents like Borrowed Time and Renewed Hope.

Think about the distinction between fast and slow weapons in WOW for a moment. At a given level of DPS, slow weapons hit harder but less often; fast weapons hit far more often, but for lower amounts. With proc-per-minute weapon buffs, a fast weapon has a much lower chance per strike to activate its effect than a slow weapon does, but since the fast weapon is hitting more often, it evens out. Each has its advantages and disadvantages.

So it is with Discipline healing. The conventional-wisdom method is equivalent to a slow weapon: it doesn’t cast very fast, but it heals for a lot. However, talents like Borrowed Time, Improved Power Word: Shield, and Renewed Hope, and the Glyph of Power Word: Shield, make a fast-but-weaker option both viable and desirable. Instead of stacking crit like CW Disc priests do to increase their chances of getting Divine Aegis off, fast-but-weak Disc priests stack haste. These priests don’t heal for as much as their crit-heavy brothers, but they are much more agile in any environment, and specialize in hit-and-run healing.

Note that this is a question of approach, of philosophy, not of gear. It’s actually improbably easy to reach the soft haste cap with this philosophy; assuming that you have Borrowed Time and Enlightenment, it only takes 66 Haste Rating to reach the 33% haste cap after you’ve cast Power Word: Shield. This cap is based on reducing the 1.5s global cooldown to 1.0s, which is as far as it will go; at 33% haste, multiply your spell’s cast time by 0.67 (1-0.33) to get the actual cast time. This reduces the cast time of Flash Heal to 1s, the channel time of Penance to 1.3s (a tick goes off at 0s, 0.67s, and 1.34s), and the cast times of Greater Heal and Prayer of Healing to 2s. Further Haste Rating will increase your haste (and lower your casting time) even more, and it’s possible to reap some advantage from this, but more haste has a diminishing effect (because of the

The idea behind this approach, therefore, is that you’ll be casting Power Word: Shield at just about every opportunity. This will both ensure that Renewed Hope stays up on your targets (both the -3% damage buff and the +4% crit chance) and allow you to cast your heals as quickly as possible. You’ll also do a lot of mitigation of damage, and since you’ll be keeping the Weakened Soul debuff up on a lot of players at once, you’ll also have an increased chance to proc Divine Aegis when you do have to heal someone.

There are, of course, drawbacks to this approach. The first and most obvious is that it goes through a lot of mana. This can be mitigated by a high Intellect and mana regen rate, and even without those advantages, I had little trouble in 10-man fights when I tested this approach last night. Another objection at first glance is that by by relying on casting Power Word: Shield before you cast your heal, you’re effectively adding to your cast time (the half-second shaved off your Flash Heal is offset by the 1-second GCD from PW:S). My only response to this is that it’s better to be casting PW:S almost constantly. By doing that, you’re keeping yourself in Borrowed Time for when you do need to heal somebody.

I admit that this is an approach that feels strange to a lot of healers, myself included; spell haste is for caster DPS, and PW:S is for emergencies! I think that’s a Burning-Crusade, Holy-style healing mindset, though, and getting used to Discipline as a healing spec means getting rid of some of our preconceptions about what it means to heal as a priest. We can start with allowing for the possibility that PW:S and our caster stats can be used differently than we’re used to.

(Yeah, I’m kind of bad at conclusions.)

After some testing, I can say that Divine Aegis works thusly:

• If you cast a critical heal on a target, they gain the Divine Aegis effect, which lasts for 12 seconds and absorbs 10/20/30% of the total (not effective!) heal amount. (I was using Flash Heal to test, so I was getting 1.5-2k DAs.)
• If you or another Disc priest casts another critical heal on the same target, the Divine Aegis effect refreshes its duration (back up to 12 seconds), and 10/20/30% of the second critical heal is added to the remaining shield from the first crit heal. For example, I establish a 1.5k DA on the Tank, the Tank takes 1000 damage, and then I cast another crit heal on the Tank for another 2k shielding, which makes the total shielding 3.5k and the remaining shielding 2.5k (since the shield absorbed 1k damage between the first and second crit heal).
• The maximum damage that a single DA can shield is the level of the target * 125. That’s 10k damage for a level 80 tank. If a DA reaches this amount through repeated application, and you cast further crit heals on the target, the DA’s duration will be refreshed but no additional shielding will be added.
• When the DA buff wears off, the next crit heal starts the process over again.
• You can have Divine Aegis going on more than one target at once. Each shield has its own independent shielding total and duration.

I hope that’s helpful!

Well, I finally did the number crunching for +haste vs. +crit. The results turned out as I expected them to – although not as strongly toward +haste as I expected them to, which was kind of surprising.

Going as far +haste as you can, you can get 1095 haste rating from gear, gems, enchants, and buffs. You can get about the same amount of +crit. However, 32.79 Haste Rating translates to 1% haste; 45.91 Spell Critical Rating translates to 1% crit. That difference is crucial.

With 1095 Haste Rating, you get about 33.39% +haste. Essentially, that means that in the space where you could cast 100 spells without haste, you can cast 133.39 spells with that much haste. (I’ll use Flash Heal as the baseline, since its casting time is the same as the Global Cooldown.) With about 1100 SCR, you get about 23.95% +crit. Without any +crit gear, I have 17.95% +crit for holy spells on my priest, so 1100 SCR would put me around 41.9% +crit for holy spells (36.9% for non-holy spells).

So in the time it takes me to cast 100 spells with my +crit gear (150 seconds for Flash Heal), I can cast 133.39 (call it 133 for neatness’s sake, and note that I’m rounding down) spells with my +haste gear. At 41.9% crit with +crit gear, 41.9 of those 100 spells will crit (call it 42); at 17.95% crit with +haste gear, 23.9 of those 133 hasted spells will crit (call it 24). Incidentally – or not, if you’re Matt – that’s 42 Divine Aegis procs for +crit gear, and 24 for +haste gear.

I’ll take a moment here to define H as the amount that a normal Flash Heal heals for.

With the +crit gear, that’s 58 casts that don’t crit, and 42 that do. Since critical heals give 150% normal healing, over 150 seconds, my +crit gear gives me 58H + 48(1.5H), or 121H.

With the +haste gear, I have 109 casts that don’t crit, and 24 that do. Over 150 seconds, I get 109H + 24(1.5H), or 145H.

In other words, with the gear that’s available on the live servers, stacking +haste provides 20% more throughput over a similar length of time than stacking +crit.

Oddly, the trick I talked about a few posts back – interweaving Power Word: Shield and fast heals to get the benefit of Borrowed Time – has sharply diminishing returns if your +haste is above 820, because the global cooldown can’t be reduced below 1 second. In other words, you can’t get more than 150 spellcasts into 150 seconds. In fact, from the point of view of only throughput, it’s better to not rely on Power Word: Shield at all, and just stack +haste and throw out Flash Heal and Penance as fast as you can. However, that’s not taking into account the mitigation of Power Word: Shield… which I’ll cover in another post.

I can’t decide whether or not it’s odd that one of the reasons I want to learn to draw is that I want to be able to draw WOW fan-art.

Matt mentioned, in response to my post on Discipline priest philosophy a few days back, that he was still divided on the issue because he “likes crit for the RNG procs like Divine Aegis“. The problem with stacking crit for Divine Aegis, at least, is that DA only shields for the amount that’s actually healed. If your heal lands on someone who’s already at full health, and it crits, the Divine Aegis that pops up will absorb zero damage. This is, obviously, less than helpful.

Also, I haven’t done the math to find out how much crit you can get on gear, but I feel like you’re actually more likely to get a critical heal with multiple fast heals than you are with fewer slow heals at a higher crit percentage. I’ll have to look into that tomorrow.