Monday, September 1, 2014

Discipline Tuning: The NeverEnding Story

You may have heard that there's an expansion in the works for Blizzard's flagship franchise, the World of Warcraft. Yessiree, they're doing another expansion, complete with flyable new areas, new races, new classes, and a whole wave of class modifications and tuning.

Tuning is a subject near and dear to the hearts of discipline priests because discipline is a spec (the only spec?) that has never been properly tuned. Oh, I'm not whining...lack of tuning has been as frequently pleasant as it has been vexing. Sometimes we're on top of the mountain, sometimes we're under it. But we're never actually among the intrepid group of pilgrims working their diligent way up Mount St. Healsmore.

Think I'm exaggerating? Let's check the tape, shall we? ("Shall we?" is always the right note of faux-outrage that informs the listener that possibly bitter soapboxing is on the way. I'm never bitter, though, so I'm just using it ironically.)

How did Mists of Pandaria treat discipline? Well, at launch, discipline was far and away the worst healer spec in the game, as their absorbs were very weak and their healing was very weak. Blizzard noticed. A month later. Incoming hotfixes changed things a bit. Instead of being the worst healer spec by a mile, discipline priests became the best healing spec...by a mile. Spirit Shell seemed to be the culprit, as it was powerful and, on a one-minute cooldown, it nicely fit the fact that seemingly every tier 14 fight was designed around the Big Bad Attack being on a one-minute timer. Or a two-minute timer. Or two Big Bad Attacks, each on two-minute timers, alternating evenly. Blizzard noticed. A couple of patches later. So for 5.2, Spirit Shell was nerfed...it had a shorter shield-building duration and mastery (which, for discipline, boosted the potency of absorbs) no longer affected Spirit Shell. Solved! Except that as time went on, and gear levels increased, an interesting thing happened: while Spirit Shell wasn't the meter-crushing force it used to be, discipline priests pretty much stopped worrying about healing and only focused on coating the raid in Divine Aegis. All the time, forever. There is no cooldown on Divine Aegis, after all. With the high crit levels possible through the final raid tier of the expansion, Atonement heals crit and indiscriminately put Divine Aegis on raid members non-stop. The level 90 tier of talents, Cascade, Halo and Divine Star, were uncapped area of effect healing spells...that means, it had equal effect no matter how many targets it hit (whereas most area of effect heals have diminishing returns once the number of targets hit exceeds six). It no longer even mattered if Cascade, Halo or Divine Star actually healed anyone...even if it hit a full-health raid, it was bound to put up lots of shields. Add Spirit Shell to that mix and suddenly discipline priests had the ability to lock other healers out of the precious, precious meters to a surprisingly large extent.

But Blizzard couldn't have known it would play out quite like that at the end of an expansion. Is what I'd be saying right now if the exact same thing hadn't happened in the previous expansion, Cataclysm. The mechanic was a little different...back ye olde days of Cataclysm, Prayer of Healing automatically applied Divine Aegis to all players healed by the spell equal to a percentage of the healing done, no critical strike necessary. With the huge mana regeneration available to healers by the end of the expansion, it was possible to just keep casting Prayer of Healing as the filler spell, even if the raid was full health, to build and maintain perpetual Divine Aegis on everyone. Whenever damage hit, it bounced off a nice, rubbery buffer of Divine Aegis.

But the similarities between Mists of Pandaria and Cataclysm didn't end there. Like Mists of Pandaria, discipline was pureblood crap at the start of the expansion, got hotfixed into godhood eventually, got panic-nerfed some time after that and then finished the expansion as I detailed above...the "nah, I don't think I'll be sharing the healing today" spec.

Discipline in Vanilla and Burning Crusade was not a viable raid healing spec, it was mostly used for PvP or ignored completely. Wrath of the Lich King was the first point at which the developers got serious about making discipline viable, so it began at the "non-viable" starting point and slowly climbed, over the course of the expansion, towards respectability. By the final tier of the expansion, Icecrown Citadel, well, discipline was certainly viable and occupied a valuable raid niche, but it's hard to say whether it was tuned properly because healing had become almost irrevelant. Mana was nearly infinite, every healer just spammed their strongest spell and you lost only if people didn't handle mechanics properly and were one-shot, essentially.

So very arguably discipline was properly tuned at that time, but I can't count that because healer tuning was impossible, in my view, to ascertain in that kind of environment. Outside of that single tier, discipline has been oscillating between entirely too good and entirely too weak. At some point, Blizzard is likely to just delete the spec from the game as an admission that they have no idea how to tune an absorb spec but, as Syrio Forel would say, not today.

Discipline has, from what I've heard, been having a rocky time of it on the beta because what else is new. Everyone, even other healers, agreed that at the start of the beta, discipline barely even counted as a healing spec. Tuning is slowly bringing discipline back towards the kind of thing you'd maybe consider looking at as a possibility to use for raid healing, depending.

I just thought I'd take this time to offer some thoughts on what I'd like to see for discipline going into Warlords of December November.


Atonement


Atonement got out of hand. Let's just admit that. In Cataclysm it was kinda-not-really optional. The Atonement itself was pretty optional...whether you healed small amounts of damage by heals or by DPS was pretty irrelevant. Follow your muse! However. However, not building Evangelism in order to pop wings was sub-optimal for your healing. But only a little sub-optimal. Using five global cooldowns, give or take, to build the stacks and using up talent points to get Atonement, Evangelism and Archangel were enough price that you could do your work pretty well ignoring the whole issue.

During Pandas In The Mist, though, it was 100% required. Penance joined the Atonement party and for a while, Penance was better if cast offensively. Holy Fire did strong healing via Atonement and when it morphed into Power Word: Solace via a talent, it was one of the best mana regeneration abilities in the game. Even if you didn't Smite, you could build up your Evangelism stacks just fine doing what you do.

Oh and did I mention? Smite wasn't bad at all. It wasn't amazing healing, you probably didn't want to spam it during high intensity healing periods, but it did respectable healing and, later in the expansion, the Divine Aegis that got stacked from spamming it was more than worth the mana it cost. Which was close to nothing.

So yes, if you were a discipline priest and not wearing out your Penance, Smite and Power Word: Solace buttons, you were, sadly, Doing It Wrong(tm). Pure Atonement wasn't the answer all the time, but it was a good answer an unfortunate amount of the time.

So it sounds, from reports, like Atonement is being nerfed. That's proper and right and all decent, moral people can get behind that. Only...it sounds like they're nerfing it into near uselessness. Uselessness for the actual healing those spells do...you still need to use Atonement to build Evangelism stacks to pop wings. That relationship hasn't changed and it seems like discipline priests are balanced around using Archangel a lot. So, discipline priests will still need to cast five Atonement spells per 30 seconds. But each of those casts might be wasted global cooldowns outside of the stacks you get, which means the whole mechanic/relationship will be pretty punitive. In order to get the thing you need to get, you must throw away quite a few global cooldowns. That seems like maybe a wild oscillation the other direction. You know, I've seen this movie before.

So what to do? What would I do? Well, you've come to the right place to hear what I would do. I would do this:

  1. Make offensive Penance strong, but not as strong as casting it "defensively" (into an ally)
  2. Make Holy Fire/Power Word: Solace strong
  3. Make Smite terrible, akin to what Heal is right now on live, which is terrible


Why is this the right approach? Because Penance and Holy Fire/Power Word: Solace (one becomes the other depending on how you talent) have cooldowns, so you can't spam them. Making them pretty good is no different than other healers having some solid but cooldown-constrained spells, which they do. Since Penance is better when cast defensively, the "right choice" will not always be to cast if offensively...in a mana-constrained world, you will often want maximal efficiency in keeping that awful tank alive, so getting the most healing by using the spell defensively will be something you'd want to do sometimes. Other times, when the tank is mostly stable, you'll be more willing to lose some healing-per-mana efficiency in order to split your Penance among three people (since each tick of the spell can heal a different person through Atonement) and get a stack of Evangelism.

And using the Atonement you can spam, Smite, during any period of tough healing will be like trying to put out a conflagration by standing over it and answering the call of nature (applicable to male or female). Ineffective, pitiful-looking and extremely painful. You can spam it, if you want, during low intensity healing, but it'll do low healing and low damage, so any Divine Aegis formed will be minor and the amount of damage you bring as utility will also be pretty minimal. It'll just be flavor...discipline's style of doing close to no healing for little mana.


Absorption


And now we come to the problem that has haunted developers when it comes to the healing side of the game ever since discipline became viable: how do you balance absorbs?

Here's the crux of the problem: unlike actual healing, absorbs reduce the incoming damage. Why is that important? Because you can't heal up death. If a player has 100,000 health and one healer can put a 40,000 absorb shield on that player before damage comes in while another healer can heal for 40,000 (or 50,000!) after the damage comes in, which do you think is more valuable when 110,000 damage hits that player?

That leads to this question. Do you, as a game developer, balance boss damage around having the wall of absorbs that a discipline priest (or two, or three) can offer, or you do balance boss damage around not having it? Either answer is wrong. If you do balance it around having the shields, then the damage is borderline unhealable if you don't have a discipline priest (or two, or three) because the damage will have to be nearly double strength since it's assumed to need to break through a buffer of absorbs first. If you balance the boss damage around not having the shields, then a discipline priest (or two, or three) can completely trivialize the mechanic.

To me, there's at least one good answer:

  1. Make discipline absorb spells extremely potent (I was tempted to stop here)
  2. Make discipline absorb spells extremely expensive--ruinously so, if cast a lot
  3. Give discipline priests adequate actual healing powers
  4. Avoid making every fight low-damage/high-damage alternating phases

(4) is surprisingly important. When it comes to meter winning (and actual effectiveness in winning boss encounters), a major benefit for discipline priests is getting to essentially pre-heal high-damage phases during low-damage phases. Discipline priests can use the final 10-20 seconds of a low-damage phase to start building their shield wall on the raid. When the high-damage phase hits, all of that prep time gets applied to the stressful phase and discipline priests can continue healing. The net effect is that they get 10-20 seconds of "extra healing time" in the phase that other healers do not.

Having that sort of design sometimes is fine...you can't completely remove one damage pattern if you want varied encounters. But every encounter can't play that way if you want to rein in the effectiveness of shields. Damage patterns like moderate damage at all times or longer stressful periods (with longer light damage recovery periods) can mitigate (pun intended) discipline's advantages in this area. The perfect fight design for discipline to exploit their strengths is a high damage period every minute that lasts for no longer than about 10-15 seconds. The further a fight's damage pattern is from that, the less advantageous for discipline.

As for (1) through (3), the idea is that if a discipline priest wants to primarily spend their mana on absorption, that's fine...but that's all they'll be able to do and they'll be paperweights for 75% of the fight due to being out of mana, so they'd better make their 25% activity time count. Alternatively, they can mostly heal like the rest of the unwashed masses and very rarely use absorption shields as emergency measures. Or they can find some pleasing mix in the middle. But they should be forced into hard choices...absorbs are great, but they shouldn't be the right choice all the time. Using your absorbs should cost you the ability to do other things.


In Conclusion


Actually, I don't have anything more to say. I thought I'd have something to tie this all together with a nice little bow, but what're you gonna do? Hope this was mildly interesting!