Wednesday, February 18, 2009

Blizz healing spell design philosophy.


New theories incoming, this time about heals.

Basically, I believe all healing spells fit more or less into these three categories:

Baseline aka Bread & Butter: Spells that have a clear defined purpose and are straight forward.

circumstantial: Spells that are better than the bread and butter spells for various special purposes. Most group heals fit in this category.

Cooldowns: Spells that are so powerful and good that Blizz gives them a cooldown to prevent only using them.

Blizzard tries to make every healer use as much as possible of their reportoire. Some spells are much better fit for either PVE or PVP, but they often try to buff the spell for the least used aspect. The baseline spells will be almost garantueed to always have a use. The cooldowns can pretty safely be buffed straight, as the cooldown will limit just how powerful they become.

For the circumstantial spells, though, they follow a different strategy. It can potentially be very dangerous to straight buff a circumstantial spell (as in more healing throughput without adding to mana cost), because you might end up it being completely superior to other tools at disposal for the same purposes. That's what happened to Circle of Healing. They can't balance it by giving it a high mana cost, because the players still will use the spell and then complain about mana. The way they end up buffing these spells are usually by adding secondary effects to them, or give them additional situations in which they might be useful. Adding another situation where the spell shines makes it less circumstantial, as the circuimstances in question show up more often.

Let's take an example, Binding Heal. When it was first introduced in Burning Crusade, pretty much every PVE priest shunned it. While it could outshine Flash Heal in throughput (healing per second), it had no significant advantage in healing per mana (especially compared to Greater Heal) and risked overhealing twice. In serious arena, it was much more common for the enemy team to focus one player rather than split the damage, and it offered no additional benefit to focused damage. Of course, there were occassions where it would shine, but it was very circumstantial. Blizz couldn't straight buff the throughput, as that would've made it put out more single-target healing than Flash Heal. They tried lowering the mana cost, which helped both in PVP and PVE. But since it still was too circumstantial in PVE, they gave it very low threat. That helped. Suddenly priests used it in heroic on threat-sensitive fights, or early in pulls before the tank had proper threat on all mobs.

Fast forward to WotLK. Blizz is talking about changing how healing works. At the same time, they can't mess up too much the existing healing style, as there is no way to garantuee that a new healing style will be better than the old. Read: More entertaining. It's hard to please most of the crowd, impossible to please all.

From the 3.1 preview, it said that holy priests now will have a faster Greater Heal after casting Flash Heal. Fairly unoriginal as it's basically the same idea that they used with both Discipline and shaman Resto end-talents, but it works. It changes Flash Heal so that instead of being a pure bread'n'butter spell (discounting Surge of Light procs), it will also be a circumstantial spell. At the moment, you get the highest throughput of healing as a holy priest by spamming Greater Heal. In 3.1, you will do better alternating Flash Heal and Greater Heal, perhaps Greater Heal several times if a single flash heal will give a stack of buffs (like Tidal Waves gives two faster Lesser Healing Waves/Healing Waves, not only one). In addition, you can also throw in a flash heal early so that you'll already have the buff ready when you need it.

What they do is give spells additional circuimstances in which they'll be better than (other) baseline spells. I offer two observations to Blizz approach in making spell more interesting.

First observation: By adding circuimstances such as the above, Blizz tries to reward planning and foresight. While combat in WoW always will have random elements, most of what happens can be anticipated. At the very least, it's easy to anticipate that the tank will take damage. Then you might also anticipate that he or she's going to take spikes, and that the rest of the group will take AoE damage from the spell the boss mob is just casting. While I think it is a good thing to reward anticipation, it leads to one of two problems. First case, fights might be tuned so hard that the healer is required to be very familiar with the fight in order to keep everyone alive. Second case, fights will be tuned so that it still is possible to improvise and catch up, in which case they might be too trivial for a healer familiar with it. Of the two evils, the second is the lesser and the most likely Blizz will follow.

Second observation: There might be problems at the horizon if you keep adding stuff to existing spells. Specifically, you will "need" to press your cooldown-heals at every possible cooldown to be an optimal healer. That, in turn, makes the healing game feels very spammy, and less dependant on good decisionmaking. Let's take an example, Riptide. For the moment, this spell has all of the following functions:

  • A direct instant heal.

  • A HoT that lasts 15 seconds.

  • A buff to your next Chain Heal with the Riptide target as a primary target which removes the HoT.

  • A self-buff to the caster, Tidal Waves, reducing the casttime of the next two combinations of Lesser Healing Wave and Healing Wave.

  • A chance to proc Earthliving Weapon, a second HoT that lasts X seconds.

  • A chance to crit, thus proccing Ancestral Healing.

  • A chance to crit, thus proccing Improved Water Shield.

  • A chance to crit, thus proccing Ancestral Awakening.



Pretty hefty list for a single spell. Why is it so long? Blizz has more or less stated that they considered it a problem that every resto shaman in a raid did nothing but spam Chain Heal all day. They wanted to buff Riptide so that it was worth the GCD to cast it for a resto shaman. In the process of buffing it (which also included buffing it's direct throughput), they added a lot of circumstantial reasons to use it. Now it is almost at the point where I feel that I need to cast it as often as I can while healing 5-man heroics, provided anyone in the group have taken damage. Holy Shock is very similar in the regard that it also has a lot of circumstantial effects attached to it.

The bottom line, I think we'll see even more of the circumstantial stuff for healers in the future. Consider the healer abilities in the Aces High daily, same as the last phase of Malygos.

(By the way, sorry for the lack of posts. Like WoW, a blog's priority is well behind both job and private life)

Thursday, February 5, 2009

The healery arts of bursting


Pretty much every healer have the potential to put out some burst damage on demand. There will be moments in both PVP and PVE where this is handy. Mostly in PVP, of course, where the ability to burst down an enemy can be viewed as healing saved. But not exclusively, there will be times in PVE when bursting is handy. For example if you pull aggro on some add and it is mostly dead already or have low health, or on bosses which have vulnerability periods in which every point of damage counts. Take the Headless Horseman, to take an example a lot have experienced.

The main difference between bursting in PVP and PVE is the risk of getting counterspelled/kicked and thus getting that school of magic locked out. This is only a serious problem on your healing school (holy or nature) and only on spells with a cast-time. Actually, it's often an advantage to get your non-healing spell school locked, because whoever blew that cooldown can't immidiately counterspell you again. In both PVP and PVE you risk getting your casttime extended by damage taken for non-instants.

For all of the bursty combinations, you want to maximize the damage output for the given amount of time. The question is, how much time do you have? You can make several combos depending on how much time you have at your disposal. There's also an advantage to put an instant at the end of a chain, because it hits the same moment as the cast-time right before it, and the global cooldown triggered doesn't "count" towards the time the combo takes. In PVP, many enemies will hit defensive cooldowns if they take incoming serious damage. By making the damage hit at the exact same time, they won't have time to react between the nukes.

I ignore spells with long cooldowns. Though you can get a nice burst as a shaman by for example using Nature's Swiftness and Chain Lightning, you usually want to save those for other purposes.

Priest


A quick and safe 1.5 sec combo is Mind Blast -> Shadow Word: Death. If you don't risk getting counterspelled, Holy Fire -> Shadow Word: Death will probably yield better damage for a holy or disc-specced priest, at a lower mana cost.

You can of course drop both for a twice as long combo, Holy Fire -> Mind Blast -> Shadow Word: Death. I like using Holy Fire first because it packs a dot, slightly longer cooldown and better damage if you can't finish the combo.

If you're discipline, you also have the option of dropping Penance into the fray, of course based on the assumption that you don't immidiately need it for healing. If you're holy, pray for a crit and drop an instant Smite in there.

All in all, priests got pretty good burst potential, and at a neat 30 yard range. Use it well.

Druid


Druids are lacking in the bursting department. First and foremost, the problem is that to deal any kind of damage (except Thorns) you need to shift out of tree form, which means that the mana cost of the combo is sort of increased by the mana cost of shifting back to tree form.

The next problem is their spell selection. I don't assume any healer specs Starlight Wrath. Starfire deals decent damage and doesn't include any risk of counterspelling, but at a 3.5 sec casttime. Wrath deals pretty low dps without serious talent investment into the balance tree, risks counterspelling, and still has a 2 sec cast time. Moonfire deals very abysmal damage compared to it's mana cost when only looking at the instant part.

There's always the option of Moonfire spam, which has the advantage of being only instants. However, mana cost and low dps doesn't make it a too powerful burst.

The only option they have is to chain Starfire -> Moonfire or Wrath -> Moonfire depending on how long time they have.

All in all, druids are seriously lacking burst potential, in my opinion.

Shaman


Shammies have some interesting options. First and foremost, Lava Burst is a garantueed crit as long as the enemy has Flame Shock on them. Lava burst deals better damage at a significantly lower mana cost than Chain Lightning, even without Flame Shock present. In addition it's handy that it's not from the nature (healing) school, in case of counterspells. Like Lightning Bolt it has a travel time, which can be both a blessing and a curse.

The special thing about shocks is that they share the same cooldown. In PVP, if you *need* your enemy snared (Frost Shock) or the shock cooldown ready to interrupt a spell (Earth Shock or Wind Shock), you won't have the option of chaining Flame Shock into Lave Burst. In PVE, you might often be better of using Wind Shock to reduce aggro rather than use it for a nuke. There's also the problem with range, 20 yards (barring gear bonuses) rather than 30 yards on cast-time nukes.

All in all, there's a lot of different combos for different needs.

First, let's look at Flame Shock -> Lava Burst. Since you have to wait for global cooldown after the first, the combo takes 3.5 sec, barring haste and interruptions. While Lava Burst is a garantueed crit, without Elemental Fury it's only a 150% crit, and the initial hit from flame shock is nothing great. In effect, while handy, this isn't a superior combo. The redeeming trait is that you only need to stand still to cast in 2 of those 3.5 seconds. Glyph of Shocking reduces the time of this combo by 0.5 sec.

If you spend 4 seconds, you can chain Lava Burst -> Chain Lightning -> Earth Shock. Frost Shock at the end deals a bit less damage, but snares if you need that should the combo fail to kill. The advantage of this combo is that Lava Burst have travelling time while Chain Lightning and Earth Shock hit instantly, meaning they can hit in very short amount of time (<0.5 sec), giving an enemy player very little time to react, provided you're at max shock range. If you're in melee, this isn't such a good idea. Of course, there's the risk of getting Chain Lightning counterspelled. Just Lava Burst -> Earth Shock is a quick and safe combo, in the sense that you can't get nature counterspelled.

Given the nature of Lava Burst, the best option for bursting is setting up a killer combo by casting Flame Shock earlier, then load up a double or triple nuke so that you get a second shock in. This is easier said than done in PVP, slightly less so in PVE.

Paladin


Since pallies lack any casttimespells, their nukes are limited only to the instants they have. The interesting part is that Judgement of Justice (or any of the other judgements) is off the global cooldown, meaning you can that and another instant in at the same moment. Seal of Righteousness is probably the seal that causes the best damage on jugdement for a holy pally. Enlightened Judgements give judgement a lot of range, in fact more so than most other offensive spells in the game.

The primary nuke is Holy Shock, which has a few disadvantages. Only 20 yard range, for a start, and the fact that you might need it for healing rather than damaging. The bright side is that you probably have a fairly good crit rate on this, and using it with Divine Favor still grants the instant Flash of Light when used offensively, thanks to Infusion of Light.

Chaining Holy Shock and a judgement gives a fairly good nuke, and is completely instant. From 3.1 you can also throw in Exorcism, which will then be usable on all targets rather than just demons and undeads. See here for more details if you missed the change.

Last words


I believe Blizz doesn't dare to give druids a baseline high-damage, low cooldown ability like Holy Fire, Mind Blast and Lava Burst because they're afraid that it might make druids overpowered in PVP (again).