Friday, March 30, 2012

Instant Crit Versus Long Tail Mastery

The orthodoxy for holy priests is that, other than spirit, mastery is the best secondary stat to gear for. Haste is also solid (many suggest stacking at least to 12.5% raid-buffed for the first healing-over-time breakpoint--where Renew and Divine Hymn attain an extra tick per cast) and critical strike is shunned. Shunned!

I'm not going to discuss haste in this post, because haste is a bit of a different's essentially always good for reducing GCDs or cast times and becomes even better if you're approaching a haste breakpoint, but it comes at a cost--to realize most of the throughput gains, you have to fit in more casts, which means more mana burned. Haste is pure throughput at the cost of mana sustainability. (Extra healing-over-time ticks are an exception, but holy priests aren't nearly as HoT-centric as, say, restoration druids.)

What I want to discuss is mastery and critical strike...the two stats that provide throughput at no sustainability cost (i.e. free healing).

What's also interesting about this question is that I'm comparing the two stats that most holy priests would say have a wide gap between them and a clear "right answer" for which is better: mastery. And yet, I'm going to argue the reverse, that critical strike is the correct option for holy priests in most cases.

Front-Loading And Back-Loading

In a vacuum, holy's mastery mechanic--Echo of Light--actually performs quite well. Due to the lack of talents that leverage critical strike (unlike discipline, which has the Divine Aegis mechanic), mastery tends to perform better than critical strike when theory-crafted.

The problem arises when you dig into how Echo of Light works. If you cast a single spell, Echo is rather straight-forward...the target of the healing spell receives a HoT that heals for a certain percentage of the original heal over six seconds.

This gets complicated as you cast more and more spells. If you already have an Echo of Light HoT ticking on a player and heal that player again, more Echo of Light has to be added to the player. Thankfully, the new Echo doesn't clip or overwrite the old Echo, the way the Prayer of Healing glyph HoT does...that would neuter the mechanic terribly. How Blizzard made this work is interesting: it adds up the healing remaining on the old Echo (some of it has already ticked away), combines it with the healing that the new Echo would provide and creates a brand new six second HoT with all of that healing. It's clean and elegant...none of the Echo healing is lost, you just have a bigger and stronger HoT.

The problem, however, is that this serves to increasingly back-load the healing that Echo of Light is providing...that is, it pushes more and more of the healing to the future. With each heal of yours that lands, you're building a stronger and stronger HoT. Chain-casting to get through a rough patch of raid healing with leave you with a super-charged HoT on anyone who received a number of your heals.

What's wrong with that? The problem with that is that the more back-loaded your healing becomes, the more likely it is that it'll get pushed to overhealing by the healing-over-time from other healers or the completions of casts in progress. Now you might fairly point out that if it weren't back-loaded, then you'd be pushing other healers' healing into overhealing and, assuming we don't care about meters performance (and we shouldn't), what's the difference?

That response would be right on point if we assumed that the same casts would be made in either case. However, healers adjust their spell choice based on the state of life bars. If all that Echo of Light healing were delivered right up front, instantly, the other healers would see the effect of it before they made choices to drop another Healing Rain or cast a Holy Radiance or begin another Prayer of Healing cast.

In other words, if the effects of your Echo changed life bars quickly, it might save your fellow healers some casts, thereby saving them some mana to be used at another time. Less total overhealing, rather than overhealing redistributed. Instead, since your Echo's real effect is still to come, they cast all those spells, largely topping everyone off and then your Echo roars in to overheal all those people.

Now, what if we had a similar mechanic, whereby each of our heals added extra healing...but that healing was reflected instantly in life bars, thereby allowing your fellow healers to adjust their casting decisions accordingly? Wouldn't that be nice?

It would, and that mechanic is called critical strike. Now, you may have heard some scurrilous rumors about critical strike, that it's unreliable and RNG...but those fears are largely overblown. And while I shan't rehash my entire post on that subject, I will re-affirm that critical strike is quite a different animal for a holy priest, especially a holy priest who is raid healing.

A raid healing holy priest is going to make heavy use of three spells primarily: Prayer of Healing, Circle of Healing and Holy Word: Sanctuary. She may or may not also make heavy use of Renew. (This is not to say such a priest would not use other spells, but those spells would do the majority of the heavy lifting.)

The common thread between all of those spells is lots of "hits" per cast. Prayer of Healing hits five targets per cast. Circle of Healing hits five targets per cast (six if glyphed). Renew ticks four times, or five times if you're at the quite attainable haste breakpoint of 12.5%. Holy Word: Sanctuary is variable, since it depends on how many people are standing in it, but assuming you choose to use it when a lot of people can be inside it, it hits a lot of people and it ticks nine times (base, without the four-piece tier 13 bonus).

When you have that many hits per cast, you're going to get a lot of critical strikes. Unlike single-target healers, where you're likely to get nothing on any individual cast, with these sorts of spells, you're quite likely to get at least one critical strike per cast, if not more. This, in effect, acts as a similar effect to mastery (I'm not speaking mathematically, here, but rather conceptually); each of your casts in effect gets some extra healing. Not RNG healing, mind you, but a pretty predictable return since each cast has enough chances for "mini-crits" that you are very close to guaranteed to get some.

The difference, of course, is that this return will be front-loaded. It happens as soon as you make the cast, or as soon as the tick happens (in the case of Renew or Holy Word: Sanctuary). You see the healing immediately and so do your fellow healers. It won't be overhealing for you (assuming you cast when there was damage to be healed) and it hopefully won't lead to overhealing for your other healers if they see and adjust.

Now, can Echo of Light still do yeoman's work? fights where your healing team is really struggling to stay abreast of damage the entire fight, and therefore everyone is not eventually topped off, then your Echo of Light will be able to tick fully and mastery does have a better conversion rate from rating to throughput gain. In cases where you are never really topping off your raid, mastery is a stronger stat.

However, fights like that are very rare unless you're trying fights that you don't really have the throughput for. In most fights, you either have periods of strenuous healing that eventually end, allowing you to get the raid topped back off for the next such period, or else you're casting continuously in order to keep the raid largely topped off to be prepared for large damage spikes. In neither scenario is back-loaded healing a big asset, really. Obviously, it's not all wasted...some of those Echo of Light ticks occur while you're casting the next spell, which presumably is still before everyone is topped off to full health. However, the larger and larger ticks keep getting pushed further into the future by each successive cast.

You won't do poorly with a lot of mastery, but you may find that you squeeze more useful healing out of critical strike.

Could this problem with mastery be remedied? I think a fairly simple "fix" would be to make Echo of Light operate the way Wild Growth does on each target it hits: front-load the healing so that the HoT starts off strong and then weakens over its duration and, otherwise, let Echos of Light combine the way that they already do. I believe that that would help balance out the continuous pushing back of stronger HoTs. You'd still be creating stronger HoTs with each successive cast, but it would deliver a big percentage of its payload quickly...which, again, has the virtue of being reflected quickly in the life bars, allowing your healing team to adjust their casting around it.

Wednesday, March 28, 2012

Priests Desperately Need Minors

That title should bring in a lot of Google search traffic, I think. That's a quick lesson to all you kids out there who want to be pro bloggers: you can't cheat SEO (search engine optimization) algorithms, so make a title that's actually honest about your subject matter.

This post, of course, is about the criminal need for more minor glyphs for currently frustrated priests.

Blizzard unveiled a fantastic (in my view) system for glyphs: remove the obvious "increase one's powers" glyphs (if they're so obvious, they should just be baked into the spells and balanced appropriately) and just have (major) glyphs that allow you to tailor your playstyle to preference and (minor) glyphs that are largely only for customizing one's visual style.

If Blizzard does a good job, major glyphs will largely be balanced against one another, so none are "clear winners" select the ones that make playing your class more fun for you. That said, I'm not going to discuss major glyphs in this post. I may re-visit them in a future post. This post is about cosmetic/stylish minor glyphs that could be fun for priests.

A bunch of druid minor glyphs were revealed during the Mists of Pandaria press tour blizzard of information (pun fully intended). I can only assume that since those were the minor glyphs displayed during a "first look" that those are intended to be the baseline of what to expect for every class. Blizzard would never resort to the cheap movie trailer tactic of only showing the best bits to lure you in, right? I would think that the cheap escort tactic of hinting at but not yet showing the best bits would serve them better.

So, in that spirit, I'm going to assume priests are going to get an avalanche of amazingly cool minor glyphs and suggest some that I'd like to see.

Glyph of the Shadow Raven

Getting to the obvious one first (but I'd be remiss in skipping past the one that many priests have wanted ever since shadow orbs were added to the game), this glyph would either turn the shadow orbs graphic in shadow ravens, or else just passively surround priests who are in Shadowform with shadowy ravens. In either case, the precedent here are the shadow priests from New Hearthglen, in Dragonblight. They have a few shadowy ravens circling them (all the time, as I recall). It's a nice bit of style.

Glyph of the Angel

This one would be for any priest who took the Archangel talent (perhaps including shadow priests, but perhaps only for healing priests). This glyph would passively provide the current archangel graphic (small, glowing wings) at all times, and when the ability was actually activated, they'd unfurl into a slightly more obvious graphic (perhaps Tyrael-like wings?).

Glyph of Holyform

With this glyph, priests in a Chakra state get the graphic that is currently given to priests who have activated Chakra but haven't yet cast a spell that puts them into a Chakra state (a lovely wreathing in golden flames). Why must such a lovely graphic be mostly never seen (though it bears mentioning that a similar graphic is used for Power Infusion)?

Glyph of Natural/Liquid/Lunar Penance

This would be three different glyphs, each of which re-colors Penance. The "natural" version gives Penance a green, leafy look (Blizzard does a nice job with putting a "falling leaves" look into spell graphics). The "liquid" version would give Penance a blue and watery look. And the "lunar" version would give Penance an otherworldly white glow. (This one is a little dodgy, since golden light is a bit of a priest staple and the "natural" and "liquid" versions would look druidic and shamanistic, might not foster a cohesive look for the class.)

Glyph of Vengeance

Your Spirit of Redemption is now a Spirit of Vengeance...instead of using the model of a graveyard angel for when you turn into a Spirit of Redemption, you gain the model of a val'kyr. Actually, will holy priests even have that ability in Mists of Pandaria? I've almost never taken the talent, so I've really not even paid attention.

Glyph of Poignant Death

When you perish, you emit of flash of golden (or shadowy, if you are in Shadowform) light. The world just got a little dimmer (or brighter, if you were in Shadowform) for the loss of a priest. So sad.

As you may be able to tell, I don't play shadow very much, so I'm rather healing priest focused with these glyphs. I think, though, that glyphs like these (including more selection for the shadowed in the priesthood) would allow for some fun flash to be added to the priesting game.

If you have ideas (hopefully, even better ones than these, though I am very enamored of Glyph of the Angel!), please share them in the comments!

Friday, March 23, 2012

Fixin' Things

So in the wake of another winding-down expansion (Cataclysm, in case you forgot), I've decided to propose some solutions for things that are either considered broken by many people (archaeology, for example) or things that are considered broken by me (transmogrification, for example). I may even fix things that aren't broken, but could stand to be improved. Hey, if I'm waving this wand, I should wave it as much as I can.


This is something I really want to make fantastic, because it almost feels like a great time-waster, but not quite. And I think solid time-wasters are important for those times when you want to be on World of Warcraft (where some of your friends are) and doing something vaguely productive with your character. Blizzard definitely seems to agree, as their next expansion (Activity Stations of Pandaria, in case you had forgotten) seems to be chock full of attempts at giving us things to do that don't take hardcore time investment.

Still, archaeology is almost good, so here's my attempt to make it good.

I think that instead of adding a handful of starter epics, as they did with the introduction of the profession, they should add an enormous number of transmogrification pieces, with varying drop rates. Many different pieces ranging across the spectrum of cloth, leather, mail and plate, and all the various weapon types (or at least the most-used types...considering there are rarely best-in-slot fist weapons, it may be forgivable not to put in a lot of fist weapon transmogrification pieces). Cloaks, too, of course.

Now here's a key: all of these pieces should be account-bound. Forcing you to grind up the profession on every character and then making the vast majority of the drops useless (a cloth piece when you're excavating on your rogue, a plate piece when you're excavating on your hunter) would be vexing. Allow players to excavate on the character of their choice and still feel pleased about drops that their other characters can use. None of this stuff should be stat-filled power gear, so game balance in being able to fire stuff all around your account shouldn't be a relevant issue.

Many of these pieces should be quite rare, others not so rare. Conceptually, I'd say that if a player was likely to find one of these items a week when doing archaeology a fair amount (but not an incredible amount), then there's a pretty good sense of reward in excavating whenever you have some free time. You can do it by yourself, so you don't need to wait on the whims of friends or guild members (as you would for transmog runs through many instances) and it feeds into one form of character progression: the wardrobe you have available to put your own special stamp on your character.

Now, you may ask, "But where are all of these transmogrification pieces going to come from? Artist time isn't infinite and they're working on a lot of things!" I'm glad I had you ask that. One source of these piece, of course, could be recolored versions of existing items. Maybe a green or silver or multi-colored version of those feathery druid tier shoulders that my friend Khizzara loves so much. Or a replica of the Starshard Edge (full disclosure: I've been desperate to get my hands on that dagger, so this example might be a little self-serving).

Another source, if recolors and replicas leave you a little cold, is clothing that currently only NPCs wear. For example the gown that Azshara wears in the Well of Eternity dungeon. Or the gauzy dress that Tyrande wears in the...Well of Eternity dungeon (they really knocked themselves out for wardrobing their dungeon dwellers in 4.3). There are quite a few objets d'art that are worn throughout the game by NPCs. Why not assume some of them ended up buried for us to find?

In addition, yes, create some new stuff. These solutions weren't meant to be effort-free, but rather things Blizzard could do (at some cost to themselves) to upgrade game systems (in my view) without it being excessive in terms of resources spent. Combine some amount of new art, with some recolors/replicas, with some NPC-only pieces and you can create a pretty large number of options without having to design an inordinate amount of new stuff.

Plus, of course, you can toss in the requisite number of mounts and pets, which I assume they plan to do anyway. This would just be to supplement those.


Okay, I don't truly think this is broken. I'm enjoying it. It's just not quite as magnificent, as conducive to self-expression, as I believe it could be.

Here's my list of fixes in a handy bullet-list format:

  • This one's a two-parter and one part Blizzard has already said they're looking into. Allow players, when transmogrifying a weapon, to choose to carry over the enchantment glow on the source weapon in addition to the weapon art. In addition, allow the enchantment glow display on one's weapon to be toggled on and off, like the display of one's helm and cloak. The second, of course, is something Blizzard has already said they're interested in doing. The first, though, I think is a fairly obvious feature to implement. There are a number of enchantment glows in the game to select from and it offers players one more customization venue. And if they think their weapon looks best with no glow, well, that's what the toggle is for.
  • Classify weapons by family and allow transmogrification within family. For example, here is a set of families: one-handed weapons (mace, dagger, fist weapon, sword, wand*), two-handed weapons (two-handed sword, two-handed mace, two-handed axe), pole-arms (pole-arm, staff), ranged (bow, crossbow, gun). You can't transmogrify a dagger to look like a staff or a two-handed axe to look like a fist weapon, but weapons of the same general class can be transmogrified into each other, so that if you have a two-handed mace, but really love the look of two-handed swords, you can still use that for your look. Blizzard already allows this with ranged weapons, so I think there's little to no conceptual leap to extend it.
  • This one is not technically a mechanical issue of transmogrification, but I think it is quite related to the new world order of customizing one's appearance. Allow the display of one's shoulder piece to be toggled on or off, much like the current toggles for helm and cloak. I'd say that shoulder pieces are even more likely to be disruptive to one's desired look than cloaks. In the past, the argument against allowing this was the prestige-identifier theory--that players identified the power level of other players by the distinctive look that their shoulders, chestpiece and leggings created. However, with transmogrification, that's all out the window can no longer have any idea of another player's gear level from the look of their gear if they choose to customize their look. So you might as well allow shoulders to be hidden, since a player intent on changing their image will likely be using the look of shoulders quite out of step with their actual ilevel.
  • Allow white (common) items to be used as sources for transmogrification. It seems like a very arbitrary limitation and, perhaps surprisingly, there are quite a few very nice or very unique looking models hidden away in the statless items.
*In Mists of Pandaria, wands will be main hand weapons


I really should have written this post four months ago or so, like I had planned, rather than being lazy, because the fix for the glyph system that I would have presented is virtually identical to what Blizzard appears to be planning for Mists of Pandaria (minor glyphs being an extensive collection of fun and cosmetic choices, major glyphs allowing customization of one's play style and the "power" glyphs--the ones that just flatly power up your spells--removed and just baked into the spells and talents).

I would sue Blizzard but I've learned from past experience that the legal system is not sympathetic to claims of "thought robbery" while one is sleeping.

Still, I guess that's one game system Fixed(tm). I'm going to take credit for it and wait for my other fixes to be put into action. It's only a matter of time.

To be continued...someday...

Wednesday, March 21, 2012

Developing A Healing Strategy

It's process time again! We do so love our organized processes. Perhaps you've always wondered how a healer should approach a new encounter and put together a coherent framework of action (doesn't that sound professional, like some business world/World of Warcraft hybrid?). Perhaps not. Actually, probably not. I'm going to tell you how, though.

Healing, in my view, can be distilled to this: matching patterns of incoming damage with patterns of healing. In dungeons, things can be a little random (what with tanks of varying skill, DPS of varying desire to pull extra packs, etc) but in raid encounters, things tend to settle into fairly understandable and predictable patterns.

Consistent tank damage is one pattern of incoming damage. Spiky, inconsistent tank damage is another. Consistent damage to the raid (like a persistent damage aura) is yet another. Bursts of single-target damage to random raid members is even still another. I could go on like this for days, but you get the idea. Damage has a shape, a affects life bars in different ways.

Those differing profiles or patterns are what make healing dynamic. If damage only had one profile (say, evenly distributed damage across everyone in the raid), it would strip the dynamicism from healing the encounter. Ultraxion is a lot like that and it's not exactly the most thrilling fight to heal (though sometimes it's fun to have a fight where your job is simply to pump out the biggest sustained numbers you can).

So you could enter a raid encounter with little preparation, beyond knowing what you need to avoid standing in, and see how the life bars change and cast an appropriate spell. Even for reactive healing specs, though, that's not a good idea, as you'll always be a step behind. The "step behind" is not that you're casting the spell after the damage happens...other than discipline priests, that's how healing specs work. Where you'll be a step behind is needing to recognize the incoming damage profile and choose the correct spell to counteract it before you can cast. That's a significant delay that adds up over the course of the encounter, leading to less useful healing than you'd otherwise be capable of.

Study The Game Film Between Games

Read about the encounter...the mechanics you'll be seeing, the damage they do and the recommended strategy. Now, the recommended strategy may not be what your raid team does, but it's valuable just to get a general understanding of what types of tasks your DPS and tanks will be carrying out and what they'll be subjected to.

Watch a video, too, if you can. Patterns can often make more sense when you see them play out.

Based on this research, start to build a plan for what spells you're going to want to use and when. When you're going to need to go all out, guns blazing, and when you'll have some respite to perhaps pop mana cooldowns or generally recover a little (or, at least, ease off your most expensive spells). Especially as you do heroic encounters, you'll pretty much need to be casting something at all times...taking time to decide what it is you need to cast will slow you down. You want to have a sense for what you're going to be casting when ahead of time.

The Scientific Method

Healers are nothing if not scientists. So you have your hypothesis about what patterns of casting will see your group through, but hypotheses need to be tested. Every attempt on an encounter is a nice controlled experiment in the lab. Start off by sticking to the plan you crafted beforehand.

Things are going to fall apart the first time, if it's a difficult encounter. That's part of the fun--data collection! So keep a sense for when things seemed to fall apart (generally it'll be memorable as the point in time when you throw your mouse through the drywall) and, after you've thrown your mouse, try to stay clinical. What happened? What went wrong? Did everyone else stick to the plan? If they did and the damage still got out of hand, then perhaps you didn't have the right spells planned for that part or maybe you miscalculated what the damage profile would even look like. Sometimes, the intersection of several mechanics can be hard to predict, a priori, in terms of how they will combine to hurt the raid.

But that's cool, you're still in good shape. Repair up and adjust your strategy a bit. Ask yourself a few questions: Did I seem to be healing the wrong people at times? Did I use spells that were too weak when bigger spells were needed? How much will patching up that drywall cost anyway? Subtly, but crucially, could you have used different spells at earlier times in the fight that might have helped the situation when things fell apart?

That last one, I feel, is easy to overlook. Generally, and understandably, people gloss over the points in the fight that seemed stable as "fine" and examine only the points in the fight where the deaths began to pile up.

Obviously, the failure point probably has most of the important post-mortem information. Failed raid encounters often boil down to certain critical junctures and once you have those sorted, you have the fight down.

However, as the ripples of anger and fear in Alberta can eventually culminate in a stock collapse in London* so too can decisions early in the fight lead to consequences later in the fight.

*Theoretically. It's probably never actually happened.

As an example from my experience as a discipline priest, perhaps you could have cast a bunch of shields while things still felt stable in preparation for the period of time that even casting all your fast, expensive spells was not sufficient. Perhaps if you had focused on keeping the tank higher in health at a different point, she wouldn't have used a defensive cooldown and she'd have had that cooldown available at a more critical juncture. Perhaps if you had matched the right spells to the situation earlier in the fight, you'd have more mana to blow at a big and important healing moment.

Figure out not only what went wrong at the time when things went wrong, but also what decisions earlier in the decision tree might have averted the crisis. Each encounter is a bunch of new data...use it to refine your hypothesis.

Insert Cooldown Here

I think it's important to first figure out your general plan of action, which is why I haven't mentioned cooldowns to this point. Sometimes it'll be obvious just reading about the encounter when cooldowns will be important and sometimes the raid leader will assign your cooldowns to certain places in the fight. That's fine.

But it's not always clear when cooldowns should be used. Not every fight has an "unusually massive blast of damage now" moment and each raid group does an encounter slightly differently (or a lot differently). In those cases, you'll have to figure out when a cooldown will do the most good by seeing where your raid team has the most trouble and mentally marking those times as the points at which to insert your cooldowns into the fight.

But remember: when in doubt, use the cooldown at the earliest moment that it'll be useful. Cooldowns are powerful abilities that Blizzard doesn't want you to be able to use too often. The less you use your powerful abilities, the longer the cooldown time you are adding to it. After seeing the fight enough times, you may gladly trade using it less for clutch usage to get you through moments you couldn't survive otherwise.

Talk To Other People (Your Fellow Healers Are People, Too)

"Talk to them" is not necessarily literal. I'm also including reading the thoughts of other people on the internet, even if you don't interact back. So many people are doing all of these same encounters that you'll see a lot of opinions. As per Minstrel's Law of Everything: 90-95% of everything (stand-up comedy, music, opinion) is crap. Especially when it comes to the Internet. But that other 5-10% can be pure gold.* If you spend time reading over what other people say, you may come away with one or two solid ideas you didn't think of to make the encounter easier.

*My blog: 90-95% crap, 5-10% gold. This is intentional; I encourage you to figure out what is what, it builds character.

Your healers are particularly worth studying as specimens, as they get to experience the same situations as you, unlike healers in other raid groups. See what they're thinking, what they're feeling, what cultural backgrounds are prejudicing their healing.* It's not impossible that a little communication can cause a revelation that will change how you address a pattern of incoming damage.

*"I'd expect you to feel entitled to our feral druid's Innervate. Everything's just come so easy for you ever since daddy got you a job as a vice president, hasn't it, Spoiled Bobby?"

The Main Point

The main point here is that your job as a healer is to identify the shape of incoming damage over time and use the right spells to match that pattern. You should have an approximate idea of what that pattern will look like in your head going in, so that you have a testable plan. Being approximate, however, means that there will be certain crucial inaccuracies, so you'll use the wipes to analyze and refine both the picture of the damage over time that you had going in and your plan to address it.

Have one part of your mind dedicated to being an impartial observer that doesn't care if you win or lose, it just wants to know what happened. Leave the fiery passions to the DPS...your job is to stay analytical at all times. The more you observe and understand, the stronger the refinements to your plan.

The most important piece of gear for you as a healer is your brain. (Seriously, guys, I just came up with that. I'm always coming up with stuff like that off the top of my head at parties.)