2 February 2011 § 5 Comments
I have lots of stuff I want to talk about – plenty of things that have struck me as worth a comment or a discussion while I was levelling in Cataclysm. I think the most pressing matter, however, is a discussion of the state of Discipline. I switched to Discipline originally at the start of Wrath, as we were starting out in Naxx, because the other healer in the raid was himself a Holy Priest and we thought having some variety would have been a good idea. It was love at first sight: I loved the mechanics of it, I loved the versatility, and I also liked the flexibility in terms of soloing and PvPing (remember, this is all pre-dual spec, albeit by only a few months). I remained Disc throughout Wrath, through our OP moment in Ulduar (before the Penance nerf) and through our slightly darker situation in ICC. However, I am now seriously thinking about going Holy, because most of the raids turn into bouts of frustration and despair. Note that this is not about ego – in our raid I have not been top healer for many fights, and in fact I think I have most often been in second place (regardless of whether we were two- or three-healing an encounter). Now, however, I find myself with about half the effective healing of our top healer on almost every fight, so I am starting to feel like deadweight for the rest of the raid. Why is this happening? And what can I or we do about it?
Atonement, once again – and basic healing strategy
There was a lot of debate in the last days of the Cata beta, as well as during the 4.0.1 time, about the role of Atonement. The question was whether Atonement was supposed to be a mana replenishment tactic, a healing boost tactic, a viable healing strategy, a dps entertainment during those boring heroics, or just a very good coffee machine. OK, I made the last one up (although, hmmmm, coffee machine), but you get the idea: lots of possibilities, not a lot of data to understand, and the little data we had changed on a weekly basis with the various patches. So, I think it may be time to put a few ideas to rest.
Atonement is NOT a way to regain mana. Casting 5 smites will cost around 12,234 mana (depending on how the game rounds fractional costs), and return you 5% of your total mana. Unless your total mana is in the region of 244,672, you are actually getting LESS mana back than what you are spending (if your mana is around the 240k mark…. I don’t know what to say). This is not to say that we will not get there: Wrath mana pools more than doubled from start of raiding to end of the expansion, so it’s possible that the same will happen this time around. But, for now, I think we’re stuck with the fact that we will gain about 50% of the mana we spend on building Evangelism stacks. There goes the mana idea.
Atonement still has a fairly low healing throughput. It has received some hot fixes to make it better: it is now affected by most healing modifiers (Archangel, notably, but also talents like Divinity), but 7k-9k normal hits, and 10-15k crits are still not very much for a 1.3 sec heal (especially given our very low crit rate this early in the expansion). Of course, it is still relatively cheap (Archangel may not return mana, but it sure makes Atonement healing very cheap, effectively halving the cost), and the fact it is a smart heal makes it very useful on situations of low diffused damage. Also, keep in mind that Atonement depends on your Smite damage: so any special situation linked to the encounter that modifies this damage could make the corresponding heal a waste (on Siamat, if you’re smiting the boss before the adds are down) or pretty juicy (in Deadmines’ ragezones).
We already ruled out the possibility that Atonement was meant to be our coffee machine, and, I don’t know about you, but Heroics are still not exactly snoozefests that encourage me to Smite to have fun (and even if they were, I doubt we could ever get any noticeable dps from Smite, really). So, that leaves the healing boost from Archangel as our main reason to build up Evangelism: that 15% is actually a very nice increase, especially when you add to that the 24% of Grace (here’s a question that needs some verification: are they multiplicative? Or additive? I think Blizz made all bonuses like this additive a while back, but it would be nice to find out). This means that our healing should be to throw mending, then alternate Shields and Smite (to benefit from Borrowed Time), get to 5 stacks and then decide whether we want to keep that stack rolling, or get Archangel and boost our healing. When Archangel is up, switch to high throughput heals (Penance, GHeal, but also PoH or Divine Hymn), and remember to refresh PoM and even a Renew or two before Archangel expires. If you find yourself with Archangel up and not very much damage to heal, you can start building the next stack of Evangelism up – rinse and repeat. Notice that nowhere have I mentioned the use of Heal or FHeal. Heal should really not be on your bars: it is for people that don’t have Atonement, as it heals for about as much and doesn’t give you Evangelism (which really begs the question of our 2pc bonus for T11: why, Blizzard, why?). FHeal… I probably went from using too many FHeals (see below) to using too few of them. In theory, if someone is going to die in the next 1.3 secs, and Shield and Pain Supression are not an option (Weakened Soul and cooldown, respectively), then you should use FHeal, while crying about your mana disappearing down the drain. FHeal should, however, be a very small part of your overall healing, all things considered.
Mana, throughput and set-up time
As I was leveling and as soon as I got to 85, the biggest problem I had was mana management. Partly because I was still piecing things together, and probably used a fairly outmoded healing model, I found myself oom more often than not. This problem however went away as I got more comfortable with the healing routine I described above, but also because of another fact. Our mana replenishment comes from Rapture, Shadowfiend, and Hymn of Hope. ALL of those are based on your total mana – so as you gear up, and your maximum mana goes up, your mana replenishment will correspondingly go up, and it really felt it went up exponentially. I did gain about 20k mana in the gearing up process, so maybe it was just physiological – I think in fact all healers now gain mana depending on the size of their pool, so this may well be a common issue.
So the mana issue was just a matter of gearing, and it may well be common to all healers – it is worth noting that I am seeing it a lot less on my paladin, but the difference may also be that I know the fights better now, and so do (most of) the other people using LFG (or my guildies). However, attaining a decent mana situation only served to expose other, and possibly bigger issues.
The biggest gripe I have is a numbers issue. I am spamming my heart out, casting constantly, and my numbers are just way too low. I mentioned my “rotation” (as much as healers can have a rotation) just above, and as far as I can tell, this is the most sustainable rotation for healing Disc priests have at the moment. The problem is, I’m still healing about half of what our paladin or druid healers can do. And I know healing is not about numbers, that latency and reflexes may mean we’re sniping heals off each other, that Disc has never been a contender for the top spot on the metres. But half the healing? That sounds way too high as a gap.
I am not completely comfortable with the choice between mana efficiency and sustainability on one hand (which is basically using the spells we mentioned above) and throughput. Our “high throughput” rotation to me is spamming shield, penance and GHeal on one target – which is basically sure to get us oom very fast, and sadly, and this is my gripe, not really getting a lot of healing out – or at least not substantially higher than what we would get from our mana efficient use of Atonement. I admit that this may just be a matter of still adjusting to the new system (you know, I’m starting to get old, it’s difficult to change…), but it feels that the difference should be higher – and it would be, if we could use the Archangel buff for high throughput, but of course Archangel is not available on use when we want.
This brings me to the third and final concern – the set-up time for Archangel. Most healers have cooldowns that allow them to just heal burst damage: druids have tree form, paladins have Avenging Wrath, shamans… not sure about them, as we don’t use them much and I haven’t delved into them. Our comparable cooldown, Archangel, requires about 7 secs to prepare. Now, to balance that, the actual cooldown of the ability is MUCH shorter (18 secs with 30 secs cooldown, compared to 20 secs with 2 mins cooldown talented for Avenging Wrath, and 30 secs with 5 mins cooldown for Tree of Life), but the set-up really limits the usefulness of Archangel as an “oh shit” button – in fact, the Smite rotation is the first thing that goes out of the window when the shit hits the fan, in my playstyle. I don’t want to stress comparisons with other classes, because it’s unfair and most especially dangerous because we really don’t want all classes to have the same abilities. The message here is that Archangel seems to be stuck in the middle between a true “oh shit” button and a regular feature of Disc healing every 20 secs.
So, in Heroics, I often have the tank plunge dangerously low in the first 10 secs, as I build up to Archangel, and then I use the next 15 secs trying to scramble them back to a semblance of health. In raids, I feel like the 5th wheel of the healing team: I’m either stuck to background healing, or I’m running out of mana. I may be doing something wrong, and I know I can certainly improve on my healing skills. I am currently using Archangel as often as I can, but given mana is becoming less of a concern in normal situations, I could probably start keeping a 5-stack rolling, and use it as an emergency cooldown much more than as a routine part of the healing strategy. Also, I don’t want to overly stress the extent of the problem: writing this post over several days has also helped me to separate temporary frustration with a bad night from a more permanent, underlying problems.
The next patch is also buffing our numbers quite considerably: Grace will apply to multiple targets, Shields will cost 30% more but heal 200% more, and Penance will cost 7% more and heal 20% more. I am not sure. The Grace change in particular will be nice: at the most basic level, it means that Grace can stay on the tank, even if we throw a Penance or a GHeal on a secondary healing target (another tank, or a careless dps). It also becomes possible to get fancy and actively keep Grace up on multiple targets, though that sounds fairly difficult. The Shield change will be nice: although the cooldown change has made shield spamming impossible, having a shield whose basic absorption is 3x the current one will be invaluable, especially in the Archangel-building times which are our weakest spot. The real question is whether the change will affect just the basic value of the spell, or the spellpower scaling too (I would guess a mix of the two). I really really hope these changes will be enough: I still adore the spec, and we have a lot going for us (Power Word: Barrier is nothing short of amazing) – which makes the current state of affairs even more frustrating.
23 October 2009 § 2 Comments
Me: “Actually, no, you didn’t die to untanked whelps, you died because you got cleaved by Onyxia”
PUGPriest: “That’s not true, I got eaten by whelps – people, let’s not get punted into the pits”
Me: *links Recount death metre, showing PUGPriest getting cleaved*
Me: “Let’s also try to stand away from the front of Ony, please – chaining cleave is really a bad idea”
One attempt later:
PUGPriest: “I was eaten by whelps again – tanks need to control them better”
Me: “Hmmm no, sorry – you were hit by Ony’s flame breath. She was in an awkward position, but we need to make sure only the tank is in front of her”
PUGPriest: “No, I tell you, it was the whelps”
Me: *links Recount death metre, showing the one shot due to a 25k Flame Breath*
PUGPriest: “Look, Tsark, I don’t care what that mod says – I know what happened!”
I don’t care who you are – whether you are my best friend or a PUG I meet for the first time. I also don’t care how you die: however stupidly you do, I probably have died in a similar way in my past, and quite likely I have died in even more stupid ways. What really bugs me is when you
- blame other people, when you are not blameless yourself;
- try to argue that Recount is wrong.
In fact, this post is really about the second issue. I don’t think I am a metre fiend. I have never kicked someone out of a Vault group because their dps is too low (and I have left Vault groups that did that on at least two occasions, precisely because I don’t think it’s the right thing to do). I also hate it when people link metres after a boss fight (malus points if they are doing it during a loot roll moment, because that forces everyone to scroll around to see who won the roll). However, I still think Recount is an essential tool to figure out what happened.
During a raid fight, we’re all going to be busy doing lots of things: healing, tanking, dps’ing, moving out of fires, etc. It is highly unlikely that we will have perfect awareness of all that happens, and that goes double if we’re getting one-shotted by an unexpected source. If it’s slow damage (e.g. the fire debuff from Gormok the Impaler), we will notice; if a mob beats on us slowly (e.g. one or two untanked whelps during Onyxia) we will notice; if we are having a lot of whelps flying around, and at the same time Onyxia fire breathes and immediately after fears, then it’s unlikely that we can know for certain what happened. That’s where Recount helps: it tells us exactly what went on. Most importantly, we can check it after the fight, when things are a lot calmer, and we can think back about what happened and how we can improve. Note that what I just said applies even more to combat log analysis sites (e.g. World of Logs, which is the one I use currently), but that requires people to be even more “offline” (i.e. not busy in actual combat) than a simple Recount link does.
Once you see a Recount link, the absolute worst thing you can do is to say “the mod is wrong”. Newsflash: computers don’t lie. Sure, they suffer from GIGO, but that’s a different story. Also, Recount is pretty solid for recording combat logs – and while it’s possible that you died because several things hit you at once, they will all show up on the combat log. Recount is not perfect: sometimes, it will just record your death and little before it, for reasons that are unknown to me. But I’ve never experienced Recount doing a false positive, i.e. recording something that didn’t happen.
Now, of course, there’s different ways to use the information Recount provides. Pointing fingers and assigning blame is never nice – and often counter-productive. Clearly, seeing how people die can easily lead to just that – so as a raid leader I normally try to avoid doing that, unless the raider in question is insisting that they didn’t die from such-and-such, while Recount shows they did. However, if you’re called out and a recount link proves you wrong – the best thing you can do is say “Ooops” and move on. Really, dwelling on it is just going to make the matter much, MUCH worse for everyone – yourself included.