I think all mana classes, including us Druids should be making the effort to be self-sufficient and not factor an Innervate into their own regen. I try and hold my Innervate back as I think it's better used on, say, a Mage who has just been Rebirthed. Another example; I am often Soulstoned and, after accepting the Soulstone rez, the Innervate I cast on myself is a invaluable.

In summary, I think Innervate should be held back for exceptional circumstances. If you find it's the same person always asking for an Innervate I think they should be addressing their playstyle and gear. With the upcoming change to the cooldown then it will become more 'disposable', but I still won't take its usage lightly.

In regards to Innvervate I think another factor is how much volume of heals are you doing for your situation? Are you one of the most critical healers? Do you have to pick up slack from other healers?

I do a lot of pug raids lol so I heal a lot more than maybe I would if I was in a raiding guild, I heal like crazy and my #1 focus is keeping everyone alive so I save Innervate for when things get crazy I know I can quickly get full mana and keep a fight going.

I dont know if this is the absolute best approach, but I think it fits the healing I do and has been fine.

I have used Innervate in situations where maybe I am healing a paladin tank and maybe they are very low on mana, and so I may hit an innervate on a paly tank to make them more effective if my mana is doing fine anyway, thats the only thing I can think of.

I am still learning though.

I believe the equation for how to use Innervate and how to gear with consideration to Innervate is a bit more situational than simply maximizing your own performance. The equation should revolve around maximizing your raid's performance as a whole. So it should be considered if wearing regen gear and using Innervate on a fellow raider (in normal circumstances) will maximize your raid's performance.

The usual equation on how to gear for regen vs. throughput, is gear for regen until you can last the whole fight, and then gear for throughput. When considering whether or not to include Innervate in that regen portion, maybe it is better to reverse the equation and say gear for throughput until everybody stays alive, and then gear for regen until you don't need Innervate. However, it seems it is simply much easier to derive how much regen we need to last a fight then it is to derive how much throughput we need to keep people alive.

Edit: After thinking about this a bit more, the reason throughput takes priority over regen is so that, at some point, you can switch a healer out for another DPS.

Since all healer and DPS classes have competent mana regen options, I don't see why you'd want to overload yourself in mana regen gear and considerably gimp your own throughput just so you will always give your Innervate to someone else. If you do that, the person that would always get your Innervate will be half-useless anytime you don't happen to be around. Also, it means that you will be much less effective in situations where high throughput is required.

I gave my Innervate away maybe 3 times total since I hit 80 and I'm fairly convinced that this was mostly the result of poor play of the person that needed it.

If you give your Innervate regularly to other healers, then there is something seriously wrong with your own setup and your raid setup because it means that you could just heal more yourself and the other healers have to do your job, so they need your mana.

I think there are certain tools at our disposal that are for contingencies, i.e. Innervate and Rebirth. I like to be sure that I have these tools when I need them. Another example which I had on Igniss recently, 2/3 through a fight two healers die and they are the only deaths. No Rebirths are available. The answer was to increase my own healing at the expense of efficiency and the savior was having my Innervate as backup.

If my entire regen including Innervate was based around predictability then what is left when the unexpected happens?

In relation to stat weightings the only gear I would say I keep for regen is my belt, Idol and meta gem. I also have Spark for certain fights. Otherwise I employ pretty standard gemming and enchanting.

Several things Mondas:

1. In a situation which approaches saturation for healing throughput (like Steelbreaker last, Mimi P2, Freya +3 in places), a Druid geared to not use their own Innervate has no choice but to throw that Innervate on a healer with higher throughput. As Resto Druids and Holy Priests are fairly even in this regard, what you're really doing is tantamount to saying "I'll let someone pick up my slack and give them my Innervate to do so".

2. In models of fights, especially models of hard modes, the "unexpected" needs to be prevented as much as possible. If a healer is running oom long enough before others that he/she requires your Innervate then their gear and/or playstyle should be addressed before you sacrifice your own throughput to bolster theirs. If it's a case of them dying and needing your Bres/Innervate to keep healing then either the healing team as a whole is not prioritising their assignments correctly or they need to stop standing in fires.

3. If you gem/enchant according to the "standard" model then you are gemming/enchanting for throughput in general and will be using your Innervate on yourself for most hard modes. By the tone of your first post though, I got the impression you were advocating altering the standard model to accommodate routinely using your own Innervate on someone else.

4. With regards to your Ignis example just bear in mind that the plural of anecdote is not data.

I must agree with most of the assessments here. Innervate may be a spell that can be cast on someone other than yourself, but I believe that the only reason for that is because we have specs who's mana does not play a factor in pve. IE Feral specs. Priests have their Shadow Fiend only work for them because every spec uses mana so regardless of your choice of playstyle, you will inevitably need the mana that this ability provides.
Innervate would be a useless ability for a feral and in many cases a boomkin spec'd druid in a raid, so Blizz have made it a targeted ability. Just because it can be cast on another player does not mean that it should be.

Resto gearing should be balanced around using Innervate on yourself. Take Hoddir for instance, During frozen blows and at least the 10 seconds before it starts, you should be GCD capped. If you gimp your throughput to the point that you do not need to use innervate on yourself in an encounter like this. Esp if you heal like most and keep yourself GCD capped on rejuvs and WG. Then in higher dmg situations you reduce your effectiveness. IE, mim P2 and P4(HM) Frozen Blows, etc.... Simply because your heals will be less effective per tick, but you will still only be getting the same amount of heals out per second.

Personally I like to gear like this. If i GCD lock myself to rejuvx5 WG rotation (With reduced rejuv cost idol), with a few lifeblooms on tanks every now and then. I have to use my own innervate at about the 40 -30% mark of the fight. Often i will end with about 8k mana in this situation. However that is a smooth fight and stays inside a very high throughput, high efficiency rotation. If things go bad and i need to cover a MT or OT because of a death or fight mechanic, I still have about 8k mana left over to dip into a less efficient but more single target throughput rotation.

If your regen is too high, your gimped throughput could be insufficient for spamming a tank in special circumstances.

In response, Lightflower:

1. I know you are speaking generally, but for me in a raid there are no healers with higher throughput so I see my Innervate as best used on myself or in one of the circumstances I described. My thought is that the other healers never needing my Innervate and most of them being 5-7% lower than me in effective healing is not coincidence.

2. I should add a postscript to my earlier post; my guild hasn't tried any hard modes yet. However, I think what you are saying is not to plan and gear for accidents or failure, but to plan and gear for an expected successful execution. If I am building contingency into my play then I am gimping my norm, so to speak. I would say that you and some other posters may be changing my mind about some of my gearing.

3. No, I would never tell a Druid to expect to give away their Innervate. I guard mine jealously. To clarify, my mindset has been to keep it for myself unless it harms the raid not to.

4. Fair point.

Feralkin, I don't think your reasoning for Innvervate in your first paragraph fully holds up as Innervate was a targetable spell when it was a talent in the Resto tree and wasn't available to ferals. I agree with your other points though. In fact, if you are finishing fights with around 8k mana you're pretty close to having a second Innervate to give away next patch assuming a new Innervate gives about 7.7k mana.

What's the difference between 1.05s and 1.00s GCD if my repetitive button mashing can't possibly reach that resolution. It ends up being 4.8% increase in throughput if you can make use of it.

Here's a bit of napkin math. Lets say I mash my Rejuv at a constant rate of 6 times per second, or a resolution of .167s. At that speed, 1.00s can be cast every 6 cycles, or at exactly once per second, while 1.05s will be cast every 7 cycles, or once per 1.167s. That's a 14.3% difference, so doesn't that prove that 1.00s is better? Not really, because it's impossible to time exact 1.00s button presses, and the only way to minimize any 1.00+ delay would be to mash buttons *really* fast. You would need to be mashing your Rejuv a staggering 20 times per second to get the resolution necessary to ensure that your 1.00s Rejuv will cast at least one button press sooner than a 1.05s Rejuv.

We can't press buttons that fast. Not even close, not on a regular keyboard during raiding conditions at least. Without a .05s or less frequency at the instant of GCD finishing, you can't consistently cast 1.00s faster than 1.05s in practice.

Perhaps I'm missing some sort of ancient Chinese GCD knowledge, but I don't see how you can make use of the full throughput increase without a godly key press speed. Then again, people could be using some sort of program like AutoIt to make your keyboard into one of those "turbo controllers" used on consoles. I would distrust using something like that since I could see it triggering some sort of anti-bot/anti-spam mechanism which temp bans me for a few hours.

I also read somewhere that spell queuing no longer works below the 1.00s GCD (and I'm also assuming that you can't queue before 1.00s for a 1.05s cast, so you still won't get much from it) so how else would someone do this?

I can't help feeling like this argument is horribly fallacious. Lets assume its true that the haste that gets you from 1.05s to 1.00s is useless. What about from 1.10s to 1.05s? Going by your argument, you'd have to press the button equally quickly for that extra speed to be utilised - and so on all the way to a 1.50s GCD. I hope nobody would agree that haste is completely useless

You're right that with a 1.00s GCD, you'll never actually cast that quickly, but you need the 'leeway' a 1.00s GCD gives you in order to cast every 1.05s. In the end, however fast your latency/button pressing allows you to actually cast, haste until the haste cap is going to allow you to cast more spells.

Think of it this way. You know how when you are in a hard mode trying to heal like a bajillion targets but you're waiting for your GCDs? Well, with .05 secs more speed, more of those heals you are spamming to go through actually will. Sure it's hard to measure consistently, but more is heals is more heals.

I think of my GCD haste cap like a a DPSer would think of their hit cap, it's a priority to reach it.