You don't know what they're doing though, 700 haste can be viable for say spamming Flash Heal or Greater Heal on tanks, or spamming Prayer of Healing.

Haste is only terrible if you're spamming PWS, because then you aren't using the haste. But if you're tank healing, depending on the situation, you might not have time to spare PWS on the raid (looking at you Algalon, pre-3.2).

Also, even with Divine Fury and Borrowed Time, you won't be hitting the GCD using Greater Heal.

Well...i guess that's partially personal preference, partially a "pre-ToC"-relic...
I personally wear a mix of ToC10H/ToC25N/Uldu-gear and noticed a drastic increase in Int + Spirit.
Nevertheless I still find myself in situations where I'm in trouble keeping up high HPS over the full duration of a fight -
e.g. ToC25H/Northrend Beasts

In P1 I usually assist our pallys/druids with PoM+Renews on 2 of our 3 tanks and top our melees with PoH+CoH after his AoE spelllock-stomp. Having an additional 200SP still wouldn't be enough to fully heal a melee-group with a single PoH, so I see no reason in gemming into SP. In the middle of P2 I'm already having mana issues because I have to use almost every GCD...

Since BC I tried to manage my gear with the following rule:
If I can heal through a perfect try of a (currently) hard encounter without having to use my shadowfiend I'm on the safe side because I have room for negative "surprises" like the death of another healer or being brezzed after screwing up in a void zone.

Imo there is nothing worse than wiping after a 6minutes encounter with the boss having 4-7% of his life because a healer ran out of mana and the team wasn't able to compensate his/her dropout. Imo healing a raid is big team effort and as long as every healer does his/her job, there is nothing to worry about. Our pallys Holy Light regularly hit the MT for 20k, it doesn't matter if my FH hits him with 6,3k or 6,5k - it's just important THAT it hits him.

I know that the theory may say something different but I think when you run around with a respectable amount of SP (which you automatically do in ToC-gear) worrying about sustainability is more important than finding a place for an additional 23SP to push your performance in the meters.

kind regards
Kyonyuu

See my armory page, I have 34k mana unbuffed.

I'm actually assuming Shadowfiend, Hymn and Mana Pot will be used, hence I stack up on Spell Power. I just can't bring myself to stacking Int - smells Paladinish

You only need exactly enough mana to not run out of it over the course of a fight assuming the use of a potion, Shadowfiend, and Hymn of Hope. Anything more is simply a waste.

Basically what Tsigo said: if you can find a fight where you go OOM with that much intellect, then by all means stack it. The only fight I've ever come close to those levels on was early in progression Ulduar.25, doing Thorim with 2 MCd mobs, tank healing using the "spam buttons harder" method. It's amazing how fast you spend mana when you sit under the GCD on almost all spells, essentially casting as fast as is possible.

The second example would be Freya.25+3, spam-shielding the raid. Even with the shields getting used up, there's not enough mana coming in to balance the mana going out, even with cooldowns. However, you can easily get to P2 before OOMing, and then who cares, because p2 is easy.

Basically, running Intellect trinkets as Disc is overkill, and you should be looking for solid spellpower / crit options in those slots. The combination of [Sif's Remembrance] and [Eye of the Broodmother] work nicely.

If you're never completely OOM during an encounter, there's no need to have more regen or longevity (mp5 / int).

It's just more reasonable to gain throughput.

Although I guess the fact that you are able to spam heal when you want (not necessarily need) isn't a bad thing.

It would be very easy to use similar, if more complicated, logic to say that more throughput it useless the vast majority of the time. If you could have kept everyone alive with less spell power than you didn't need the spell power you have and you certainly don't need more. If everything goes according to plan we don't need more throughput or more longevity, and we could wonder why we want either. Obviously, however, we want more of both because both help when things aren't exactly according to plan.

Here are some real world examples of a huge mana pool saving the day:

Fighting Twin Val'kyrs on 25-man we lost some people to silly mistakes and ended up not breaking on their shields, allowing them to heal 20% (before the buff). I managed to - just barely and with a mana potion - spam shields on the raid for the entire fight with the occasional penance despite this, clocking in about 9k healing per second. Regemming for throughput would have hurt the healing I could have done.

Single healing 10-man hodir usually has me standing in a light beam during frozen blows alternating PoH on the two groups with Mass Dispels to break freezes. I overheal substantially, and could reduce the frenzied pace of casting slightly to save mana, but then I would be letting people take two ticks before I healed them, which would put them in danger if they fail to dodge an icicle or if I have to dodge one myself.

I lost my healing partner halfway through 10-man XT hardmode. I moved directly on top of XT to keep everyone in range and started firing off heals with complete abandon. If he would have gravity bombed me we might have lost, if he had done it during a tantrum we would have lost for sure, but he didn't and we won. We won because I could chain PW:S and PoH during tantrums. The amount of spell power I would have needed to make that situation viable if I had to worry about mana is unattainable.

Your tank (and pretty much every other raid member) usually returns to full health every few seconds during the course of a fight. Basically, throughput solves crises that last a few seconds. Regen solves crises that last a few minutes. That is not saying that regen is better, that depends on your distribution of crises. I lean towards dealing with longer crises because: 1) That's what lets your raid get more practice during attempts on new bosses and 2) I feel that I have stat-independent tools to deal with short term crises such as Pain Suppression, Power Infusion, yelling at the tank, and good reflexes.

Plus I'm not that convinced that you can't save more lives by having enough mana to spam significant overheal than you can by healing for more when you do heal.

Edit: The argument that if you don't run out of mana you don't need more int is not dissimilar from the argument that if your house doesn't burn down you don't need fire insurance.

Your argument is dangerously close to suggesting that neither extra throughput nor longevity are helpful, which means healers don't need gear upgrades. In some sense this is true-- there's only a fixed amount of healing to be done. But all upgrades are good somehow. If you cannot spend all of your mana, then at least extra throughput will decrease the odds of someone dying due to random spikes. Extra healer throughput might reduce the odds of a tank death from 2% to 1.8%. But if odds of running out of healer mana are 0% (and thereby causing a death), then gaining more mana doesn't help at all.

Remember, it's possible to have enough longevity. But there's always a chance of tank spike death in pretty much any encounter, so extra throughput always has an effect, even if it's minimal. Spell power can be "not terribly useful" and still be the best stat to stack.

There is a maximum useful throughput just as there is a maximum useful longevity. The difference is that we all agree that we could never hit the maximum useful throughput (e.g. your flash heal hits for the tank's maximum health and you have a 1 second GCD), but maximum longevity seems quite reachable. I disagree, however, that it has been reached in current gear, and I certainly don't think that something like a 30k mana pool gets us to the point that we get zero return from more longevity.

I've probably averaged about 50 boss attempts a week for about 38 weeks of wrath, and I've listed two examples I can think of where having what is normally a very excessive mana pool saved my raid from what would otherwise have been a wipe. That gives having a big mana pool about 0.1% rate of saving me from wipes in my own experience. Add to that the first time we beat heroic beasts - 9 seconds after the berserk timer and with me having contributed about 120k damage from holy fire and smite because I could afford to dump my mana when I wasn't healing, and now it's 0.15%. There may be more examples, too. If I thought that regemming for spell power was going to give me a 0.2% better chance of saving the tank then I would consider it, but I really don't think it's anywhere close to that high.

When we talk about spike damage getting our tank, we are talking about the tank dying very quickly, certainly in less time than it would take to land two heals. In that case, the 300 more spell power you would get from regemming spell power instead of intellect would only save the tank if he receives a heal within that second or second and a half, ends up at 250 health or less after that heal, and then goes on to survive. I honestly don't think I've ever seen that happen. I recognize that it could, but it seems like an exceedingly rare event.

If the tank is below maximum health for a long time (six or more seconds, say, since that is a very long time for a tank to be below maximum health) then you might be healing the tank for upwards of 3-4% of his maximum health with that 300 spell power if you are spamming heals that entire time. Again, if it's been six seconds since the tank was at full health, you've been spamming heals the whole time, and he is knocked to 3% of his health then he goes on to live, that would certainly be attributable to that extra spell power. But is that going to happen 0.2% of fights? That just seems like an overestimate to me. Admittedly, I cannot get my head around how to begin modelling this kind of thing, so unfortunately that's a gut feeling.

If you don't smite with your spare GCDs, if you would never stand in a fire for an extra couple of seconds and use binding heal instead of stepping out of the fire and using flash heal to stabilize the tank, or if you choose not to pre-shield three people when you know one of them is going to be damaged at random and the other two shields mean nothing, then more mana has would have more than a 0% effect on you. There are lots of ways to use mana to make things safer for your raid.

Sorry if this seems off-topic, but I read Elitist Jerks forums to hear from folks that are on the cutting edge of content. If you haven't done any Hard Modes in TOC-25 at least, then I do not value your opinion as much as someone that's already cleared all of HM TOC-25.

I don't care what numbers you throw out or your theorycraft - if you haven't done the hard content, you really don't know what you're talking about. So I don't see how arguing that you should stack int/regen when you haven't even seen the content really gives you a valid foot to stand on.

Mod me down if it seems ad hominem.

Not quite. I've done every fight in the game. I know for a fact that I don't need more mana for the current content.

My thoughts exactly. The only thing I care about for this level of content is speed and reaction time for Anub'arak Penetrating Cold, so I stacked haste. Should Icecrown provide longer fights that require more mana, it would take all of 15 minutes to re-stack Intellect.

Granted the current tier of content isn't that difficult, but there's no reason to be ending fights with lots of mana and cooldowns remaining unused. You'd be gimping yourself and your raid.

There is a thought that going into the next tier (fresh into 3.3, for example), you should always have max regen available for the "unknown" factor. The way our 245/258 gear works, though, is to stack very, very high levels of intellect and spirit without even socketing for it. Combine that with [Solace of the Defeated] items, and our regen will already be at historically high levels going into IC.

So, in a roundabout way, agreeing with Tsigo, and promoting the idea that going into IC, consider wearing a regen-heavy set, whether by gemming for it, or just switching around pieces to get the regen higher. It's always been useful in the past (BT, Sunwell, Naxx especially with the gear reset, Ulduar, and ToC).

The length of fights you encounter will be randomly distributed. Even doing the same fight multiple times will have a random distribution of lengths. If everyone lives and nothing goes wrong then ignoring escalating gear and learning curves, which will tend to make the fight become shorter and shorter rapidly, the length will probably be roughly normally distributed with a small standard deviation. In that case it's easy to say you have enough mana. But overall the length of fights is not normally distributed because every time someone dies or makes a significant mistake the fight jumps up in length.

But I'm wrong about that because fights these days are set up in such a way that if you lose a significant fraction of your dps then the fight does not get longer but simply results in an automatic loss. If a clump of your dps run into two orbs and blow up then you might not be able to break the Val'kyr shield, in which case you simply can't win, the fight doesn't drag out. Beasts has a short berserk timer, low dps on Jaraxxus means getting overwhelmed by adds, champs is all about having them lose people before you lose people, and any significant loss of dps on Anub means either falling behind on adds and having your tanks die or doing no real damage to the boss at all. Ultimately the length of all fights, while randomly distributed, will end up with a cap on the end caused by the dps check.

I will continue to gem for int because my guild whenever possible runs 25-person normal modes with 14 or 15 people to gear up for the 10-person hard modes, and those fights can take 14 or 15 minutes, but clearly that is just us being an exception and has - as many posters above said - no relevance to guilds who are trying to defeat 25-man hard modes.

It's not just about the tanks dying. Lots of fights provide instances where random people can have spikes. For example, the arcane blasts on Algalon. Or someone messes up and takes two orbs in 3 three seconds on twin Valkyrs. Someone has taken a few stacks of scarab damage during Anub'arak kiting and takes one hit of a passing burrow. These are all situations where extra throughput will absolutely save someone from dying and extra mana does not.

I would also suggest that if you are so close on the beasts enrage timer that holy priest DPS is making the difference, the solution isn't to stack more mana. You want one of:

- Put more DPS in the raid so the fight is shorter
- Do a better job of keeping DPS alive so the fight is shorter
- Have DPS learn to play better so the fight is shorter
- Recruit better DPS so the fight is shorter

Something is wrong with your raid if you hit the enrage timer and stacking extra mana is fixing the wrong problem. You'll note that all the posters in here that are clearing most of heroic ToC25 are claiming mana is a non-issue. And it's only people struggling with the fights who say mana is a problem. I think you're trying to solve a DPS problem by improving healer longevity (instead of throughput), and that's just going to make things worse.

It's not like we have a lot of gear options. Most of us will go for 4 set tier 9, with the BoP chest. All the other cloth slots are basically one option. Enchants are pretty standard per slot. All it comes down to is Trinket choices and gemmings. Overall, I've always (since MC in Vanilla) favored balanced stats over anything else due to it helping you in all cases, instead of specific ones.

Red slots = SP
Blue slots = Int/Spi
Yellow slots = Int or Haste depending on my stats and what I feel like

And a Nightmare Tear for good measure.


Trinkets:
I use Eye of Broodmother and Sif's Rememberence (until I get a ToC one).

If you run RAWR, and optimize your gear/gems/enchants, you basically get the same advice when you set it up to min/max your throughput and efficiency at the same time.



Maxing out your Int is fun for e-peen, I do it when I swap to Disc for kicks. But MT healing really doesn't need more than 30k or so raid buffed. You really should focus on SP way more at that point as Disc doesn't have Spiritual Guidance and @245+ ilvls, that can be a huge deficit.

On a side note, Blizzard really should have added Int based +% to healing for Disc. It would have had minimal impact on Arenas, as mana pools there are very gimp.

Actually pocketmage, for 245 ilvl I found that [Legwraps of the Demonic Messenger] is my preferred option, and for 258 I'll prolly be going for 4pc bonus + [Handwraps of the Lifeless Touch]. Pre-258 I'll gladly be breaking the 4pc bonus for 258 upgrades, although in reality this means just that I'll replace the gloves before I have the tier legs.

Also I agree, I don't really see a huge advantage in going beyond 30k mana right now and will prolly inch towards gemming rep sockets pure Sp, otherwise maintaining current gemming(yellow sp/int, blue sp/spi). I still can't ever justify ignoring a socketbonus other than Spi.

If you want to balance the stats you get from gems, that's a suboptimal way of doing it. Putting a [Runed Cardinal Ruby] in a red slot and a [Seer's Eye of Zul] in a blue slot gives you 10 int, 10 spirit and 23 spell power. Putting a [Purified Dreadstone] in a blue slot and a [Luminous Ametrine] in a red slot gives you 10 int, 10 spirit and 24 spell power.

Runed Cardinal Rubies should only be used by players who gem for spellpower in nearly every socket.

Obviously the example where we hit the beasts berserk timer is an example of the raid doing a poor job. It is, however, an example of us doing a poor job and winning as opposed to an example of doing a poor job and losing, which is what we had done in our previous attempts. Wins count, and stats turning a loss into a win makes those stats count. The following week we took two tries on beasts and won with plenty of time left, last week we one-shotted it and I expect we will continue to do. There was no long term or recruiting problem to solve, it was simply a situation where we needed more practice and to play better, and a little bit of priest dps got us our win a couple of attempts earlier then we would have otherwise had it. What more could I want my stats to do?

As I acknowledged in my previous post, my analysis of fights getting longer when people die is largely wrong because most fights seem to have a natural limit to how long they can be, whether it is a berserk timer or some other kind of dps check. I also think my view is skewed by playing 10-mans and mostly only experiencing 25-mans on normal modes with small numbers of raiders. On 10-mans it is not usually that simple to change the number of healers you bring to a fight. Almost all fights require two, and going to one is just out of the question regardless of spell power, so putting one healer in a position to contribute dps during lower healing times in the fight is an alternative.

Anyway, I am clearly going on against popular opinion expressed by people who know what they are talking about so I am almost certainly the one who is wrong. It is possible that 10-man guilds who do 10-man hardmodes just experience a different set of challenges than 25-man guilds doing 25-man hardmodes, and that both views could be right from their own perspectives, but it still seems more likely I am wrong. Unsatisfying as it is to do so without intellectually convincing myself of the merits of it, it's only a couple thousand gold to regem, so I'm going to see if it makes a difference to our phase 3 Anub where the win or loss is really on the healer's shoulders.

Of course throughput is better in such situation. But such situations are actually also an argument to socket haste instead of spellpower because landing any heal at all in such a situation is so much more relevant than if the healsize varies by <5% (or whatever the difference between socketing spellpower and not socketing spellpower actually is).

Regarding l337n00b: Your right, 10 man and 25 man aren't really compareable. The combination of lower gearlevel and that you seldom have all possible manaregen buffs in 10 man can make socketing for int still very viable imo.

This is why I prefer having a nice mix of all stats: SP, Regen, and Haste/Crit.

Regarding dmg spikes, I definitely think that haste matters quite a bit more than SP but only to a degree. If you're fast enough, then extra haste does nothing for you and SP is the clear winner. So there's a bit of a hand wavy haste cap that is highly affected by your raid's healing composition and how nice you play with the other healers. The magic number for haste varies and is a very hard number to nail down.

I find that 15% haste works perfectly for me.

I'd like to get my crit back up a bit, and it seems like my next couple upgrades will do that nicely.
Spell Power increase is always welcome, as HP pools are greatly increasing with ToC gear, and stuff is always hitting harder.



Good to know, I'll use that sort of gemming in the future. Not worth regemming 7 slots for just three extra SP.
Thanks