I have not yet gotten around to handling T8 in Rawr so it rates items on pure stats, not taking set bonuses into consideration. Also have to add that Trinkets are in for a major rehaul as no Ulduar trinkets are currently handled.

I can confirm that this bug is still up on live. I saw it last night during an XT fight.

source: Wow Web Stats

The big problem I see with Disc raidhealing is the big waste of all my crit I have on my gear, and if you spam PW:S on raid, you will be soon haste capped, so a lot of stats will be wasted.

More than often, I shield some raid members in any "free GCD" on the tank , to keep BT and gain 4pc bonus, but only raid healing seems a little weird to me.

Sry for my bad english.

Hi Guys,

I have questions reguarding stats for Disc priest. What is the best correlation between haste, crit and spell damage?

I know the answer may varry depends of the healer role in the raid. So, I'm playing as a tank healer the most time and rarely as a raid healer. My stats are about 2,3 spell power, 35% crit rating and 10% haste(raid buffed). I neved use gems for a spirit or intelect since I rarely even use Mana potions.
My main spells:
Glyphed Penance - which like crit, haste and spell power
Flash Feal
Shield on cooldown - which counts only spell power
POM when moving

Is 2.5k spell power, 40% crit and 10% haste the target for me?

This is mostly how I play, though I feel I've solved the "throughput without borrowed time" problem via short duration haste buffs. Every time there's a damage spike, whether on a tank or the raid, I have a haste cooldown available: Power Infusion(20%, 15s, 1.6m c/d), Hyperspeed Accelerators(340 haste, 10s, 1m c/d), and Scale of Fates(432 haste, 20s, 2 min c/d). Even in hard modes, there isn't really a fight where I need more throughput than I have cooldowns, and most of the time Power Infusion is even free to throw on a dps if I want to. Taking the Scale of Fates was primarily PvP motivated, but it's actually very good in PvE and I wouldn't be surprised if I consider it at least situationally best in slot by the end of ulduar. I believe Engineering is also underrated for PvE.

I stack full spellpower gems with dragons eyes, prefer crit and often regen over haste (but generally take the highest ilvl even if it means haste over crit), and my mana and throughput are fine so far. I passed a Spark of Hope to a good resto druid since I'm generally not having mana problems and he probably benefits more (plus we run, on average, 6 druids - our moonkin usually eats 4-5 innervates on longer fights, so I could just steal one if I wanted). I'm currently planning on using a Spark of Hope (when another drops) and the Scale of Fates for the foreseeable future.

I'm not really conviced by the 6% haste rule. Yes, I know that this is the GCD cap with BT up.
However, this cap occurs mainly when we chain shields, and in this case, crit or spirit / MP5 are not really needed either. Basically, all what we want in this situation is intellect and spellpower, but that is two stats we can't gear for.

When tank healing, the GCD cap of 1s is not often problematic. We usely use a long cast spell after BT. Only POM after a PW:S leads to a 1s GCD.

So, I would just conclude that as Disc, you should focus on Int (for regen) and spellpower (for throughput) when you can (gems and enchants), and otherway, equilibrate your crit and haste.

I agree. I especially don't like the idea of capping haste at 6%. For one thing, raid compositions often vary, so you are not guaranteed all available raid haste buff. If you only consider enlightenment and borrowed time, then the "soft cap" with the borrowed time buff up is 13.2%.

But also, one should keep in mind that borrowed time isn't going to apply to every single spell you cast, and without borrowed time up, you're not going to be anywhere near the soft cap. In other words, those casts not benefitting from borrowed time will definitely benefit from haste well beyond the 13.2% level.

This matters when tank healing since weakened soul limits the uptime of borrowed time. "Just throw a shield on someone else" doesn't make a lot of sense in the context of maximizing throughput on your main target since you have to leave your primary target for at least a GCD. Good luck earning that time back. You'll have to use borrowed time on 5 seconds of cast time in order to recover every 1.0 seconds of lost cast time.

So if you're wanting to increase the maximum HPS you can put on a single target, haste is good stat for you -- especially if you only have 6% haste from gear it is almost inconceivable that haste rating wouldn't beat your other options for throughput. The formal relative weights (source) for throughput on a particular spell are:







where the constant (k sub sp) is the spell power coefficient for the spell. Note that PW:Shield throughput doesn't benefit from crit (although the glyph does). However, even if you get the full benefit from crit, if all you care about is throughput, then haste will beat crit until you are nearly capped by the 1.0 sec GCD cap. Importantly, the (throughput) value of haste relative to crit is not dependent on your spell power at all and is:



Any value > 1.0 means a point of haste rating gets you more throughput than a point of crit rating. Since your crit percentage is nonnegative, the numerator (1 + 0.95*crit%) is always >= 1, which means haste flat out wins over crit until 1 + Haste% = 1.4738, i.e. when you have 47.38% haste. If you have 1.9% or more crit%, then haste rating beats crit rating until you hit 50% haste.

Having said that, I think healing throughput is far less valuable than many players assume. Most of the time, healers are not operating near their maximum HPS. Unless you are running into your throughput ceiling, it is unclear how important it is to raise that ceiling still further. Nevertheless, make your own determination -- just don't apply DPS thinking to HPS. The situation is not the same at all.

Also, crit has several benefits to disc priests, namely inspiration procs and DA shields. The above calculation takes into account DA shields for throughput calculations, but "throughput" is maximal by definition -- i.e. it ignores overhealing. Since DA shields are resistant to overhealing and don't interfere with other healers' heals, that adds some value to crit which is difficult to quantify. Inspiration also may add significant value, but it varies and is also hard to quantify.

On top of that, haste burns your mana faster. When spamming flash heal, a point of haste rating will increase your mana burn by over 0.5 Mp5 [5 * (base mana cost / base cast time) / 3279]. Sure, you won't always be spam healing and then the effects on mana are blunted. But if you're not spam healing, then you're not near your maximum HPS anyway, so if you can't stand the increased mana burn from haste, you probably shouldn't be stacking it. Stack spell power or crit, which increase your throughput without increasing your mana burn.

Bottom line: for throughput alone, haste is very valuable and not just up to 6% or 13%. But if you take into account other factors, it is easy to construct a reasonable argument for spell power and even crit rating being more valuable.

PS: Incidentally, it is a lot harder to argue for crit as a throughput stat for holy. The haste/crit rating throughput valuation is:



Ouch! That means at zero haste and crit, haste rating gives 2.8 times as much throughput as crit rating, and there is no point where crit rating is even nearly as valuable as haste for throughput. That's enough of a disparity that you can argue that combinations of haste and spirit are always better than crit. For instance, you can make the case that 1 haste rating + 1 spirit > 2 crit rating for thoughput and barely different in terms of mana effects.

Curious, how fast are you going OOM on Freya hard?

Now with Ulduar and some people picking up Pandora's Plea I thought I'd figure out how much Intellect you actually need for 8 Spirit to be superior to 3 mp5 (as per the half gems we might be picking for socket bonus)

I believe the current formula is this:

Mana Regen = 5 * squareroot of Intellect * Spirit * Base_Regen

What needs to added here is:
Meditation
Kings+Enlightenment (2-(0.9*0.94)=1,154)

Now since I don't know how formulas are made it's a bit primitive:

3 mp5 = 5 * squareroot of X * 8 * 1,154 * 0,5 * 0,003345

X ~ 1510 Intellect

This ofc doesn't mean that mp5 pieces are worse once you reach 1500 intellect but just for those of us that are close to that anyway it's good to know that spirit might not be such a bad thing

Unless you're specced Disc 100% of the time, even at lower Intellect levels taking Spirit gear is a much better idea. Even as low as 1300 Intellect the difference is only ~10% (and that's assuming 100% I5SR), and that gap is made even smaller when you consider that Spirit / mp5 regen itself is only a small amount compared to our Intellect regen / mana pool.

The advantage of having a great Holy set (since Spirit is actually a decent stat for Holy) far outweighs the tiny gap between Spirit and mp5 for Disc, and as you gear up and get more Intellect the gap becomes smaller and smaller until you reach 1500, at which point Spirit is better in all circumstances.

// Edit: Hmm, actually that's not quite right. 8 spirit is slightly higher item budget than 3 mp5 (3 * 2.5 = 7.5, not 8), so just redoing the calculations quickly for Spirit vs mp5 generally rather than specifically for that gem:

1 mp5 = 5 * squareroot of X * 2.5 * 1,154 * 0,5 * 0,003345
1 mp5 = sqrt(X) * 0.024126

X = 1718

So you need 1718 Intellect before Spirit is greater than mp5 for equal itemisation budget, assuming 100% I5SR.

Greater Heal already has much higher HPS than Flash Heal. I feel like I have copy/pasted this about a thousand times this week but here it is:

At 2400 SP, 30% crit and 25% haste:

Flash Heal
(2040 + .81(2400))*1.04*1.09 = 4516
4516*.7 + 4516*.3*1.95 = 5803 average healing including criticals
5803/521mana = 11.1 HPM
5803/1.20seconds = 4836 HPS

Greater Heal
(4270 + 1.61(2400))*1.04*1.09 = 9221
9221*.7 + 9221*.3*1.95 = 11849
11849/1050mana = 11.3 HPM
11849/2.00seconds = 5924 HPS

I also disagree with this idea of "filler" spells. I assume you mean that while Penance is on CD and the tank already has Weakened Soul, you should be casting "filler" spells. I think it is a waste to just mindlessly cast Flash Heal (or anything else) while waiting for Penance or Weakened Soul to come off CD. If your tank is at 100% why mindlessly cast Flash Heal? I would rather start casting Greater Heal and then cancel it at 0.5 seconds left if my tank is still at full HP.

I don't mean to single you out or anything, the mindset you have is pretty common among posters here. The main arguments against Greater Heal are that it is too hard to heal reactively and that it is more likely to overheal than Flash Heal. I agree with both points to a degree but I think that both those negatives can be mitigated enough to warrant using Greater Heal over Flash Heal when tank healing. Precasting Greater Heal and canceling it at the last second if it is not needed mitigates both of those arguments in my mind.