You make a good point, but I don't think the system is likely to change. It's not a huge problem if the designers understand the issue. The (mp5_out - mp5_in) term is a crucial quantity affecting longevity, and so care must be taken to limit regen stats that could make that value too low. For instance, higher tier gear could increase throughput stats mostly with only minimal gains to regen stats.

Alternatively or additionally, harder content could demand much more healer longevity. It's a matter of taste, but I kind of like a design where harder bosses take increasingly longer to down, demanding more and more of your dps and healers, partly because it makes "easy" bosses really, really fast and easy for an overgeared group.

So I'd prefer they change their itemization and raid content to fit the mechanic over changing the mechanic. But if they cannot manage that, then maybe applying some diminishing returns to mp5_in may work. As Elimbras noted, this problem is just like the problem with avoidance stats, so it could be dealth with similarly. I don't like that option because it's a messy hack, and the same thing could be accomplished more elegantly by simply changing the stats on items.

I have a question about Disc vs. Holy healing. I ask this mainly because two of the best healers in my guild are at odds, and finding some kind of evidence or technique to make them work well together is going to be important for us when Ulduar comes out. (We are a 10-man guild.) In terms of keeping in line with this thread, it might be worthwhile to have a section on how Disc and Holy Priests can synergize with healing Shamans, Druids, and Paladins, but also how they can synergize with each other.

Basically, the Holy Priest states that running with a Disc priest nerfs the efficiency of her Prayer of Mending, which she says is her most efficient heal. Because PoM procs only on damage dealt, not the damage absorbed by PW:S and DA, she feels that the Disc priest is causing her PoM to die before it's able to get all it's ticks in. The Disc priest, on the other hand, believes that shielding raid members is a critical part of his healing kit, and tends to shield anyone who takes heavy damage in the raid to keep Borrowed Time alive.

Is this just a matter of calling the Disc priest a tank-healer and the Holy Priest a raid healer? How can they both raid heal without shields and PoM procs stepping on each other's toes?

For Disc the use of shields is quite important. When 3.1 hits it will be even moreso... This doesn't make Holy/Disc an incompatible composition.

Shields are not going to normally going to cause a PoM to "die out" prior to when it would anyway. What it may do is slow the bouncing though or if it is "started" on the tank delay its first bounce. In raid damage periods often the damage exceeds a single PW:S amount so the shield will break eventually and the PoM will clean-up the next damage. The slowing of the PoM bouncing may cause the Holy Priest to recast PoM expecting all jumps are complete before they really are... That is either laziness, inatteniveness, or wasted bounces not being a concern.

The presence of shields needs to be something that your other healers need to possibly change their playstyle around and can improve overall raid performance. I very often raid with a Holy Priest and we work extremely well together. On a fight like Grob we both tend to PW:S the disease target to limit the healing they require. Now he delays any PW:S he would do so that I have a moment to do it first due to my shield being better, and unless his shield is required he avoids them in general on most fights.

PoM is a great tool but it is by no means "better" than your Disc priest paying enough attention to PW:S people in danger. PW:S is a great way to halt the influx of damage so that heals can land before death. There is no guarantee where the PoM will bounce in raid damage situations. Damage not taken is better than damage healed, generally. Another way to think about it is would you tell the raid to "stand in fire" so that the PoM bounces more & faster?

It is important to remember that healing is about keeping people alive, not about leading a meter.

We raid in 25 with four priest often, and that's not a problem (except that it means we lack diversity in healers class). Yes, Disc is shielding (and a that's one of his most efficient tools), and it has somehow a bad interaction with POM.
But what your healers needs to understand anyway, is that any healers has negative interaction with other healers : you have a (significantly) increased chance of overhealing one target if there is other healers healing (that target). And luckily, shields are among the lest "concurrential" heals, meaning that one healer can shield to prevent further damage and the other effectively heal past damage without both heals landing at the same time and one of them being wasted overheal. So, your holy priest may have slower POM, but she also has less overhealing on raid for all heals she uses.

The main question is however assignation. It's good to know who will be using which ability on which target and case. If your disc priest is on duty on the tank, then he might throw occasional shields on the raid when he has time, but it shouldn't be so often. And for most fights, he can specifically cast it on ranged (or melee) such that the other healers know where this backup (if any) will be. For most of the fights, it is good to know which priority in healing every healer will have, and that would solve the biggest part of your problem.

One case where you want (if not on emergency with tank healing) your disc priest to throw shield is when there is random secondary target on the raid (for example, ice tombs on KT). PW:S is then mostly efficient because that's instant, prevents deaths, and doesn't interact with other heals from other healers (typical case of six heals all landing 1.5s- 2s after the damage, and only one or two being effective).

Although it will most likely not change the outcome, I'd like to point out that you forgot to account for the different item cost of crit and haste per % (not rating, compare [Battlemaster's Alacrity] and [Battlemaster's Cruelty] for an example), so haste is cheaper than crit.

But there is even another problem when discussing haste as a stat to increase the Holy Concentration uptime and that is the fact that it works by increasing your n per HC cycle - which in turn will not be the case for every HC cycle. In other words, you increase your capability to have a higher HC uptime for a lot of cycles (Unless I'm wrong and we're spamming all the time - I was only able to test in 10-man so far). We trade in more regen for more invested mana.

Your crit chance in contrast will increase the HC uptime in any case with every n.

As I said, nothing new in the end, but I think we should treat haste and crit less similar as a regen stat.

Just a general question on the behaviour of POM on shielded targets. I've often seen '0 (XXXX) Prayer of Mending' flash across the screen when we've had a disc priest in the raid on a POM proc. Doesn't this indicate that POM procs even on shielded targets that absorbs all the dmg? I'm not sure, but hope some people can confirm wether a full PW:S absorb will block a potential POM proc.

Yeah, I guess I also suspected server-client lag could be the cause.

On the topic of mutually exclusive traits of PW:S and PoM I try to cast it on dps, preferably a retri paladin or shadowpriest(too bad SW won't proc it in 3.1). I agree with TheDoctor though, I wouldn't concern myself too much about not having POM bounce around because of shields. POM is an outstanding tool to control hard and wide damage with a single cooldown for the healer, I couldn't ask for a better spell as a holy priest. On predictable damage the holy priest has other and better tools to heal the raid with. And yeah, heal meters tell nothing. A raid leader should know when he has a good healer in a raid without having to use meters to confirm it. And a healer should know if he or she is doing a good job without having to use meters to buff their confidence.

My take on this is a little different. Quite honestly: it sounds like the real problem may be that your healers are more interested in outshining each other than in keeping the raid alive. Obviously, I'm an outsider and can't say that's true for sure, but the whole discussion seems a little disturbing to me. Why is there concern about a PoM not bouncing? Did someone die because of that?? A lack of a PoM bounce resulting in a death? That seems quite unlikely. Or is someone worried about how it makes them look less good as a healer? Hmmm....

To be fair, there is "a little" interference there, but it's quite minimal -- definitely a molehill and not a mountain. There is a much bigger interference on any PoM the disc priest might cast since these are likely to get blocked and consumed by the (probably) larger PoM from the holy priest. So the complaint seems ironic to me.

On the other hand, I would tell your disc priest that keeping borrowed time alive all the time isn't crucial either. If someone is in danger, and you have time, by all means shield them if it reduces the probability that they'll die. But the focus should be on whether it's needed to save raid members, not on whether it keeps borrowed time up just for the sake of it. Especially when tank healing, it is always a bit dangerous to leave your main target and do anything to another target. You have to wait at least a global cooldown before getting back on your main target. Borrowed time makes that decision less costly, but there is still some risk.

As others pointed out, shield is actually very good about not interfering much with others' heals. Maybe you can point out that if your disc priest threw out flash heals instead, your other healers would be "losing" a lot more healing to overhealing. PoM not bouncing is a comparatively small problem.

Yeah, I could have explained more but didn't bother because it does take some time to explain fully. Sorry.

The prior discussion about haste vs crit was in fact haste% vs crit%, and that is slightly different than comparing haste rating versus crit rating. However, the results would be the same when comparing ratings, except there would be an additional constant conversion factor thrown in. So you'd come down to saying that the value of a point of haste rating relative to a point of crit rating is:



where k is the % haste per haste rating divided by the % crit per crit rating. I believe the above function is still always less than 1, but it's more of a pain to show clearly. At least I don't have a simple argument for it offhand...

As for the number of casts per HC cycle not being true for every cycle, I guess I'm not sure I'm understanding you. Haste will affect the average number of casts you can make per HC cycle. Some cycles you may squeeze in more than others, but that shouldn't matter since we're looking at an average which won't apply to every (or perhaps any) particular HC cycle.

That's easy. ln(1-c) isn't greater than 1 until you hit about 73% crit (1/e). 1-C is NEVER greater than 1. 1+H is always greater than 1. K isn't going to do much more than tip where in the mid 80s-90s haste becomes better.

Crit % that high are just things we don't worry about.

Thanks for covering my lazyness and putting in the numbers.


You got my point and convinced me.

As a clarification: I was not sure about the statement "haste will increase your average cast frequency in an infite fight" (1), but yes of course, either you cast the same amount of spells per cycle or you have more. So only the worst case is the same cast frequency. My wrong assumption was somehow based on the current content. Most of the time, haste only makes you overheal less, because your heal will land 0.05s sooner on the target than your fellow healer's one.

Since (1) applies, haste does increase your throughput, your HC uptime and therefore the mp5 generated by HC. But when using this mp5 value, we still have some issues in terms of longevity. For example, when you look at this relation someone mentioned a few pages ago:



Since haste increases your mp5_in and your mp5_out at the same time, you will most likely not gain longevity. But you can win some throughput and cover the mana cost with the higher HC uptime (to a certain extent). Modelling this could be a pain though since it depends on a certain FH/GH/BH ratio.

Well, sorry to burst your bubble, but you need to read the background more. I had already shown in a previous post that:



is strictly less than 1 for all values of C between 0-100%, not just ones below 73%, and that a percent of haste is never, ever better than a percent of crit. Showing that a point of haste rating is never better than a point of crit rating involves showing:



is less than 1, which is somewhat harder to show but still should be true I believe.

edit: Here's a more formal mathematical argument: The derivative of -(1 - C)ln(1-C) is -1 - ln(1 - C), which has one zero at C = 1 - 1/e, corresponding to a maximum of 1/e = ~0.367879 for -(1 - C)ln(1-C). Since k < 2, -k*(1 - C)ln(1-C) is < 1. And since 1 + H is always >1, then -k*(1 - C)ln(1-C)/(1 + H) < 1. So haste rating is never better than crit rating at increasing your HC uptime.

Okay after reading all this, i'm confused cause I see contradictory talk. As a soon to be 80 holy priest, what stats on my gear should I be looking for?

Mp5>spellpower>Int>Haste>Crit?

You need a balance of regen (spirit, crit, int) and throughput (spirit, crit, haste, spellpower) stats so that you keep everyone alive while also not running out of mana.

Just to add another dimension to the whole haste vs mana debate: Theoretically you having haste saves your other healers' mana, since they'll have a higher chance to cancel their cast after yours has landed. Personally I believe the whole debate is pointless, since if you're going for high HC uptime you'll not have haste as your primary stat to go for.

From a Holy POV, my recommendation is to stop going for crit when you don't have any mana problems and go for haste / SP since the throughput gains are much higher. For the people talking about inspiration uptime on a holy priest, I'm actually skipping that talent entirely since we'll have a disc priest full time in raids and he/she will be able to keep inspiration up on as many tanks necessary.

The more interesting discussion would be what yellow socket you'd want to use for longevity. Intellect for pure longevity, or crit for longevity with added throughput. I'm pretty sure intellect wins, but not by how much.

The best parallell to draw is the one of caster classes with hit. Just as they start capping hit, before going for other throughput stats, you'll want to "cap" regen before going for throughput. As a healer this is ofcourse alot more complex, since it's very hard to theorycraft exactly where that "cap" is when your healing will vary in all fights, which adds the element of personal playstyle matching your gear set as a priest.

There really is no clear BiS gear set you can read up on and just go for, you'll have to tailor it around your own needs. As an example I very often have gone for max haste and max spirit, since that makes me able to burst heal to the extreme and then have sessions of extreme mana ticks when I'm not casting (this won't work in 3.1 but it is an example). Others have gone for max regen stats and very balanced throughput stats so they can keep a steady healing style through all fights.

As for the "noob" friendly way to gear, I'd go for as much intellect as possible for raiding and a mix of spirit / intellect for 5-mans. Big stats are the easiest way to play for a new holy priest.


Edit: merged posts.

Double edit: stay away from mp5!

So Int>spellpower>spirit>haste>crit>mp5

What I'm getting from this all is that there isn't a stat weight system like there is for other classes when it comes to healing. So one stat isn't necessarily "better" than another, you just need to cap the different ones in order to be effective.

Yeah thats not confusing at all :-)

Follow question though.... isn't mp5 going to become more important in 3.1 when they nerf mana regen?

On the contrary, spirit will become more important. Spirit will continue to give the same I5SR regen that it does now as well as the spellpower bonus through Spiritual Guidance. In addition Holy Concentration only boosts spirit-based regen. So the mana returns from Holy Concentration and thus crit rating will scale with spirit but not mp5 (this is also true now).

This maybe is a silly question but I'm having priests in my guild ask me what the haste cap is for Holy priests. Is anyone able to answer this for me?

The fastest your global cooldown can go is 1 second, so haste becomes less effective when raid buffs + gear haste bring you to more than 50% haste. It's still not useless at that point, since it gives you faster greater heals and prayers of healing.

The haste cap is 1639.5 haste rating or 50% haste, since that reduces your GCD to 1 sec. this is what most consider the cap, although it is true what the above poster says, more haste will lower the cast time of spell with a cast time above 1,5 sec even beyond that point.

This is yet another post asking about spirit vs. mp5 in 3.1. I am posting because I seem to have gotten into an unfortunate argument with my priest class leader about spirit vs. mp5 in 3.1, with me advocating for spirit, and him arguing that mp5 is much better and that spirit "sucks" because he loses a lot of mp5 when stacking spirit. I am here because I would like my reasoning for my stance to be critiqued.

For the record, I do not have access to PTR while he does, so this is what I have gathered from regularly reading these boards; I will admit that I am only theorycrafting here, but I do believe him to be wrong, regardless. My understanding in preparation for 3.1 is that spirit > mp5, even with the spirit/oo5sr nerfs and the HC nerf/IHC loss, due to the following reasons:

Spirit grants spellpower from talents, while mp5 does not. Furthermore, spirit also interacts with HC (while mp5 from gear does not), along with crit. During HC we gain 150% benefit from spirit. The frequency of HC procs and the overall uptime of HC relies not only on crit, but on play style, and from what I gather, the demands of the encounter and raid damage. As such, the value of spirit is heavily intertwined with HC uptime and thus cannot be as easily quantified as mp5. I am unsure of how the new Serendipity will change my play style, if at all (though I am assuming that it will), but even if I am being overly optimistic here, I believe that it will also interact positively and synergize with the other new/revamped talents to promote HC uptime as well. My understanding is also that should spirit be worse than mp5, it would only be marginally so, and that the spellpower from talents should also be taken into account when considering the value of spirit.

In conclusion, due to spirit's interaction with talents to potentially scale and increase, and mp5's more static nature and lack thereof, I have surmised that I will opt for spirit over mp5 in 3.1 even with the nerfs. Any input would be appreciated. Thank you.

Unless I'm really badly failing at math, with the new spirit coefficient, 1 point of spirit will be equivalent to .529/.264 O5SR/I5SR regen. Ignoring all other factors and assuming 100% I5SR time, you need a HC uptime of about 52% to make Spirit equivalent per itemization point to mp5 (at .4 mp5). At 10% O5SR, that drops to 42%. This is also completely discounting the effects of Kings and Spirit of Redemption. Assuming that you can keep up 50% HC uptime (and you should be able to), Spirit is better for regen - and that's ignoring the spellpower conversion entirely.

Yeah, spirit should be pretty clearly better for holy than mp5, largely because of the spell power bonus. It's a closer call for disc priests.

In addition, spirit regen scales synergistically with int and crit (assuming HC), which mp5 does not. So as other stats increase, spirit will become increasingly valuable while mp5 will not.

Kings and Spirit of Redemption also matter quite a bit, increasing spirit regen by over 21%: