The problem with a throughput argument of hasted penance would seem to me to be the cooldown, at least from a tank healing point of view. Playing disc at 3.0, I felt like penance was fast hasted or not because of its 3-tick nature. Sure extra haste is nice, but you're talking about a spell that already hits immediately and then twice more in rapid succession. The elephant in the room question is about that other 6+ seconds of cooldown. It seems to me with the other 6 seconds you'd be casting either gheal or fheal waiting on cooldown. Ignoring PWS cast/cooldown issues, wouldn't flash heal spam in that interim period be optimal for returning mana via rapture and minimizing overheal (especially given you don't have the deep holy talents that improve gheal)? I guess I don't see how haste is the "reason" for dropping gheal from a disc rotation. It seems that decision is made more by talents and cast style, the haste is just a bonus.

edited the rest, 3 hour essay exams = jumbled incoherent thoughts.

Agreed. As haste doesn't reduce spell cooldown, I really don't see how 20% haste is going to help you stop GHealing. It's partially a matter of style, and I'm still trying to figure out the optimal pattern, but cutting out GHeal as a result of gaining haste makes no sense to me. Infact, cutting out GH no matter what your stats is pretty much wrong. You're fine to FH spam (if you can sustain it..) while your Penance is available, but if it's on CD I'm of the opinion that you need to be GHSC (greater heal stop casting) because FH doesn't have the throughput to catch up from a big spike (at least in the theoretical "hard" fights that I'm hoping to see in the future).

Ah, I think I understand now - your point was more that you wouldn't have to GHSC because a reactionary Penance would heal the tank up quickly enough not to require the GH throughput?

The top end Holy talents actually buff Flash Heal more than Greater Heal. It's the bottom end ones that buff Greater Heal unusually. Of these, only Improved Healing is really of consequence - it is very likely that a Discipline Priest wouldn't take Improved Healing.

However, Greater Heal is still more efficient and about 25% more throughput (@ +2500) than Flash Heal even with 0/3 IH (and ignoring Rapture returns). So in a solo-healing environment where you're not particularly concerned about other people sniping your heals, Greater Heal is probably a better choice.

In a multi-healer environment, you have to accomodate the fact that Greater Heal and Flash Heal are markedly worse than what other healers bring to the table.

With only JotP and Healing Light, Holy Light is about 35% more efficient/66% higher throughput than our non-Rapture Greater Heal. Note that Illumination is equivalent to 60% * critical rate cost reduction and Paladins have +7% more critical than Discipline Priests.

Ignoring IWS and TW, Lesser Healing Wave is about 37% more efficient/3% less throughput than our non-Rapture Greater Heal. Note that IWS is equivalent to around 80% * critical rate cost reduction and Shaman have +5% more critical than Discipline Priests.

Comparison to Druid healing are a bit too complex for this sort of handwavey sum-up, but suffice it to say that Druids can outperform Greater/Flash Heal as well.

So in a multi-healer environment, the best spell for you to cast in between Penance/PW:S uses is... nothing. Someone else has a better (in most cases, a lot better) spell to cast. This leads me to believe that the best technique would be to cast-and-cancel your Greater Heal, only letting it resolve if the tank has taken truly huge lumps of damage.

It ticks twice, 3 seconds and 6 seconds after the PoH lands, so yes, it would be useless for Loatheb.

As an aside, this glyph appears to be gaining double benefit from Focused Power. Obviously the +4% healing increases the damage of the Prayer of Healing, but it is then applied to the Glyph HoT as well - test I just ran, the PoH healed for 3274, so you'd expect 327.4 from each tick of the Glyph, but instead I got 340 and 341. 681 is exactly 4% more than 20% of 3274.

So I guess that makes the Glyph 4% more attractive to Discipline priests. You may well consider that "4% of not much!" though.

On disc raid healing (2-3 posts back):
I'm sorry to see that PW:S and DA can't be parsed by WWS yet. (or parsed easily. If someone has a way to track the absorbs and hence the mitigation and rapture returns, that'd be VERY appreciated) A 6k PW:S can be an extremely efficient "heal" on a raid member. I'm thinking of a case like Malygos phase 2 where back-to-back Arcane hits from Scions usually spell doom. The shield can help to buy a bit of time for the rest of the raid to get that guy up.

So maybe I'm seeing a role where a disc priest can be focused on the tank a great deal, and throwing out a PW:S on a raid member between Penance CDs and Weakened Soul debuffs. Giving someone 6k hp instantly can sometimes make or break a raid.
And let's not forget that Disc is probably the fastest reactive healing - barring NS. Shield+hasted Penance+hasted FH=a ton of healing fast fast fast

Disc would almost surely prefer a PW:S glyph. I could have sworn I was getting ticks of it the other night, but Nib and others pointed out some pretty indisputable math on the topic.

Sadly it comes to the same. The final glyph slot basically falls to what you use, I tend to find atleast some situations were a POH is useable, so the glyph works. If your casting PW:S a bunch thats easily the right glyph. The argument will likely end up as PW:S for disc, SOR for holy since PoH will quickly get overscaled by COH in any situation, and maybe... just maybe one day it will proc at the right time to save a boss kill.

Changing gears.
Distomos and I discussed this briefly, but I'd like to gauge reactions as a whole.

As the only holy priest, (assuming a large portion of BIS gear) would dropping DS for GS be viable? losing 6/9% intel and mental agility along with DS for Divine providence and Guardian Spirit.

I'm not sure about being the only holy (since Cahrin runs Disc full time), but I have yet to have Mental Agility or Mental Strength (you can see in my armory), I'm full blown holy. I love Guardian Spirit, it has it's uses which is exactly what I like it to be.

Personally, I feel that if your regen is getting to a point where you are comfortable with it you should be heading to Divine Providence., the extra 10% on those spells is pretty substantial.

Right, except you're completely ignoring any gains from overhealed DA, which is the entire point of FH overheal spamming. GHSC provides no benefit if the tank doesn't get hit, which is why I prefer not doing it unless I think I'm going to need the throughput (i.e. if Penance is down).

I'm pretty confident I can solo heal through most of Naxx aside from a few fights. My healer leader just never gives me the opportunity.

By all accounts those are amazing talents, but so are the others. For the sake of simplicity lets say that MA/MS=DP/GS with DP/GS pulling slightly ahead at higher gear levels.

Which really makes DS the tipping point. assuming no disc priest is present, 80 spirit is a chunk of regen (exactly how much depends on the rest of your gear) and 16 spell power.

I'm not sure what you're talking about. You don't actually gain mana by spamming Flash Heal. Both DA and Rapture work off of effective healing, so throwing a Flash Heal on a full health target just means you spend full mana for no healing.

And when I last checked, DA wasn't returning mana anyway via Rapture anyway. If it now is, then you could potentially argue that since the negative impact of overhealing is doubled (from the initial Rapture return as well as the smaller DA Rapture return) so a faster spell would be more worthwhile.

None of which changes the fact that Greater Heal and Flash heal really, really suck compared to what other healers use.

"Critical heals create a protective shield on the target, absorbing 30% of the amount healed. Lasts 12 sec."

Throwing FH at full health is a chance to crit, thereby reducing the next hit by around 2k.

I've yet to really see conclusive evidence either way on whether it's returning mana. From anecdotal evidence I'd guess it's bugged in the sense that it's not stacking properly, but I'm pretty sure I still get mana back. I've been meaning to test it for myself, so I'll probably do that tonight.

No idea what you're trying to say here. How does DA returning mana double the negative impact of overhealing?

entr0p -
Thanks for posting the WWS. It's nice to have actual data rather than hypothetical models. That said, I agree with constantius: the fact that you were able to do this healing with your chosen gearing (high crit) tells us nothing about whether some other choice of gear might still be even better.
One observation is that Circle of Healing, which is your primary heal throughout this run, does allow crit to be a throughput stat. Between the small size of each heal and the fact that you have so many chances to crit per unit of time, you end up invalidating two of the objections to crit as throughput. And you get the effective regeneration from Surge of Light as well.
Looking at the log summary, it seems that Improved Holy Concentration and Serendipity aren't doing that much for you here. But then, your role is raid healer on this run, so the spells you're casting most of don't showcase those talents.
Anyway, I wanted to focus on one particular fight that emphasises AOE heals: Sapphiron. You cast Circle of Healing about 60 times (hitting 331 times) and Prayer of Mending 13 times (for 48 bounces) in this 5-minute fight, along with a few free Flash Heals and other spells. This gives us a good starting point for estimating what you would gain or lose from various stats, and should definitely give crit a fair trial, since CoH and PoM both work well with it. I'm going to ignore the few other spells you cast just to simplify the analysis. It looks like you spent somewhere between 22 and 28 seconds outside the 5-second rule; I'll go with 8% (of 300 seconds).

The baseline (per-target) healing coefficient for Circle of Healing is 40.2%. Multiplying for Spiritual Healing, Twin Disciplines and Divine Providence gets us 51.1%.
The coefficient for Prayer of Mending is 80.8%. With bugged scaling, this goes to 107.3%.
Your observed crit is 30%, which adds another 15% to these values.

331 * 0.511 * 1.15 = 194.5, less 21% overheals is 153.6
48 * 1.073 * 1.15 = 59.2, less 25% overheals is 44.4

So for every point of spellpower we could expect to gain 198 more total healing in this fight just from CoH and PoM. The free Flash Heals push this closer to 210 healing per point of spellpower.

The observed heals per mana for CoH in this fight is (1128977 / (60 * 927)) * 0.79) or 16 healing per mana.
For PoM, the value is (263987 / (13 * 463)) * 0.75) or 33 healing per mana. Weighted averaging gets us 19 healing per point of mana (neglecting the additional free Flash Heals in this case). With 60 5-second periods in the fight, every point of mp5 gets us 60 mana and therefore 1140 additional possible healing.

Every point of crit rating gets us an additional 1/46 chance to crit. Using base values of 1800 and 3400 for CoH and PoM (these seem to be approximately correct based on the WWS), we get for one crit rating an additional

331 * 0.50 * 1800 * (1/100) * (1/46) = 64.8, less overheals is 51.2
48 * 0.50 * 3400 * (1/100) * (1/46) = 17.7, less overheals is 13.3

So every point of crit rating gets us 65 or so more total healing. But we also need to account for our Surge of Light chances. You used 14 of 31 SoL procs in this fight; let's go with 50% utilization.

Chance of SoL proc on CoH is (1 - ((1 - (crit%*0.5)^(number of targets)). Using 5 as a typical number of targets and 30% as our current crit percentage, we get a 55.629% chance to get SoL per cast of CoH. Rather than try to take the somewhat scary derivative, I'll just recalculate based on a crit chance of 30.0217% (one point of crit rating): it comes out to 55.658%, or an additional .029% chance of SoL per point of crit rating at this threshold.
The value of an SoL proc is 4796 (avg Flash Heal) * 50 % (utilization) * 0.79 (overheal adjustment) or 1894 healing.

60 casts * 0.00029 * 1894 = 33 healing via CoH/SoL combo, per point of crit rating

For PoM, the calculation is slightly differen: we'll count SoL proc chances based on the number of bounces, which means it's just (crit% * 0.5), which is to say that we get an additional 1/92 of 1% chance ( = .0109%) per point of rating.

48 bounces * 0.000109 * 1894 = 10 additional healing per point of crit rating

So accounting for these factors, each point of crit rating gets you 65+33+10 or something like 108 additional healing.

Haste appears to be fairly useless in this fight, since most of your casts are spaced 4-5 seconds apart, and the issue seems to be endurance rather than burst healing.

Let's also look at Spirit. With the 5% bonus from Spirit of Redemption, each point of spirit on gear gets us 1.05 points of usable spirit. Your raid-buffed Intellect is probably about one thousand (just looking at your gear). So each point of spirit (gear) will get us 0.873*1.05 = 0.917 OO5SR regen, and 0.275 mp5 (with Meditation). Using weighted averaging based on the 8% OO5SR time observed in the log, we get 0.326 mp5 per point of spirit. Further, we get 0.25*1.05 or 0.2625 spellpower per point of spirit. Thus the total value of a point of spirit (incorporating earlier values for mp5 and spellpower):

0.326 * 1140 + 0.2625 * 210 = 427 additional healing per point of spirit.

Bottom line: even with spells that make crit look good, crit doesn't look that good. Values:

MP5 = 1140
Spirit = 427
Spellpower = 210
Crit = 108

I also think in looking at this WWS that a strong case could be made for dumping the points in IHC into Mental Agility instead. But that's probably better left for another post.

Cahrin I believe Kortar is referring to the phenomenna of DA only proccing of effective healing but not over healing, something that I'm inclined to believe through my own experience. It's generally a bad idea to take tooltips as gospel given the abundance of bugs in this game. It is also my understanding that DA isn't returning mana through Rapture, at least not reliably. Maybe it is prone to the same Rapture bug as PW:S? Bearing both those things in mind I believe the final point was that your efficiency is penalised more as you overheal for higher amounts - the overhealed amounts are not included in Rapture calculations and do not contribute to DA through crits (though the effective portion of the crit will).

Bearing all of that in mind the only real incentive you ever have to over heal is keeping Grace stacked up. As disc you have good preventitive options through PW:S and DA when it happens but also good reactionary tools via Penance and Borrowed time. Don't be scared to let the tank dip before you cast and remember you're losing the benefit of a lot of your talents by overhealing.

As I now understand the issue, I ran some tests on live. Here are my findings:

- DA works fine with overheal with respect to both the absorption amount and the return from Rapture
- PW:S and DA seem to stack, and the Rapture returns are still active
- There's a bug (reported and known, just confirming) with Rapture in which the mana returned is based on the target's mana pool - this essentially means you get nothing back from PW:S or DA on a Warrior. I hadn't noticed this before because our MT is a Paladin, and when I got Rapture ticks from shields, I didn't bother thinking about the actual amount.

The third point I tested with three people: myself, a DK, and an Enhancement Shaman. I got around 15 mana for around 350 absorbed with DA/PW:S on myself, half the from the Shaman, and 0 from the DK.

So yes, in the case that Rapture is horribly bugged with respect to regen, it may not be wise in terms of mana regeneration to spam FH on a full HP tank. However, the absorption still works on overheal, so from a throughput point of view, it's still useful.

Edit: I have logs available if anyone's interested :P

The problem with crit (for Holy, not necessarily for Disc) is that while it's a better regen and efficiency stat than Haste (simply because Haste is not a regen stat at all), it's still not very good. If you want endurance, regeneration, mana efficiency, you are better off with spellpower, spirit or mp5, or for that matter a switch to Disc with Rapture and a giant mana pool that has these things in spades. See the numbers I put together for entrop's WWS of a 5-minute Sapphiron fight.
If on the other hand you want throughput, crit is both unreliable and not quite as good as spellpower or haste even on average. I'm not saying you shouldn't take crit if it's cheap; really the problem comes from the fact that you need so much rating to get 1% crit. Things like Holy Specialization or Moonkin Aura, that get you 5% easy, are obviously very good. It's just that if you want to get 5% from items, the cost in other stats is too much.

This suggests you haven't spent more than ten seconds thinking about this. Holy Specialization plus standard raiding Int plus baseline crit gets you to 14% without any buffs. A Holy priest would need negative crit rating to aim for 10%. I would expect 20% rating in a raid setting even for someone like me who values crit gear very very low.

Can you provide and example (like WWS) for a Holy priest? Inspiration uptime can be very valuable, but that isn't usually the Holy priest's role since Disc priests can do this more effectively. And the other benefits of crit in the Holy tree are underwhelming. Look what it costs you to get from 20% to 30%, calculate the actual benefits, and try to show how it's worth it.

ok i'm curious....

if crit got such a boost why is so much of the BiS gear stacked with haste. and the BiS gems +haste for yellow and +power/haste for red... no crit?

i'm around 16% crit and 14% haste atm. pretty solid imo but i don't even have half the BiS gear (which almost all have more haste than what i'm wearing)

how can the rule of thumb be 10% haste but have all the top gear have an extra... almost 8-10% extra on haste. but be subpar on crit

Also, haste...I'd definetly aim for a soft cap amount, 10% for that magic 15% with the totem.


We're raiding tonight, I'll get some from that.

But in the future, if you want to scrutinise what I, or others say, at least read the context its in so you don't appear the fool.

Because cloth gear has been scaled for Priests, Mages & Warlocks together, not for a singular class' needs. Thus you have an assortment of stats thrown together so they appeal to all of the aforementioned.

BUT ... for BiS 2h staff... is the spire... with 99haste on it. instead of Damnation with 91crit on it..


OR both rings have haste? i'd agree with the badge one but the second one should be Lost Jewel to at least be semi even on stats to reach that rule of thumb mark on crit

First of all [Damnation] is bugged. It should have more spell power that it does. If you notice the [Spire of Sunset], which is of the same item level as Damnation, has 461 while Damnation only has 408 (should be 461).

BiS gear, if you read what Nid posted below his BiS set, is objective at the moment. He said the gear list he had was in his opinion.