@ Perkyone

multi-doting and cleaving on protectors lowers raid dps it does not increase it. A DPSs job is to maximise effective damage. For a healer your job is to maximise the amount of healing you put out. If a DPS fails or is tunneling the boss and ignoring raid mechanics then its my job to keep them alive.

Top parses are at the top because the players have the gear and the understanding of the spec to produce maximal HPS. That is the only reason. They are not playing ineffectively or hold their raids back. Its the other way around.

Bubble spam is not heal snipping, its the most effective way for discipline to heal. If the other healers heal less, then there is nothing wrong with that. Less incoming damage = higher healthbars = lower chance of random deaths.

Now on some fights where the boss has a damage buff atonement can be be a powerful tool, especially on 10man and those hc modes where the boss has big damage buffs. But even in those 10man hc fights, many top disc priest parses still show clear evidence of PoH spam and it still has high overheal. Atonement still has extremely high overheal in low damage parts of the fight and the bigger it gets the worse crit becomes.

Details for Alaxana - 02-01 20:04 - Nightmare Asylum - World of Logs
Details for Claire - 15-12 17:59 - 灵�*�圣域 - World of Logs
Details for Tîgér - 19-12 18:23 - organised confusion - World of Logs

I already explained why proc haste increases the value of haste rating. If you disagree post your calculations or your explanation, otherwise its just your opinion with nothing to back it up.

@ Poloptress

In 5.1 mastery had high scaling with spirit shell, but it started at a low level. So the graph started low but had a high slope. In 5.2 mastery starts high, but now it slopes really slowly. That is why the nerf didnt affect PoH much. It has a higher baseline benefit from mastery, but additional mastery adds bugger all to spirit shell. Shedding 10% mastery reduces spirit shell by a tiny amount. So you can remove all your mastery and you will notice a comparatively small decrease in spirit shell healing. Thus you lose little by reforging your mastery away and you gain lots by replacing it with haste or crit.

Haste is unaffected by overheal. What it does is it multiplies your effective DPS. If you cast 10 spells with 50%OH and 10 spells with zero OH without haste you will cast 11 spells with 50% OH and 11 spells with 0 OH with haste, which is 10% more HPS.

Haste is better for stacking aegis. Casting 11 PoHs with 25k aegis per target each is a bigger stack that casting 10 PoHs with 26k aegis per target each.

Breaks in casting don't affect haste either. If you are casting for only 80% of the fight you will still cast 10% more spells in that 80%, which is still 10% more healing.

What can affect haste scaling is if the proportion of low and high overhealing spells changes. If for example you can't get an extra spell in during the high damage phase and you only get extra spells during the low damage phase, then haste will give you less return.

So in reality the only thing affecting haste is breakpoints, but that works both ways and disc by having proc haste can easily tip the scales to favour haste.

So as long as you are not limited in the number of casts you can make by other concerns (e.g. mana) haste doesn't care about anything that people think affects it. Breaks in casting mean nothing if they are enforced by the raid mechanics. They actually make haste more valuable. Overheal means nothing and again the existence of high and low damage phases makes haste more valuable potentially if you can hit the right breakpoints.

Thus as long as you can exploit those break points by pre-cast, using proc haste and so on, haste can actually end up giving a higher than nominal return and all the things that most people think reduce haste effectiveness actually increase it. If you have to run from something then haste can make it easier to get casts off and there is a better chance that you won't have to cancel a cast.

Cooldowns also don't necessarily affect haste negatively. Some cooldowns are there to increase the healing of other spells rather than provide a higher HPS benefit themselves. Rapture and AA/evangelism are an excellent example. The presence of such cooldowns INCREASES the effectiveness of haste. Even cooldowns which do increase HPS can get a positive impact through haste, because it can make it easier to use them at the right time. Cascade is a prime example.

So basically haste is a very very stable stat and the only limits are mana and breakpoints. That is all there is to it.

For spirit shell even if you are actually hitting a break point haste is still helpful. We don't live in a perfect world, sometimes cooldowns get unsynched, or raid mechanics get in the way. Haste can reduce the loss from effects like that.

When comparing mastery with haste the thing to be aware is that mastery does two things differently from haste 1) It increases in value as OH gets higher 2) It increases spell HPM.

1) Is not a problem, because just like intellect, haste is still better than mastery for aegis even at 100% overheal. There is no point where mastery becomes better and in fact even in 5.1 haste is better. 2) HPM however is a key concern. If you are HPM limited haste is pointless.

At the moment mastery has a sufficiently high value and spirit a sufficiently low value to make spirit/haste stacking pointless, but as soon as 5.2 hits, spirit is going up hard and mastery is going down like a rock. Crit was bad and it will always be bad, so there is little doubt at this point in time stacking spirit/haste is much more effective than anything else for bubble spam, as both crit and mastery are just terrible for it. Spirit will be better than intellect for disc when 5.2 comes out, while you are mana limited.

Haste is also better for atonement spam, but if you use atonement a lot you reduce the need for spirit, which means you can replace some of your spirit with int (or crit if you cant get more int), if you end up with excess mana.

I am planning to do all I can to increase my static spirit as much as possible shedding any proc spirit after that I will reforge all the mastery my mana regen can support to haste.

Now you are certainly quite right to say that this need for extra spirit devalues haste somewhat, but the values for crit and mastery are so low and spirit is so strong now, because our mana regen will decrease sharply that they make this a moot point.

No point in bothering with crit. Even in fights which specifically buff atonement haste is better and extra mana can always be used with enough haste.

You dont understand at all do you? Cleaving and multidotting on any multi target fight is a dps increase on the meters. The problem with that particular fight is that it is a clear case where the extra effort and damage goes to waste. Over heal is not "effective" healing, just as cleaving and multidotting on that fight is not effective damage.
Long story short, casting speed procs like borrowed time and power infusion are multipliers. A haste junkie popping PI gets 20% increased casting speed and a crit junkie gets the same boost. Yes your spell haste will be different and one will go up by a slightly larger number, but both players got a 20% boost RELATIVE to their previous state. PI will only ever make you 20% better than you were before.

To illustrate this lets say I take off all of my gear. I have 0 haste rating. For this example lets say while naked i can do 10k hps with greater heal if i spam cast it. Now I pop PI, which for the duration increases my hps with greater heal by 20%. I am now doing 12k hps. Now i put my gear back on. lets say i have 10% haste. When I pop PI time now I get 22% haste. Lets say for example my greater heal is 55k hps with all my gear on. When I pop PI that goes up to 66k. Because 55/1.1*1.32=66 it is still only a 20% increase relative to the previous state. Having established that.. it doesnt matter what level of haste you have, a casting speed buff will give you the same boost. The only time you could ever conceivably get a tiny benefit from a temporary casting speed buff (over someone with less haste) is if you hit a break point of some kind.

That's not really a very fair comparison. Because of DA, a PoH that has 100% overheal still has value unless there is no raid damage before it expires. For Protectors, cleave DPS is of no value because it is entirely negated whenever one of the bosses dies (discounting things like cleave possibly increasing the chance for procs that can increase your primary target DPS). There hasn't been an encounter since... Chimaeron? I think, that is a fair analogue for Discipline, where any healing above a certain threshold is entirely negated.

Just a point both of you seem to conveniently ignore:
Disc has two ways of over-healing . One is direct over-healing that's also shown logs. The other one is shields that run out/get capped instead of getting broken. Assuming damage comes at somewhat large intervals, you'll have one form of overheal no matter whether you use PoH to stack DA (heal gets wasted) or use PoH to heal players back up (DA gets wasted). In the end it's about keeping the party alive and if your mana allows for it, there's no reason not to pre-shield people to give them a higher EHP when damage occurs. It also takes stress from healers when damage actually occurs, possibly allowing them to use more mana efficient heals instead of spamming high HPS/low HPM heals just to keep the raid alive.

I understand perfectly. It does not matter what it says on the meters, its not effective DPS, the actual raid dps on the protector that you are going to kill first is lower when you do this. Overhealing isn't effective HPS and its NOT shown in the meters. All that you see in the meters is effective healing. Even if you overheal tons with PoH you still do a lot of healing. That healing is effective healing, because no one intentionally sits in the fire to boost your WoL records. That was damage that occured in the fight and you healed it. Actually its better than just healing because it reduces incoming damage. If they do sit in the fire by accident having a 100k aegis shield is going to do much more for saving them than a 35k smite hit.

Minimising overhealing is of no consequence. The only thing that matters is who produces the most HPS and reduces the likely hood of someone dying the most. When the necessary conditions for bubble spam are fullfilled it is 100000% better to bubble spam and only use atonement to stack evangelism. Bubble spam has high overheal and that is fine. Its not wasted healing, its not wasted mana, its not pointless, its not "padding" the meters. Its the most effective way to heal, because it produces the most HPS.

You just explained yourself why haste increases the value of temporary procs, you just didn't realise it: Proc haste helps you hit breakpoints and its not a tiny increase its a humongous increase. Burst high damage phases is an important source of healing and getting 8 spells instead of 7 represents a very high increase. If 30% of your healing comes from high damage phases and you are getting 8 instead of 7 spells in each that is an increase of 4.2% of your total healing, which is massive. Haste on gear makes it easier to use proc haste to get breakpoints that you would not be able to hit otherwise. Thus haste increases the value of proc haste, rather than devalue it, because of the multiplicative stacking. Crit buffs and mastery buffs in constrast devalue crit and mastery on gear.

I already explained this but obviously you can't read long posts carefully:

@poloptress: I already took into account that mastery increases in value with overheal.

But I guess you are right. I probably made a mistake somewhere. But lets see

R= Rate of aegis stacking = 0.5*(base)*(1+crit)*(1+mastery)*(1+haste) = k*(1+mastery)*(1+haste)

Change 60% mastery 5% haste to 40%mastery, 16% haste

(1+0.6)*(1+0.05) = 1.68

(1+0.4)*(1+0.16) = 1.624

3.5% loss of aegis, but 11% increase in the healing. I thus need the healing from PoH to be 31% of the aegis part in order to be equal. Since typically aegis is 0.7% of the healing part I would need 80% overheal before mastery becomes better.

With average overheal of 50-60% haste is much better.

Reducing mastery does not increase the absolute amount of overheal, it only increases the amount of aegis, and even though I was not quite right in thinking that haste is better than mastery for aegis, it still takes a very high amount of overheal to before mastery becomes better. At the values most ppl have haste is much better than mastery.

However this ignores a key factor. Aegis is also not necessarily absorbed completely. You can keep it alive when you bubble spam, but breaks due to raid mechanics make it harder to keep aegis alive. Haste increases your ability to keep aegis alive, which can reduce the loss of aegis to expiration.

You can't tell for sure that it will be like this the 50% overheal phase is a phase where there is constant low level damage the overheal will pretty much be the same there or perhaps it will be lower because you have a faster cast speed. In the high damage phase yes the last spells have more overheal, but because of the haste you can almost certainly get an extra spell in the low overheal stage, especially if you use proc haste. You are assuming that its the last one that is added by haste, but that is not necessarily the case, You would have you could have 3 at 0%, 5 at 40% and 4 and 100%, which would actually reduce your overheal.

That is why Haste is completely unaffected by overheal. If anything more haste helps you get there first. The only thing that matters for haste is being able to hit those breakpoints and with proc haste effectively at will for disc there is a big range of breakpoints that you can heal.

I never assumed all the breaks in casting are in one block . Whether you are moving, waiting for events pausing to make a decision, to drink, to wait for a cooldown it makes zero difference. Those are the same with or without haste, subtract all of these and what remains is your total casting time. If you cast x spells in that casting time with 0% haste you will cast 1.1*x spells with 10% haste. Exactly 10% increase.

Far from decreasing the value of haste those "microbreaks" increase it. You are more likely to be able to finish a cast and you can potentially use proc haste to increase the PoHs you get in there. The Flanking boss in spirit kings is a good example, haste makes it easier to finish a cast. If you are kiting something for example it is much easier to get casts off when you have haste. The only kind of break that affects haste value is a break because you are oom.

The number 1 mistake made by ppl thinking that breaks other than oom breaks affect haste. That is just another breakpoint. You have a window from when you started casting to the point where the "important action" will come. If you had haste you could have squeezed an extra cast in there, because your previous cast might have finished 2.3s before the important event. Or if this is with haste then without haste you might have had to cancel your previous PoH with 2.2 seconds remaining, which is an even greater loss in casting time. Proc haste helps you extend the range at which you can hit windows. For small windows sometimes haste can be a problem, but for large windows it is not.

So neither overheal nor breaks in casting affect haste. There are only two things that matter, (a) are you managing to hit important breakpoints (b) do you have enough mana. That is what determines the value of haste.

Going back to a point from a few posts back about the viability of atonement on elegon.

Disc priest, rank 2 on wol, solo healing heroic elegon, 80% of their healing done is atonement. Their dps spikes up to 300k for the last phase.
Healing done - 09-01 21:30 - Insignificant - World of Logs
Another solo heal as disc on heroic elegon, very high atonement.
Healing done - 09-01 19:07 - Wólfmother - World of Logs
Plenty of other examples on wol of very very heavy atonement, even in the last phase.
Very impressive in my opinion. Doubt we will see any solo heals from any other specs for a while, for now its probably only possible with heavy atonement.

I would argue that the same effective healing can be done with out endangering your raid members while also dealing damage. It seems like you have one strategy for all possible scenarios.
No damage: pom on cd, cascade on cd, cast poh
Low damage: pom on cd, cascade on cd, cast poh
Moderate damage: pom on cd, cascade on cd, cast poh
High damage: pom on cd, cascade on cd,cast poh, pop a cooldown

So you get 1 extra cast during spiritshell once every 2 minutes. You also dont get that 1 extra cast for the other half of your spirit shells, so youre 14% better for 15 seconds every 2 minutes? Uh cool? Come 5.2 that won't even matter (because ss wont be a throughput cooldown) unless the mechanic requires huge effective health for the entire raid.

No need to get snarky. I didn't insult you for not doing your math correctly with the healing buffs.

The logs you posted show maximum 250k very short spikes during heroism. The other log is 150k spikes. Look at overheal on atonement and especially on crits. Look at divine aegis "pverheal".

Now compare it with this one I did yesterday taking into account that the gear I use for this fight is 493 ilvl and without a sha touched weapon.

Healing done - 09-01 23:34 - After Hours - World of Logs

I stacked Spirit shell with Arch just before we went in, then stacked atonement straight up + an extra penance after procing archangel, then straight PoH spam. Look at overheal on atonement, look at the aegis "overheal" and there is no difference betweeen 10man and 25man for this since the raid is stacked and taking heavy damage and PoH only hits 5 targets. That is without heroism too.

Basically, solo healing it with atonement is for speedkills. Atonement is generally used a lot on elegon, because the damage buff increases its HPS, and because PoH just doesn't work for big chunks of hte fight. In the last phase you can without question hit much higher HPS with PoH spam (no spirit shell, no PWS).

When the conditions for bubble spam are fullfilled its better than anything else so I bubble spam but what you said is definitely not the strategy I am using

Low damage, --> Inner will, PWS/PoH spam with holy fire and penance on CD + 1 smite, pop archangel on CD if no high damage phase is coming within 25s. PoP spirit shell on CD with archangel unless a high damage phase is coming before spirit shell will be ready. Cascade only just before small spikes if it won't overheal too much. PoM during movement. Chain inner focus with spirit shell

Moderate damage, ---> Inner fire PWS/PoH spam with holy fire and penance on CD + 1 smite, archangel on CD spirit shell on CD chained with archangel, powerinfusion and borrowed time, unless a high damage phase is coming before the combo will be ready again. Cascade just before a spike hits, PoM during movement. Chain inner focus with spirit shell

High Damage ---> Inner fire, min PWS max PoH, with holy fire and penance on CD, archangel on CD, spirit shell chained with archangel PI and borrowed time. Cascade on CD, PoM during movement, Chain inner focus with archangel and try to hit low health targets

Very high damage --> inner fire, straight PoH spam with prestacked archangel, PI, borrowed time only for spirit shell, cascade on CD, intellect potion. Use inner focus on CD.

When the conditions for bubble spam are not indicated (i.e. not enough targets hit by PoH, constant random hits) --> PWS on tank, atonement, cascade, PoM spam, PoH on 3+ targets only if low overheal, Flash heal/gheal for high deficits and renew chained with borrowed time and glyph.

High damage buff on boss --> atonement spam + PWS on tank, archangel on CD. Chain PI with archangel and spirit shell if it will hit enough targets, Chain inner focus with spirit shell. During archangel PoH spam if it will hit enough targets with low overheal otherwise continue atonement spam.

=======================================

You don't seem to realise. 14% boost for 20s with PI, is very large compared with the benefit of the haste on gear needed to achieve it. More importantly this is a high HPS section, so its value to overall HPS is much higher than the low damage sections. The important thing is taht you can do 14% more HPS during power infusion with haste than you can without it. Thus haste dramatically improves the value of power infusion. In real terms the benefit is not that large, because you won't reliably hit all breakpoints at 100%, but what is left over is a very significant buff to proc haste. Or if you want to view it the other way a very significant buff to the value of haste on gear.

Without proc haste, haste on gear would have not be as good, because the range of windows where you can hit a breakpoint would be smaller.

The question is whether atonement is better HPS at the last phase and looking at the last phase only since I suspect we have similar gear:

Healing done - 12-12 21:50 - Celestial Rebirth - World of Logs

Healing done - 09-01 23:34 - After Hours - World of Logs

I zoomed your numbers in the last phase, which is what we are discussing. The 350k spike is your spirit shell. Your atonement/cascade is only hitting a 250k HPS spike. Although my numbers don't look that much different from yours, notice the lack of heroism. No heroism used at all.

The rest of the fight is immaterial, this is hc elegon not normal and the role you are assigned in your raid can limit your access to big chunks of the healing.

Also feel free to look at what your atonement is healing:

Expression Editor - 12-12 21:50 - Celestial Rebirth - World of Logs

I pulled out the numbers quickly

Total healing: 3852670
Total overheal: 606423
Total healing on army of the dead: 480652 (12% of total)
Total overhealing on army of the dead: 10032

Effective healing: 3372018
Effective overheal: 596391
Overheal rate = 15%

Effective healing as %total healing = 76%

In contrast I healed pets for 5.8k and that was again through atonement.

I think this is a fairly dramatic demonstration of how much more effective PoH spam is.

Here is the PoH spike versus the atonement spike:

Healing done - 12-12 21:50 - Celestial Rebirth - World of Logs

Healing done - 09-01 23:34 - After Hours - World of Logs

And remember you have a 30% increase in casting and I don't. PoH spam is dramatically better HPS than atonement and on top of that it does not spaz out and heal pets.

Incidentally check this out:

[02:03:52.992] Shaarra Atonement Nadrain +*389259* (O: 64424)
[02:03:57.424] Shaarra Atonement Mouninja +*265011* (O: 101166)

370k and 450k crits get zero benefit from mastery and over 50% of the crit value is lost to overheal

10 item levels difference in our gear. I'm done arguing about this. It doesnt matter if I show you math or logs, you just stay in your "stack haste spam poh" bubble (pun intended).

That is because the maths and the logs show that PoH is more HPS than atonement.

These are my unbuffed values, using 0 haste cast times.

With 8x modifier on atonement (5 stacks + archangel)
Penance: 26700*(1+0.12*(1+1.6))*3/2*8 = ~420k HPS, but in reality it is only 375k due to divine aegis overcapping: 430k crits still leave only 160k aegis
Smite: ~165k HPS but in reality it is only 150k
Holy fire: ~242k HPS but in reality it is only 212k
PoH +archangel = 223k HPS but it is buffed more because of inner focus.
Cascade (10man - 30k average hit) = 300000*1.5*1.12*1.2 = 400k HPCT

PoH+HF+Penance+cascade on CD is clearly more HPS than pure atonement, especially considering that you will only have on average 3.2 stacks of evangelism or so.

The only advantage that atonement has is 10man specific: Stacking evangelism to 5 ASAP with HF+3smite+penance will bring 7 ppl to 100% and any crits will also overcap aegis, which of course makes it pointless to cast a PoH until the debuff stacks high enough to allow it.

The original argument you made was that atonement is more HPS than holy. The logs show that this is not the case. disc produces more HPS overall, but holy produces more HPS and more uniform HPS in the last phase.

The other question is what is more HPS pure atonement or PoH+HF+penance+cascade. The answer is without question the latter with the caveat that on 10man you need to wait until the debuff stacks high enough. I know this is the case because I have tried every possible approach repeatedly (and I have done it on 10man too both normal and hc). When the logs the maths and my personal experience agree I know I can't be wrong.

I will naturally do what works for me (and I know that pure atonement doesnt) and you will do what works for you and as you say there is enough discussion here to let people make their own minds up, but don't pretend that I am somehow ignoring a reasonable argument. You only have to look a few posts back to see how I react when someone convincingly refutes an argument I have made convincingly.

If this goes live, I would feel even more encouraged to stick to my current playstyle: dps as much als possible!
Especially in 10man I can contribute quite a bit to the overall damage of the raid group (somewhere around 25% of a dps player, or around 5-10% total damage depending on setup and other duties) and dont feel like a burden to the group if I still use SS wisely.
Our Lei Shi HC first kill consisted of mostly atonement, which is amped due to the scary fog, and Spirit Shell before Get Away! phases... this resulted no where near a good ranking on WoL, but that doesn't bother me because we had some close wipes to the enrage timer earlier that evening.

If they buff penance too much, that might even lead to us keeping our T14 4piece bonus in the next content on fights where heavy atonement is favorable. The 4T15 does not sound overwhelmingly good (especially if the travel time of those light thingies is too slow), so for certain fights the cd reduction on penance might be well worth it.