Stat Priority come 4.3 for Holy Paladins.

[Shinchib]

I honestly cant tell if this guy posted here and trolled us ALL, or is now just really angry?

Nor can I. I did quite enjoy the thread. I'm fairly new to healing and have found this thread very informative (some stuff was over my head but I still found it fascinating).

It's quite amazing that this thread went to 6 pages. It started off with the OP posting about how he felt that mastery > haste builds (capping haste at 1k). It progressed into a number crunch discussion for the next 4 pages on people that are proving the OP wrong while the OP has done 0 mathematical testing himself, basing all proof as a "I did it on the PTR means I am right". The last page and some has just been a flame war where it seems the OP has become a 12 year old child.

Thanks to the OP for:
1) The entertainment
2) The information that I have learned about Mastery and Haste
3) The information I've learned about 4.3
4) The knowledge that a feeling apparently outweighs number crunching (even if it may not be 100% accurate, as one of the other posters said something along the lines of "Just because they did some math with spreadsheets doesn't mean they are gospel")

A last note for a post that I read somewhere towards the end of the thread:
It was regarding some math that a poster made regarding the omission of crit because it's too RNG in their calculation for haste vs. mastery.
I will have to disagree with that. If you are casting more frequently because you have a lower/faster cast for spells, you significantly increase your chances of getting a proc. The probability stays the same, but if you are casting more frequently, statistically, you are bound to get more procs from crit; therefore increasing throughput even more then your original calculation (even if it is wrong)

[Tobley]

Don't take Nillo too seriously.
He has a propensity to state facts with out any tact.
He knows what he's talking about and knows the math.
That being said he has no clue about personal feelings, etc of other people.

And despite my constant efforts to state that healing is not a numbers game he shuts down myself and anyone else that believes something that isn't supported by math.

[Elix]

So 7 pages and still no answer.

[Deleted]

So 7 pages and still no answer.

If i understand most of the talk about stats in the mist of all the personal attacks and what not i think haste will be better for 4.3 due to the amount of healing you will do with HR and that HR gains the most healing from haste because of the extra ticks and the ticks will not provide a mastery shield.

[Deleted]

I suspect that the difference he is seeing has more to do with the fact that shielding pre-empts direct heals, so stacking mastery results in more shields "sniping" HoTs and direct heals from other healers, reducing his overheal while increasing everyone elses. It's the same reason why, in 4.2, mastery stacking paladins can often put up very impressive parses, especially on extremely high overheal fights like H Majordomo even though all of the theorycrafting suggests that mastery stacking is a theoretical HPS loss. All that it does is increases your output by raising other people's overhealing. There's a pretty good argument to be made that while straight mastery stacking may look good, it's questionable how much of a benefit it has to the raid.

I wouldn't call shielding sniping. When you put a shield on a target, other heals can still hit without overhealing. In fact, shields works extremely well with HoTs as they add effective hp margins allowing the HoTs to tick more safely. Yes, on extreme overheal fights mastery "steals" hps, but that is because shielding is plain better on such fights. Whereas most other healing goes to waste, the shields are always effective as they are preemptive healing.

And yes, the power of shielding is one reason why mastery always tends to get underestimated. While other stats such as crit and haste needs to deduct something like 20% from their raw efficiency due to overhealing (not total throughput mind you. Just the part of throughput that the actual stat brings) , mastery pretty much always goes strong. The only case where mastery doesn't run at full efficiency is when a shield expires. But from a healing purpose, if a shield lasts for 15 secs it has pretty much filled its purpose. (if you are AoE healing you will rarely see it expire though)

The second big factor is mana efficiency. Running AoE with haste (below softcap) instead of mastery is pretty similar to healing with FoL instead of DL in terms of efficiency. Haste vs mastery HR lose you about 1.65% efficiency for every 1% of throughput (FoLvsDL is slightly higher at 1.75). That means that you can run a far more aggressive healing pattern with mastery. The last factor is that haste has a ramp-up time in that you need to get in extra casts with it, whereas other stats benefit more from frontloading with IoL, holy power (LoD/WoG casts) and pre-casting.

That said, I question the 20% claim just like I question those who are saying that haste burst is so much stronger (2k haste vs mastery only differs with 5.7% in raw total burst AoE throughtput). Really, I think you can probably go both ways and heal well. I do think that mastery is the more basic choice in 4.3 as it overall provides a better balance. But that is just my opinion. The really interesting factor for me is the 3.5k haste build and how well it will work trading spirit for AoE throughput. Can that build actually be sustainable in top end gear?

[darkwarrior42]

Don't take Nillo too seriously.
He has a propensity to state facts with out any tact.
He knows what he's talking about and knows the math.
That being said he has no clue about personal feelings, etc of other people.

And despite my constant efforts to state that healing is not a numbers game he shuts down myself and anyone else that believes something that isn't supported by math.

With all due respect, while healing does have a certain amount of intuition involved, to contest that it is not a numbers game is a flat-out falsehood. There is a huge amount of math involved; just look at calculations done regarding restoration druid and haste breakpoints.

Here's what we know are facts; I'm doing numbers again because I did them before assuming no buffs or talents, to look at a worst case scenario for haste. I won't make that mistake again. I shall be far more thorough. Anyway, the facts:
1) ~128.05716 haste rating = 1% haste. If you take into account 5% raid haste + 9% haste from JotP, plus 3% haste from Speed of Light, 128.05716 haste rating = 1.1788% haste, or ~108.63 haste rating = 1% haste.
---This second value reflects the fact that haste scales multiplicatively; the more sources of haste you have, the more powerful all of it becomes.

EDIT: As Nillo points out later, the amount of rating to get an amount of % does not change based on buffs. What I was attempting to say here was that with all the haste buffs active, you will see a 1% increase in the haste displayed on your character sheet with 108.63 haste rating, instead of the 128.06 you'll see listed as the standard conversion. I mention this only because so many people are confused by how haste works, and I was attempting to head off confusion about why I said 128 haste rating grants 1% haste, but in-game you see your actual haste on your character sheet go up by more than 1% with that much rating. As Nillo pointed out, the standard haste buffs also affect the power of mastery, by allowing you to cast more spells (and thus generate more shields), but that power is not directly reflected on your character sheet, so I did not mention it, not because it isn't important, but because I did not believe there would be any confusion on the issue.

2) ~179.28 mastery rating = 1 mastery = 1.5% absorption.
---So, 179.28 mastery rating = 1.5% absorb, 108.63 haste rating = 1% haste. That means 162.945 haste rating = 1.5% haste.
-------For the record, 200 mastery rating = ~1.6734% absorb, 200 haste rating = ~1.8411% haste. This is simply to get percentages based on equal ratings, for future use.

So conclusion #1) In an absolute sense, haste scales faster than mastery. If you dispute this, feel free to demonstrate where my numbers are wrong. Keep in mind I have said nothing about playstyle, specific circumstances, or application, thus far I have simply discussed raw numbers. Any counter to this point that involves playstyle is completely false, because it doesn't prove that haste rating does not scale faster than mastery. It might prove that mastery is better for specific situations because of how it functions, but that is a separate argument altogether.

So having come to the conclusion that haste scales better than mastery, let us return to the original premise: that mastery is, in fact, better than haste. We have demonstrated this cannot be the case because of the inherent values of the stats. Let us look at the next option: that the way in which mastery interacts with the spells a holy paladin casts is superior to the way in which haste interacts with those spells. Keep in mind that any spell that benefits fully from both mastery and haste will, numerically speaking, benefit more from [edit: haste], because it takes less haste rating to get a 1% boost to your haste (and thus your hps) than it takes mastery rating to get a 1% boost to your absorb (and thus to your effective hps).

-Holy Light/Flash of Light/Divine Light: The Standard three. All three scale linearly with both mastery and haste. A 1% boost to haste increases average hps by 1%, and a 1% boost to absorb shields boosts the effective healing of the spells by 1%, assuming the shield is fully consumed.
-Word of Glory: As with the standard three, benefits fully from haste and mastery, until GCD capped (which, assuming the same haste buffs as above, would take around 3489 haste rating).
-Light of Dawn: Benefits fully from Haste + Mastery, until GCD capped.
-Holy Radiance (Initial heal): Benefits fully from haste + mastery.
-Holy Radiance (HOT): Benefits fully from haste, no benefit from mastery.
-Beacon of Light: Benefits fully from haste, no benefit from mastery.
-Protector of the Innocent + Enlightened Judgments: Benefit fully from haste, no benefit from mastery.

So we see that there is no spell a holy paladin uses on a regular basis to heal that does not benefit fully from haste; indeed, holy radiance double dips and benefits both on the initial cast and on the hot. However, there are numerous spells that do not benefit from mastery at all, and holy radiance only benefits from mastery on the initial cast.

For clarity, here is the list of spells that benefit from Mastery: Holy Light, Flash of Light, Divine Light, Word of Glory, Light of Dawn, Holy Radiance (Initial portion only). Here we see that if you are focusing only on a single target, and not using the beacon to cross-heal, you would in fact benefit from mastery on almost everything you do (you still wouldn't benefit from mastery on PotI heals on yourself, or on the subsequent transfer through the beacon).
---I would argue, though, that healing in this way is by definition inefficient. Using the beacon effectively tremendously increases your output without requiring extra mana.


So, we now know that the nature of holy paladin spells does not lend itself to receiving greater benefit from mastery than from haste; all spells cast by a holy paladin benefit fully from haste (and in the cast of holy radiance, double dips), while the same cannot be said for mastery. What, then, are other possible arguments in favor of Mastery? There are three that I can see.

1) Absorb shields can increase a tanks effective health above maximum, allowing the tank to take more damage before dying than normal.
1a) While this is true, I do not find it to be a compelling argument. In the case of a heal that does not restore the target to full health, the absorb shield generated by the heal is no better than extra healing; indeed, if the shield is not used, it is in fact worse, as the target is still not at full health. However, assuming the shield is fully consumed, there is no difference between a heal for 1000 that generates a 200 point damage shield, and a heal that simply heals for 1200. To put it in terms of haste, there is no difference between 5 heals that heal for 1200 each (combined value of heal + shield) and 4 heals that heal for 1500 each (combined value of heal + shield). In the case of overhealing, the shield provides value where the overhealing does not; this is mitigated, however, by the loss of mana in overhealing in the first place. If you can exactly heal the target to full health, the mastery shields are clearly superior... however, this is so rare an occurance it is hardly worth mentioning.

Peronsal conclusion regarding point 1) There may be isolated situations where the shield from Illuminated Healing is very handy. However, it is expensive and difficult to work with such situations, and mastery is only truly superior in such situations if the targets are healed to full in the process, AND take another hit in the next 15 seconds... and the haste build still has damage shields, so we're not talking about the difference between having them and not having them. Haste will allow faster actions after the next damage, and achieve the same end result in the hands of a skilled paladin... but be more versatile elsewhere as well. It is simply too mana inefficient to spam heals on a target that does not need healing for the sake of generating a damage shield.

2) Haste increases mana consumption, mastery does not.
2a) This point cannot be denied. Casting spells faster means using mana faster. This only becomes a problem, however, if the increased mana usage causes the paladin to go OOM, or if it causes the paladin to drop down to weaker heals when stronger heals are needed. This is something that cannot be gauged pure mathematically, as it depends on far too many variables that change per group and per fight. In the end, however, skilled paladins are able to function at very high levels of haste without going oom; I have personally seen many paladins using expensive spells frequently, and having no issues lasting out fight after fight.

--Personal conclusion: A potential issue, but one that is readily overcome by skill.

3) The OP simply likes mastery better than haste.
3a) This cannot be debated; either the OP likes mastery or he doesn't. Liking it better doesn't make it better in an absolute sense, but if he enjoys that style more than that's all that really matters... unless he's trying to say his way is better simply because he enjoys it more.



I have not touched on the less mathy aspects of haste vs. mastery, which in my opinion clearly favor haste. The ability to get off a critical spell more quickly (which can be the difference between saving the tank and running back from the graveyard), the increased generation of holy power (through tower of radiance) allowing for increased usage of WoG/LoD, etc. This is more personal of course, but I favor haste for the same reason every paladin spec I put together either specs into pursuit of justice or enchants lavawalker; increased ability to react is power that cannot easily be measured, but can easily turn the tide of battle.



There is one case in which I would argue in favor of mastery... or rather, in which I would argue for a balance of the two. That is the case where you are focusing solely on spells that fully benefit from both haste and mastery. Let us examine such a case, assuming 5% haste aura, 3% speed of light, 9% JotP, 8 baseline mastery, and 4000 points of secondary rating that we can split between haste and mastery however we like. For simplicity's sake, I am going to ignore beacon of light and protector of the innocent, and assume we are spam casting a spell with a 2.50 second cast time that heals for 10,000 hp.
Baseline haste: 17.88%
baseline mastery: 8 (12.00% absorb)
Baseline cast time: ~2.12 sec
Baseline power: 10,000 + 1200 absorb, 11,200 total
Total effective hps: 5,283.02

Case 1: Pure haste. 4000 haste rating = 31.2361% haste. Total haste: ~54.71%. Total Absorb: 12.00%.
Case 1 cast time: ~1.62 sec
Case 1 power: 10,000 + 1200 absorb, 11,200 total
Case 1 effective hps: 6913.58 hps

Case 2: Pure mastery. 4000 mastery rating = 22.3115 mastery = 33.47% absorb. Total Haste: 17.88%. Total absorb: 45.47%.
Case 2 cast time: 2.12 sec
Case 2 power: 10,000 + 4547 absorb, 14,547 total
Case 2 effective hps: 6,861.79

Case 3: Even split. 2000 haste rating = 15.62% haste. 2000 mastery rating = 11.1557 mastery = 16.73% absorb. Total Haste: 36.30%. Total absorb: 28.73%.
Case 3 cast time: 1.83 sec
**Case 3 power; 10,000 + 2,873 absorb, 12,873 total.
**Case 3 effective hps: 7,034 hps

**Fixed the error that earsofpower was kind enough to point out

As you can see, an even split of haste and mastery yielded far more effective hps than stacking one to the exclusion of the other. I'm not saying an even split is the highest hps (I suspect that slightly favoring haste would yield higher hps, but haven't done the math), but clearly stacking only one is inferior. That assumes you benefit fully from both stats, however, which is not the case in most situations.. hence why paladins tend to favor haste. However, if you find yourself in a situation where you are focusing mostly on spells that do benefit from mastery, it's not a bad idea to have more mastery... which is why people in this thread have said that having mastery isn't a bad thing if you are exclusively tank healing.

Edit: For clarity, these last cases only apply to situations where you are using spells that benefit fully from mastery. That means you are not using Holy Radiance, and you are either not using beacon of light at all or are only casting direct spells *at* the beacon, and not transferring any heals through the beacon. Anytime you use the beacon to transfer heals or use holy radiance diminishes the value of mastery.

TL;DR: Normally I don't even pretend to have these sections. This time I add one only because my post is longer than usual (and coming from me, that says a lot). The summary is: in most cases, haste > mastery. When focusing strictly on healing a single tank, haste still > mastery, but you want a much closer balance of the two. If you want reasons or explanations, read my post.

[pepizaraus]

I don't really care to be honest. Tell them that in a private message, not me in my thread.

I believe moderator recruitment is currently closed, although you may wish to apply in the future as you seem so very concerned about it. Stop saying this is YOUR thread, it's not, this is a forum, one with fairly open access. Discussions arise that may not take the course you would like them to, but that is how a forum works, as they are not merely for your benefit.

[Deleted]

It's not important, but you've calculated "Case 3 power" (and so effective hps) using your value for total haste instead of your value for total absorbs

[Nillo]

Don't take Nillo too seriously.
He has a propensity to state facts with out any tact.
He knows what he's talking about and knows the math.
That being said he has no clue about personal feelings, etc of other people.

That's me! :P

[deviancee]

2) Haste increases mana consumption, mastery does not.
2a) This point cannot be denied. Casting spells faster means using mana faster. This only becomes a problem, however, if the increased mana usage causes the paladin to go OOM, or if it causes the paladin to drop down to weaker heals when stronger heals are needed. This is something that cannot be gauged pure mathematically, as it depends on far too many variables that change per group and per fight. In the end, however, skilled paladins are able to function at very high levels of haste without going oom; I have personally seen many paladins using expensive spells frequently, and having no issues lasting out fight after fight.

--Personal conclusion: A potential issue, but one that is readily overcome by skill.

Not entirely true. Assuming you use right spells at right time, since its not a big deal, keep judgement buff up (4.3) and Divine Plea whenever it is possible (no damage taken on raid), you are entirely limited by your gear (spirit, and yeah, haste). Curently on nerfed farm content its not a big deal, but when you will be progressing in heroics with mostly 391s you are going to feel it.

[Nillo]

1) ~128.05716 haste rating = 1% haste. If you take into account 5% raid haste + 9% haste from JotP, plus 3% haste from Speed of Light, 128.05716 haste rating = 1.1788% haste, or ~108.63 haste rating = 1% haste.

I'm sorry I have to stop you right here, but this is wrong.
No matter which buffs you have you always need 128 haste rating for 1% of haste. Those buffs merely reduce the BASE casttime of your spells, but do NOT reduce the haste required to get 1% of haste.
Those buffs also affect all other ratings as much as they affect haste (because they already increase the number of spells you cast no matter what haste level you are at).


Mastery gives slightly more healing per point, but the reasons haste is better is that mastery is usually far from 100% effective due to it's short duration and limitation to direct heals (for me it's ~60-70% except for baleroc where it is higher than 100%) while having shorter casttime allows you to react better to people dropping low.

[Katfish]

I seen many posts about gaining the extra tick of HR with haste. But since our HoTs don't stack you would have to cast it once and let all ticks count in order for it to really count. In a aoe spam issue you would be overriding your HoT anyway so that extra haste is wasted. For that issue alone though. On the ptr I'm finding that I'm casting HR back to back a lot and then direct heals some then back to HR spam, etc etc. Haven't seen a time when I just cast HR once.

I can see both sides on Mastery/Haste argument. Really looking forward to some real math being done on this as we move closer to 4.3 release.

[kiristus]

Surprisingly enough, your attempt at trolling has actually spawned decent theory-crafting, despite your best (worst?) efforts. - Malthanis

[Barthalas]

Came here to see the stat priority in 4.3, the things he posted were interesting but sadly he was unwilling or unable to back up his claims with mathematics. Even still, now paladins have a nice little home to start baking up some theory crafting.

I wonder what Malthanis removed, regardless... Malth is right, lol.

*EDTI*: Can anyone say counter productive on Kiristus' end?


Making alternate accounts to circumvent bans is quite against the rules. - Malthanis

[Deleted]

Those buffs also affect all other ratings as much as they affect haste (because they already increase the number of spells you cast no matter what haste level you are at).

This pretty much. You can't reduce rating needed when then stat has multiplicative buffs, as those buffs effect all stats equally. What you can do however is add rating needed for stats that have non-multiplicative buffs. Mastery for example don't need 179.28 rating for its first actual 1.5% throughput increase. In reality it needs 179.28*1.08 rating. This same thing also comes into play when stacking a single stat. The more you stack a single stat, the more all other stats benefit. This is why you are always encouraged to use simcraft when balancing secondary stats as a dps. (and why it sucks to not have any simcraft healing models)

Anyway to answer the person you responded to also. It always comes back to burst vs efficiency. If efficiency didn't matter we would all be spamming FoL and HR all day long. And if burst didn't matter we would cast nothing but HL/HS/WoG/LoD.

I seen many posts about gaining the extra tick of HR with haste. But since our HoTs don't stack you would have to cast it once and let all ticks count in order for it to really count

Other threads I have seen indicate that you can only place one HR buff on a single target, but have multiple HoT buffs up by using different targets (however annoying it may be). It would be nice to get it verified though.

[rinleezwins]

All that has really changed is 60% more benefit from spirit.
Also, around 1600 haste from items + judgement haste buff and divine favor/blood lust = 5 HR ticks.
You will always reach 4 ticks raidbuffed, 5 ticks with DF/BL only so it's pointless stacking more haste than ever.
Yes, HR does not stack but why would you want to cast it on just one person?

[darkrealm]

So I read through all 7 pages (as much as I could with the random trolling) and I wanted to pose a question. If, we know in general that haste scales better than mastery (this is an assumption), then why not roll Haste >mastery> spirit> to a more balanced number system to make a more balanced healer? From what I have read there is a bunch of AOE damage incoming, but it still would not preclude us from direct heals. I have to say, I have some mixed feeling about casting one spell as an end-all to our healing. Sure it would be easy, but it sounds rather boring. Currently I enjoy the challenge of holding on procs for big heals and heavy damage for the tanks. DO not get me wrong, I like the sound of having more raid healing ability, but I will miss the "tank healer" tag I have earned over the years.

One last thought, currently I noticed something with the abosrb created by mastery, it seems to stack to "X" on single target. Believe it or not I had never noticed it stacking until recently. Does anyone know if this mechanic still works the same on the PTR?

[Lylethantos]

So I read through all 7 pages (as much as I could with the random trolling) and I wanted to pose a question. If, we know in general that haste scales better than mastery (this is an assumption), then why not roll Haste >mastery> spirit> to a more balanced number system to make a more balanced healer? From what I have read there is a bunch of AOE damage incoming, but it still would not preclude us from direct heals. I have to say, I have some mixed feeling about casting one spell as an end-all to our healing. Sure it would be easy, but it sounds rather boring. Currently I enjoy the challenge of holding on procs for big heals and heavy damage for the tanks. DO not get me wrong, I like the sound of having more raid healing ability, but I will miss the "tank healer" tag I have earned over the years.

One last thought, currently I noticed something with the abosrb created by mastery, it seems to stack to "X" on single target. Believe it or not I had never noticed it stacking until recently. Does anyone know if this mechanic still works the same on the PTR?

Generally you'll want int>spirit>haste>mastery as your stats simply because spirit will give you longevity that you'll need to cast expansive spells, while haste will allow you to cast more spells in a given time and then mastery will allow those spells to be slightly stronger with an shield that's left behind on the target. The biggest blow to mastery is that it doesn't proc off of beacon heals. If this happened it could very easily change the stat weighting to look like int>spirit>haste=mastery>crit and you would want to probably keep them semi balanced to allow for better synergy between the two.

[Tiberria]

So I read through all 7 pages (as much as I could with the random trolling) and I wanted to pose a question. If, we know in general that haste scales better than mastery (this is an assumption), then why not roll Haste >mastery> spirit> to a more balanced number system to make a more balanced healer? From what I have read there is a bunch of AOE damage incoming, but it still would not preclude us from direct heals. I have to say, I have some mixed feeling about casting one spell as an end-all to our healing. Sure it would be easy, but it sounds rather boring. Currently I enjoy the challenge of holding on procs for big heals and heavy damage for the tanks. DO not get me wrong, I like the sound of having more raid healing ability, but I will miss the "tank healer" tag I have earned over the years.

One last thought, currently I noticed something with the abosrb created by mastery, it seems to stack to "X" on single target. Believe it or not I had never noticed it stacking until recently. Does anyone know if this mechanic still works the same on the PTR?

Because, until you are at the point where your regen has no impact on your spell selection (i.e you can spam Divine Light non stop or Holy Radiance post 4.3), more regen will always equal more output at a greater proportion than throughput stats (haste/mastery/crit) can provide. This is going to be especially true in 4.3, with the value of Spirit being increased essentially by 60%. Spirit will be the clear #1 stat.

[darkwarrior42]

I'm sorry I have to stop you right here, but this is wrong.
No matter which buffs you have you always need 128 haste rating for 1% of haste. Those buffs merely reduce the BASE casttime of your spells, but do NOT reduce the haste required to get 1% of haste.
Those buffs also affect all other ratings as much as they affect haste (because they already increase the number of spells you cast no matter what haste level you are at).

There were a few errors in my post I didn't catch, I'll go back and fix those in a sec.... but the point of my post was to indicate the way that haste functions for those not already familiar with it. I"m well aware of the fact that those buffs increase the power of other ratings by virtue of allowing more spells to be cast (as is seen in my case #2 calculations), but one of the most common questions I see regarding haste is "I just gained 128 haste rating, why did my haste go up by more than 1%?" or something along those lines.

I could have been more clear; what I meant to say was that with all relevant buffs, that is amount of haste rating you will need to see a 1% increase on the haste displayed on your character sheet. The haste buffs don't directly change any of the other numbers on your character sheet in a similar fashion, so I did not mention it... but you are right, those buffs will increase the power of mastery as well, by allowing more opportunities for the mastery to come into play.

In attempting to make my post idiot-proof (no offense meant to anyone here), I seem to have brought in additional errors.... I feel like Microsoft now

Going back to clean up numbers in my last post now.