[Holy]Cataclysm Holy Compendium (4.0.6) - Page 12

Malleus · 01/02/11, 7:40 AM

Originally Posted by Shldnhearth

Not sure why you assumed we failed before this.....

Beg pardon if I've misread you, but you implied it when you said you switched on Cho'gall because you had no Resto Shamans that night.

Anyway, you've defeated your own argument here. If DL is good enough when you have no Resto Shamans and Flash is not, then it's better to use DL.

amnerys · 01/02/11, 2:17 PM

For those of you who may be interested in Blessed Life gains, I went 2/2 into it for this week's raids with this spec.

Baradin's Hold:
- Argaloth: Meteor Slash

Bastion of Twilight:
- Halfus Wyrmbreaker: Shadow Nova, Furious Roar
- Veralion & Theralion: None
- Ascendant Council: Water Bomb, Glaciate, Chain Lightning, Quake, Thundershock
- Cho'gall: Depravity, Incidental damage during MCs (e.g. searing totem, certain disorients), Malformation Shadow Bolts

Blackwing Descent:
- Omnotron Defense System: Electrical Discharge, Arcane Annihilator
- Magmaw: None
- Maloriak: Flash Freeze (both application and shatter), Scorching Blast
- Atramedes: Modulation
- Chimaeron: Massacre
- Nefarian: Onyxia Tail Lash, Electrocute, Shadowflame Barrage

Throne of the Four Winds:
- Conclave of Wind: Wind Chill
- Al'Akir: Wind Burst (possibly only the phase 1 version)

And obviously any fight where the adds may aggro onto you and get a hit off will proc Blessed Life as well.

Sodak · 01/02/11, 2:34 PM

Originally Posted by amnerys

For those of you who may be interested in Blessed Life gains, I went 2/2 into it for this week's raids with this spec.

Baradin's Hold:
- Argaloth: Meteor Slash

Bastion of Twilight:
- Halfus Wyrmbreaker: Shadow Nova, Furious Roar
- Veralion & Theralion: None
- Ascendant Council: Water Bomb, Glaciate, Chain Lightning, Quake, Thundershock
- Cho'gall: Depravity, Incidental damage during MCs (e.g. searing totem, certain disorients), Malformation Shadow Bolts

Blackwing Descent:
- Omnotron Defense System: Electrical Discharge, Arcane Annihilator
- Magmaw: None
- Maloriak: Flash Freeze (both application and shatter), Scorching Blast
- Atramedes: Modulation
- Chimaeron: Massacre

Throne of the Four Winds:
- Conclave of Wind: Wind Chill
- Al'Akir: Wind Burst (possibly only the phase 1 version)

And obviously any fight where the adds may aggro onto you and get a hit off will proc Blessed Life as well.

Personally I think 1/2 makes more sense since the ability is capped @ 8 seconds, its not like 100% vs 50 is going to net you double the HP procs, I would guess its going to be almost the same - correct me if im wrong but world of logs has no way to track blessed life gains?

Parodia · 01/02/11, 2:46 PM

WoL does track Blessed Life gains. Go to the individual player, click Buffs Gained or Buffs Cast (both work) and it'll show Blessed Life in the far right column where it shows mana gains.

Sodak · 01/02/11, 3:14 PM

Originally Posted by Sicarri

WoL does track Blessed Life gains. Go to the individual player, click Buffs Gained or Buffs Cast (both work) and it'll show Blessed Life in the far right column where it shows mana gains.

thanks. this is pretty interesting, in my last raid as 2/2 blessed life and 3/3 ToR I had 40 HP from BL and 11 from ToR o.0

Noules · 01/02/11, 4:02 PM

Originally Posted by Shldnhearth

Not sure why you assumed we failed before this..... This was our first night of attempts ever on Cho'gal. Also, our kill was quite clean with no deaths. I had been quite sick and did not raid at all prior to this week aside from Conclave and BH. I'm not sure why it's so controversial to say that it's possible to use FoL over DL in certain setups, I'm not trying to argue with anyone here, I'm just saying it can be done.

It's not 'controversial' exactly, but the math doesn't really seem to work out. The issue is that you are suggesting a FoL/HL split - but this is never optimal HpM for the given HpS if the healing is fungible.

As an example, my tooltips currently show (HpS is neglecting haste, which should affect all spells equally)

HL: 8772 per cast (3509 HpS)
DL: 23400 per cast (9360 HpS)
FoL: 15743 per cast (10495 HpS)

The exact values aren't important because almost everything scales proportionally, these values are merely representative.

Now the question is is HL/FoL combination ever more mana efficient than pure DL? The answer is no assuming you can straight replace DL (this is not always possible due to mechanics or HS CD).

In the above example, pure DL spam is 9360 HpS. In order to match that with FoL/HL spam, you need a FoL fraction of:

x = FoL time fraction

9360 = x * 10495 + (1-x) * 3509 = 3509 + 6987 * x =>
x = 0.84

So roughly you need to be casting FoL 4 times as much (in terms of cast time, not per cast) as HL to match pure DL spam. A rough rotation would be then 35 FoL + 4 HL = 25 DL. From above example values, this is :

35 * 15743 + 4 * 8772 = 586093 (over 62.5s => 9377.4 HpS)

vs DL spam's 9360 HpS. The mana cost of each rotation is:

35 * 6323 + 4 * 2107 = 229733

vs

25 DL = 7026 * 25 = 175650.

It should be clear then that if you need less than DL spam, you are better off doing DL/HL instead of the equivalent fraction of FoL/HL (since the DL portion is already strictly better than FoL/HL mix) and if you need more than DL spam, you are better off doing FoL/DL. Therefore FoL/HL doesn't make sense barring actual cast time issues if you are optimizing. The difference is about 20%, so if you have more mana regen than strictly necessary you can of course make the less efficient strategy work.

Note though that the cast time issues are actually quite common and could constitute and argument for using FoL/HL, but an argument for that sort of mix should be presented in that form (e.g. HS CD interaction). As a general rule healing should ramp up from HL -> DL/HL -> DL -> DL/FoL -> FoL. Currently there appears to be nothing that actually requires high percentage FoL spam (and FoL is only slightly higher HpS than DL in any case) and FoL's role seems to mainly be to address certain encounter mechanics (e.g. Chimaeron).

Zaroua · 01/02/11, 4:46 PM

You guys are making the DL/FoL about a billion times more complicated than it needs to be, give or take one or two.

Is my target going to die in less than 2 seconds if I don't heal them? If yes and you have 2 or 3 Holy Power then Word of Glory. If yes and no Holy Power then Flash of Light.

Is my target in need of critical amounts of healing but isn't in danger of dying in less than 2 seconds? If yes and you have 2 or 3 Holy Power then Word of Glory. If yes and you have 1 Holy Power and Holy Shock off cooldown then Holy Shock + Word of Glory. If yes and no Holy Power then Divine Light.

Is my target in need of healing and in no danger of dying anytime soon? If yes and Holy Shock is off cooldown then Holy Shock. If yes and Holy Shock is on cooldown then Holy Light.

In the end you want to cast as few FoL as possible because the spell is atrocious when compared to Divine Light.

amnerys · 01/02/11, 5:03 PM

Personally I think 1/2 makes more sense since the ability is capped @ 8 seconds, its not like 100% vs 50 is going to net you double the HP procs, I would guess its going to be almost the same - correct me if im wrong but world of logs has no way to track blessed life gains?

Most of the abilities that trigger Blessed Life, such as Massacre, Glaciate, etc. occur on timers more than 8 seconds apart. Even the most frequent ability (Modulation) seems to trigger it every time it hit. I can see this argument if you spec for PvP, where you will be hit quite often, but from a PvE perspective, having a guaranteed HP proc also means you can sort of "plan" around having 2 HP before each ability and then cast Light of Dawn, for example, once it does proc. It worked out especially nicely for me on Ascendant Council, where Glaciate's proc almost always coincided with the tics of Rising Flame. I will admit that it is wasted on the 3-in-a-row Furious Roar stuns.

By the way, if anyone else would like to see what procs Blessed Life from WoL, go to the Log Browser, click paste, and type in the following expression.

[{"spellNames": ["Blessed Life"], "eventTypes": [8]}, {"eventTypes": [1], "targetNames": ["Your Name Here"]}]

And then just search for "Hit" and "Proc" combos next to each other, for example:

[19:45:57.093] Argaloth Meteor Slash Amnerys 52777
[19:45:57.484] Amnerys gains 1 from Amnerys's Blessed Life

Sethronu · 01/03/11, 5:39 AM

On topic of Holy Radiance in general and Haste making it even more amazing in particular, I'd like to point out that HR is leaps and bounds ahead of any other spell we use in terms of overhealing (with the obvious exception of BoL and PotI). It averages about 35-40% overhealing throughout full instance clears, and even on Chimaeron it's the highest overheal ability with 20-25% overhealing on HR there being not uncommon.

Sodak · 01/03/11, 6:26 AM

Originally Posted by Sethronu

On topic of Holy Radiance in general and Haste making it even more amazing in particular, I'd like to point out that HR is leaps and bounds ahead of any other spell we use in terms of overhealing (with the obvious exception of BoL and PotI). It averages about 35-40% overhealing throughout full instance clears, and even on Chimaeron it's the highest overheal ability with 20-25% overhealing on HR there being not uncommon.

That is actually not that important because presumably we're using it when it becomes effective to do so, so for example if 5 people in your raid are full health but the other 20 could use healing its going to be efficient to use HR despite the fact several people will recieve overhealing.

elvez · 01/03/11, 10:42 AM

Hello, it's my first post here. I've been lurking here for a while and I think it's time to give it a shot and try to contribute something, even if it's not 100% correct. I forgive in advance for oversights or imperfections as I'm pretty new to theorycrafting and holy paladins in general. I hope you can give constructive feedback so I can correct my mistakes.
What I've been trying to do is to lay down some napkin math as to get a rough estimate of stat values, with the exception of Spirit (which has already been discussed in detail in the previous pages).

For simplicity, my calculations assume full raid buffs, and no procs or cooldowns of any kind are involved.
My focus is on single target healing as it's the easiest to model.

Intellect
As stated by Nodrak, buffs give a 1.1025 multiplier to intellect in spellpower calculations, which in turn receive a 10% overall bonus. So the additional spellpower gained from X points of Intellect is derived from:

SP = ((INT + X) * 1.1025 + WEAP_SP) * 1.1 = (INT * 1.1025 + WEAP_SP) * 1.1 + X * 1.1025 * 1.1 = BASE_SP + X * 1.2127

So each additional point of intellect gives 1.2127 additional spellpower. INT is the unbuffed intellect of the character and WeaponSP is the spellpower on gear.

From now on, I start to get less confident about my calculations...

We know that [Walk in the Light] gives a 15% bonus to all healing, [Divinity] Gives a 6% bonus to all healing, [Glyph of Seal of Light] gives a 5% bonus to all healing and [Conviction] gives a 9% bonus to all healing (if we assume 100% uptime). Assuming all of these bonuses stack multiplicatively (EDIT: they don't. Even if the values are different, the relative results should stay the same), the overall bonus to healing is:

1.15 * 1.06 * 1.05 * 1.09 = 1.3951

Knowing this, and using the spell coefficients provided by Nodrak (base values are averaged), we get the following values for our single target healing spells:

FoL = (7328  + SP * 0.8632 ) * 1.3951       = 1.3951 * (7328  + BASE_SP) + X * 1.3951 * 1.2127 * 0.8632
HL  = (4400  + SP * 0.432  ) * 1.3951       = 1.3951 * (4400  + BASE_SP) + X * 1.3951 * 1.2127 * 0.432   
DL  = (11733 + SP * 1.15306) * 1.3951       = 1.3951 * (11733 + BASE_SP) + X * 1.3951 * 1.2127 * 1.15306
HS  = (2738  + SP * 0.2689 ) * 1.3951 * 1.3 = 1.3951 * (2738  + BASE_SP) + X * 1.3951 * 1.2127 * 1.3 * 0.2689 (Assuming 3/3 [Crusade])
WoG = (6399  + SP * 0.627  ) * 1.3951 * 1.1 = 1.3951 * (6399  + BASE_SP) + X * 1.3951 * 1.2127 * 1.1 * 0.627  (Assuming [Glyph of Flash of Light])

Following Nodrak's convention, if we set X = 100 Intellect, and if we ignore for now critical heals, the increase on the amounts healed by each spell will be (values are rounded):

+FoL = 146
+HL  = 73
+DL  = 195
+HS  = 59
+WoG = 117

(I didn't account for the crit gained from +100 intellect, as it's not going to make a noticeable difference in such rough estimates.)

Even though I despise doing this, as the gains from crit, haste and mastery are relative to the amount of spellpower, some reference values for the aforementioned healing spells are needed. To do this, I set X = 0, INT = 4000 (a value attainable in mostly heroic dungeon gear) and WEAP_SP = 1729 to get 6752.9 buffed spellpower.
From this, the value of each uncritted healing spell can be easily calculated (values are rounded):

FoL = 18230
HL  = 10208
DL  = 27232
HS  = 8259
WoG = 16318

Intellect, Haste and Mastery are indirectly affected by crit too. Even though my parses show higher crit percentages than expected (I'm hoping it's because of [Divine Favor]), most raiders should sit at roughly 20% buffed crit. This gives a 1.1 multiplier to the gains of each stat, with the exception of crit, of course. And the average value of each heal, taking into account critical heals:

CFoL = 18230 * 1.1 = 20053
CHL  = 10208 * 1.1 = 11229
CDL  = 27232 * 1.1 = 29955
CHS  = 8259  * 1.1 = 9085 (EDIT: This is wrong, it's not accounting for [Infusion of Light])
CWoG = 16318 * 1.1 = 17950

These reference values will be needed soon, but for now we can estimate the overall healing gains from intellect, taking into account 20% critical chance:

+FoL = 161
+HL  = 80
+DL  = 214
+HS  = 65
+WoG = 132

Haste

I follow again Nodrak's convention of using 100 haste rating, corresponding to 0.78% haste.
Haste should stack multiplicatively with the 9% haste given by [Judgements of the Pure], the 3% haste given by [Speed of Light] and the 5% haste given by raid buffs, so the buffed haste value is:

0.0078 * 1.09 * 1.03 * 1.05 = 0.0092

On a given timespan, 1% more haste will (on average) net 1% more healing, and applying this principle to the reference healing values calculated before (CFoL, CHL, CDL...) will get the increased healing for 100 haste rating:

+FoL = 184
+HL  = 103
+DL  = 276
+HS  = 0 (Direct effect of haste on Holy Shock is null, as it's instant)
+WoG = 0 (Same as above)

Even though it may look like haste should give more benefits than intellect, it's important to remember the statistical distribution of the heals during the fight, as the lost throughput from [Holy Shock] and [Word of Glory] are far from negligible as they often account for 30% or more of overall direct heals.

Crit

Again, 100 crit rating is the reference value, which corresponds to 0.5578% crit.
Nothing relevant to my knowledge affect crits, so we can calculate the average increment given by 100 crit rating using the reference uncritted heal values (FoL, HL, DL...). Crits heal for 50% more so it's just a matter of multiplying by 0.005578 * 0.5:

+FoL = 53
+HL  = 28
+DL  = 76
+HS  = 23
+WoG = 45

Crit seems incredibly awful, and is the main reason why I ask for feedback, as in my opinion it's not possible for a stat to have such a low value in comparison to haste or intellect.

EDIT: [Infusion of Light] should increase the value of crit, by how much I still have to find out (thanks to Drekor for pointing that out).
And there's obviously the mana saving component of Crit (when it doesn't result in overhealing), altough we could say also haste can result in mana saving by allowing more frequent usage of efficient but weak heals. Crit also somewhat interact with mastery by allowing larger shield, and that can factored into the calculations by setting a reference mastery value as it's been done for intellect and crit (thanks to thedudeabides for pointing this out).
The value of crit is also slightly increased by [Protector of the Innocent] healing through the beacon.

Mastery

Mastery rating has the same conversion factor of crit rating, so 100 mastery rating will buff [Illuminated Healing] by 0.5578%. In the best possible case, the absorb can be seen as a "deferred" heal, healing a small part of the next damage taken. So the value of mastery can be calculated similarly to the one of crit, but using the reference "critted" heals (CFoL, CHL, CDL...):

+FoL = 112
+HL  = 63
+DL  = 167
+HS  = 51
+WoG = 100

It's important to note that [Illuminated Healing] doesn't proc from heals generated by [Protector of the Innocent], the beacon and Holy Radiance, and given its short duration and the fact that the shield is easily overwritten, the effective value is much slower.

Weighting

To estimate the relative value of each stat, I define a distribution on the healing spells, to account for how much of a loss is for haste not affecting instant heals or mastery only affecting direct heals. This step is entirely arbitrary, as the distribution will vary greatly between different fights, but it wasn't my goal to provide absolute answers.
My chosen distribution is:

FoL 5%  
HL  24% 
DL  32% 
HS  20% 
WoG 19% 
(Direct heals account for 50% of overall healing.)

So the weighted average of the incremental gains we estimated before are:

Int:        +134        
Haste:    +122  
Crit:       +47  
Mastery: +104 (Only direct heals)
Mastery: +52   (Direct heals account for 50% of overall healing)

Which gives a rough stat weighting of:

100 Int = 110 Haste rating = 285 Crit rating = 129 Mastery rating (100% direct heals) = 258 Mastery rating (50% direct heals)

If this is correct (and that's a big IF), it should allow some better theorycrafter to work out the other small details and maybe integrate Spirit to give a better overall picture. It should also be easy to put this stuff into a spreadsheet or software to allow customization for base intellect and spellpower values, crit and healing spell distribution. In any case, the math is very simple and it's easy to derive the values for specific configurations.

Nodrak · 01/03/11, 12:00 PM

Originally Posted by elvez

If this is correct (and that's a big IF), it should allow some better theorycrafter to work out the other small details and maybe integrate Spirit to give a better overall picture. It should also be easy to put this stuff into a spreadsheet or software to allow customization for base intellect and spellpower values, crit and healing spell distribution. In any case, the math is very simple and it's easy to derive the values for specific configurations.

It is more or less correct with the only thing being 'wrong' is that some of the healing modifiers are additive with each other. Where your analysis breaks down, aside from mana as you point out, is that you don't seem to consider that Int/SP is an additive increase to the spell's power, causing a relative diminishing return (relative to other stats). The severity of this is only enough to have influence over comparisons to Spirit and Haste, as both Crit and Mastery are inferior to the 'potential' increases from Spirit or Haste. We will not reach Int levels high enough to cause the other two HPM influencing stats, Crit and Mastery to be desirable.

This however is the easy side of the equation. There are far more factors influencing total available mana, hps requirements and spell selections then those effecting just the basic heals. One has to consider that the heals are [ (FoL/DL/HL/HS) + (PotI) ] x [Beacon], and that PotI has spell power scaling.

Upgrading from HL to DL is not like upgrading from FoL to HL in 3.X was. In those days, this was roughly an even trade in HPM, putting extreme value into mana related stats. These days that is not the case, using a DL is in every single situation a loss of healing per mana. This is not to say that it is not a bad choice, rather it is just not as easy of one. In practical terms though, the obvious choice will be the trio of stats, Int, Haste, Spirit, simply due to basic scaling deficiencies in Mastery and Crit and constraints on itemization and customization.

I also updated the OP a bit since I had some time, still need work though.

nikitabanana · 01/03/11, 12:39 PM

Originally Posted by DiamondTear

This will of course depend on you raid comp, but people might be putting too much weight on holy power generation. We usually have three paladins, two shamans and two priests in the raid. With all that AOE healing to throw around, it might be beneficial to focus more on bringing the people that are really low on HP up. That's why I've taken all the WoG talents and end up usually casting DL and WoG on the people below 50k health. I'm not falling far on the meters from people in the double tide group focusing mainly on AOE healing and shamans and druids certainly can't touch my output.

Beacon makes us very effective at healing single targets and there is often need for it. You don't have to always be maxing out your LoD output anymore. I've never used CS and I don't always have LoD glyph on.

I'm not interested in the EJ heal at all, even though I usually judge on cooldown. The tank either needs heal or doesn't. If he does, you're casting DL on him, if he doesn't, he'd get the needed healing from the next spell you cast on a raid member anyways.

It still surprises me how many people (including our other paladin) spec 2/2 EJ. I still think 1 pt in blessed life is far superior (or another pt in the WoG talents). There are only a couple of bosses where I've found CS to be fairly useful. You can definitely CS on fights like Conclave (trivial now, though), Chimaeron since his hotbox is massive, Nef, Maloriak, and Atramedes if you run through him to dodge something).

For Chim, I've definitely found the LoD glyph to be fairly useless since ground aoe's more then suffice. I'd rather have higher regen from Plea/Divinity due to having to use FoL enough as well as divine protection since you can stay out of the danger ranges fairly easily.

Originally Posted by DiamondTear

Often one DL won't bring the target even close to full health, so topping them with FoL and still having mana after that becomes an impossibility. I won't touch FoL outside Chimaeron.

This is 100% true. The only fight I even remotely cast more then 1 FoL on is Chimaeron depending on how often my group gets hit. The AoE phase is trivial with 2 shaman and 2 priests in raid, so after a holy radiance the HP is better spent using WoG during double attacks on the tank. Obviously, a typical aoe style paladin spec is essentially useless on that particular fight since you only really need to heal around two conditions occurring - double attack being queued or your group taking damage. It turns out it's a pretty good fight to spec offensive abilities (denounce, etc) because there's quite a bit of downtime where you can spam dps into Chimaeron.

elvez · 01/03/11, 1:26 PM

Thanks for the feedback Nodrak.

Originally Posted by Nodrak

It is more or less correct with the only thing being 'wrong' is that some of the healing modifiers are additive with each other. Where your analysis breaks down, aside from mana as you point out, is that you don't seem to consider that Int/SP is an additive increase to the spell's power, causing a relative diminishing return (relative to other stats). The severity of this is only enough to have influence over comparisons to Spirit and Haste, as both Crit and Mastery are inferior to the 'potential' increases from Spirit or Haste. We will not reach Int levels high enough to cause the other two HPM influencing stats, Crit and Mastery to be desirable.

Do you know where I can find which modifiers are additive? I didn't research much but it doesn't seem like such information is easy to find out. (EDIT: Nevermind, I found it. Don't know how it could've escaped me)
As far as the additive nature of Int/SP, it is somewhat implied in the calculations. Using a fixed SP value allowed me to dodge the issue, someway. Though you stated before too that Int may be surpassed by haste at some point thanks to its scaling nature.

This however is the easy side of the equation. There are far more factors influencing total available mana, hps requirements and spell selections then those effecting just the basic heals. One has to consider that the heals are [ (FoL/DL/HL/HS) + (PotI) ] x [Beacon], and that PotI has spell power scaling.

Sorry if I don't understand, but do you mean those additional factor could make Crit or Mastery more desirable? Or is it just an observation about my calculation being oversimplified (I know they are)?
Regarding PotI, if anything, it only devalues crit and mastery even more (or can it crit?).
While the beacon is in the best case a 50% more healing when not healing the beacon-ed target, so I didn't think it could have a significant impact on my calculations.

Upgrading from HL to DL is not like upgrading from FoL to HL in 3.X was. In those days, this was roughly an even trade in HPM, putting extreme value into mana related stats. These days that is not the case, using a DL is in every single situation a loss of healing per mana. This is not to say that it is not a bad choice, rather it is just not as easy of one. In practical terms though, the obvious choice will be the trio of stats, Int, Haste, Spirit, simply due to basic scaling deficiencies in Mastery and Crit and constraints on itemization and customization.

Indeed, but in case of heavy damage or when healing the tank we can't simply use HL instead of DL, do we?

But if you say I didn't make major errors when comparing the value of Crit or Mastery with the other stats, does this mean I shouldn't ever upgrade a heroic dungeon drop with an epic if the epic will make me lose a lot of haste?
It sounds a bit ridiculous to me when half of raid drops and the tier pieces don't have haste, and reforging isn't going to make up for the loss of haste when "upgrading"...

Drekor · 01/03/11, 1:27 PM

Crit seems incredibly awful, and is the main reason why I ask for feedback, as in my opinion it's not possible for a stat to have such a low value in comparison to haste or intellect.

You'd want to account for IoL procs which would increase the value of crit for HS.