Cataclysm PvE Holy Paladin Theorycrafting

Prev 1 2 3 26 Next
The spec I recommended has 1/2 Enlightened Judgements giving 15 yards on Judgement.

There's still 2 floater points. I took 1 of those into 3/3 PotI for the extra healing (esp. beacon heals), and 1/2 Improved Judgements for 25 yard range. It makes it easier to hit ranged with LoD when they cluster.

Again, those are floater points you can put where ever you like.
From my latest heroic run, mastery consisted of 6.8% of my total healing. If it was used to full effect, it should have consisted of 12.5% of my total healing. This is not sufficient analysis, since there are all kinds of short cut's I'm taking, but it gives you a rough idea of how well it's showing in Cataclysm.
Thank you so much for the info, tried it out and it works thank you again.
Me and a guildie, both Hpals are having an arguement that maybe you can resolve. He says that the hpal mastery is a better stat then spirit and I believe the exact opposite. Any numbers you have on this to resolve it? Any help would be much appreciated.
Your friend is wrong. Until you can heal without going OOM, you need more spirit. Mastery by my initial estimates is working at about 1/2 effectiveness (half your shields are wasted). This is an initial estimate, so don't quote me on that.

100 Spirit will give you about 40-50 in combat mp5. 100 Mastery will give you 0.7% extra shielding, but at only 1/2 effectiveness, that's about 0.35% extra shielding that's used.

Though now that I look at these numbers, at later tiers of gear, mastery might beat crit, but don't quote me on that yet.

EDIT: Not paying attention to spell-check fail.
Ty for the quick and accurate reply :D Yea I knew that half the shields were being wasted. If you get any recounts on attempts for actual numbers please update :D I am always looking for new information about my class. Thank you very much. I hate that alot of Hpals think that our mastery is some all powerful stat when at the moment it isn't.
Great post. Found it completely informative and helpful. Still trying to get used to the new playstyle of Holy, I healed my first dungeon last night (Stonecore) and we didn't wipe at all, guess I am doing something right after all? All thanks to finding this thread and reading your guide. BTW I haven't played holy since pre-WotLK so this is almost like playing a completely new class. Anyway just wanted to say thank you for helping an old school pally become somewhat competent in this ever changing World of Warcraft.

Sidebar: My main spec is Ret but I do roll Holy mainly so that I can get faster dungeon queues :D
I made a mistake on my analysis of major Glyphs.

We're using LoH only once per fight, or twice at most. So considering the mp5 of Divinity is silly. It's better to consider the exact amount of mana.

So the way to do this, is consider how many cleanses it takes to match the mana returns on Divinity, and the number of beacon swaps it takes to match the mana returns. For me, this is 11 cleanses, and 5 beacon swaps.

I haven't tested the 20% returns, so you can keep it in mind if this is the case. However, I officially can't condone creative use of game mechanics :)

Thinking more into it, glyphing Lay on Hands may be considered a waste of a glyph. I haven't seen any raids yet, though I find it hard to believe they're going to last that long. If LoH is still only going to be used 1/fight then the glyph only lets us use LoH during trash pulls, or if your group is fast on wipe resets. I find it difficult to argue for glyphing for trash if this is the case, but again, I need to get into some raids to say for sure.
I haven't found myself needing to use LoH more than once in an encounter unless things are going really, really bad and with the current gear you are probably going to wipe if you lose more than 3-4 people (>20%) because of either a berserk or healers going oom trying to compensate for the loss of other healers. I have swapped between the Cleanse and Beacon glyphs depending on the encounter so you may not really ever have a "definitive" answer.

PS. Nice guide!
EDIT2: Confirmed that CS can proc SoI. I haven't however, considered Eternal Glory. If I accounted for EG, then CS for HP is only worth it if you plan on "replacing" FoL healing, which frankly seems silly. The mistake I DID make was an inaccurate K factor. However, that "bad" K is consistent across all spells and the conclusion does not change.

I made another mistake and let CS's GCD be affected by haste. This makes CS a worse option.
Leave it to Adt to figure out how to give me more work. :)


First, we need the proc rate of Seal of Insight. There are a lot of mechanics that relate to DPS theorycrafting that I'm not as familiar with. Correct me if I make any mistakes. As I understand it, ppm converts to weapon proc rate by the formula:

Proc Rate = PPM/(60/Weapon Speed)

My caster weapon has a swing speed of 1.6. I assume most, if not all caster weapons have the same speed. As best as I can tell, SoI has 15 PPM.

Consequently SoI has a 40% chance to proc.

I assume that Crusader strike can proc SoI, and not just white attacks. Likewise, I assume that the weapon swing timer continues during the GCD Crusader Strike consumes. If Crusader strike can not proc SoI, the relative mana cost of CS goes up.

Our chance to hit is:
1 - Chance to Miss - Chance to Get Parried - Chance to Get Dodged

Chance to Miss can be dropped to 0 with sufficient spirit. If you didn't take enlightened judgements it's 0.08.

Chance to get Parried can be dropped to 0 if you strike from behind. Otherwise it's 0.14.

Chance to get dodged is 0.065. We have no expertise and can't remove this.

So there are 2 "swings" that we need to consider
1) Crusader Strike itself.
2) The partial auto-attack.

Crusader Strike
Assume we strike from behind, and we have enough spirit to hit cap. Then the chance to hit with CS is 93.5%. SoI has a 40% chance to proc. SoI will return 4% base mana, while CS costs 10% base mana.

Auto Attack
We will get in partial attacks. We'll pretend this goes on long enough that we can pretend say 0.8 "attacks" is possible. In reality, some GCD's will have attacks, and some won't.

# of attacks = GCD/Weapon Swing Speed per CS attack.

The GCD = GCD0 (base GCD) / (1+total Haste)

The Weapon Swing Speed = Swing Speed0/(1+total Haste)

Haste divides out. If caster weapons have 1.6 second swing timers, we will get in 0.9375 attacks per GCD.

The auto attack will have the same proc rate, and the same miss chance as regular crusader strikes.


So the mana returns from SoI per Crusader strike are:

0.4 * 0.935 * 4% Base Mana (Crusader Strike) + 0.9375 * 0.4 * 0.935 * 4% Base Mana
= 0.02899 Base Mana =679 mana returned.

Crusader strike itself will Cost 2342 mana.

So on average, Crusader Strike will cost 1663 Mana to use.

So the question is, is 1 HP worth 1663 mana?

Give me some time and I'll let you know.
*Sigh, this post is now outdated... AGAIN
12/14/2010 1:37 PMPosted by Nethaera
These hotfixes were intentional and we wanted to share more information about the "why" for you all here:

Mana is supposed to matter for healers, and in the case of Holy paladins, that just wasn’t happening. We were seeing raid groups attempting to learn new bosses where the other healers were out of mana while the paladins were still at 90% of mana. We also thought the experience of paladins healing harder heroic dungeons was inconsistent with that of other healers, or our design intent. We thought it was only a matter of time before groups started going out of their way to stack paladins for raids or recruit them for dungeon runs, which of course is not our intent. For this reason, we thought it was important to address this point now rather than sitting on it until a future patch.

The main culprit for the increased paladin efficiency this was Holy Power from Tower of Radiance. We were seeing a strategy develop where paladins would cast Holy Light on a Beaconed tank, and then cast Light of Dawn on the raid (and use very few other spells). This strategy was remarkably successful considering how simple it is. The Holy Power-based heals are supposed to be an important component of the paladin kit, but because Holy Light is designed to be super efficient, the overall strategy was too efficient. The intent of Tower of Radiance was to make directly healing a Beacon of Light target less punitive, but instead it was becoming the only smart way to play. You can still generate Holy Power by using Flash of Light or Divine Light on a Beaconed target, or through a variety of other ways. Remember, overall it isn’t balanced if paladins are just as mana efficient as other healers and have several mana-free spells.

Light of Dawn was adjusted because it was completely eclipsing Word of Glory. We lowered the healing of Word of Glory late in development, but also redesigned Light of Dawn. They had fallen out of sync, and Light of Dawn was just too good.

TL/DR: Holy Light doesn't generate holy power with Tower of Radiance and Light of Dawn got nered.
So the question is, is 1 HP worth 1663 mana?

and a GCD
EDIT: I did NOT make a mistake. Confirmed that CS can proc SoI. I haven't however, considered Eternal Glory. If I accounted for EG, then CS for HP is only worth it if you plan on "replacing" FoL healing, which frankly seems silly. The mistake I DID make was an inaccurate K factor. However, that "bad" K is consistent across all spells and the conclusion does not change.


Note: All these numbers assume no crits, and ignore mastery. If you'd like to include these, the easiest way to do so is to increase K appropriately. Otherwise, have fun doing a Monte Carlo sim.

I don't trust the LoD #'s anymore, but the WoG numbers should be accurate.

So everything here, is going to use WoG to evaluate 1 HP.

In general, the average, non-crit heal for spells are:

HL = K*(4400+0.428*SP)
DL = K*(11733+1.148*SP)
FoL= K*(7328+0.865*SP)
WoG/HP = K' * (2133+0.198*SP)

Glyphed, I assume K' = K+0.10 for Word of Glory.
In general, because Conviction is no longer steady, K=1.05 (SoI Glyph) *1.15 (Walk in Light) * 1.06 (Divinity) = 1.27995 and ignoring conviction for now.

Mana Costs are:
HL - 2107
DL - 7026
FoL - 7328

General HPM formulas are:
HL= K(2.0882 + 0.000203132*SP)
DL= K(1.66994 + 0.00016339*SP)

So we want to divide the heal of WoG by the HPM of these spells to determine how much mana it would take with the matching spell to fill in the heal WoG would have put out.

Mana Value of 1 HP = K' * (2133+0.198*SP) / HPM Formula
At 5000 SP, K=1.27995, K'=1.37995
Mana of 1 HP relative to:
HL - 1084.75
DL - 1377.082
FoL - 1826.955

As SP goes up, the relative mana value of 1 HP goes down, to the extremes of (Calculating this involves dividing by A+B*SP, where A and B are constant. Yes it's possible):
HL - 1050.888
DL - 1306.477
FoL - 1560.424
It costs 1663 mana to crusader strike. (This accounts for mana returns from SoI.) The math to derive this was shown here:
In conclusion:
Using Crusader strike to generate HP will make you LOSE mana. The only time Crusader Strike is worth the mana, is if you do nothing but use FoL and have crappy gear.

It might save you mana if you hit enough targets with Light of Dawn. I can't tell you how many targets you need to hit until I know the new base and SP coefficient of LoD.

Is Crusader Strike worth using for HPS bonuses?

Each Crusader Strike will either cost you (pretends you only used CS to build HP):
2 GCD if you dump w/ 1 HP,
1 2/3 GCD if you dump w/ 2 HP or
1 1/3 GCD if you dump w/ 3 HP

For example, the 2 GCD cost comes from 1 GCD to GS, and 1 GCD to WoG.

The 1 2/3 GCD's comes from 2 GCD's for 2 CS, and 1 GCD for WoG. The other HP might not necessarily come from CS, but the CS itself still is going to take "more" time to use than 1 GCD. So I'm breaking it into thirds.

Each Strike will generate 0.935 HP on average due to miss chance.

Let H=total Haste
Normally H = 1.03 (Speed of Light) * 1.05 (Raid Buff) * 1.09 (JotP) * (1+hr/12780 )

hr = haste rating
GCD Time = 1.5/H

From above, 1 HP = K' * (2133+0.198*SP) Healing.

1 CS provides 0.935 HP, so 1 CS provides: 0.935 * K' * (2133+0.198*SP) Healing.

The cast times vary as shown above, depending on the amount of HP you have at the time of dump.

General Solutions
if dumping @ 3HP, HPS = 0.935*K'*(2133+0.198*SP) / (1.33333 * 1.5 / H)
if dumping @ 2HP, HPS = 0.935*K'*(2133+0.198*SP) / (1.36667 * 1.5 / H)
if dumping @ 1HP, HPS = 0.935*K'*(2133+0.198*SP) / (2.0 * 1.5 / H)

For comparitive purposes, I'm going to approximate the glyph as stacking multiplicatively, so K'=K*1.10 instead of K'=K+0.1. This makes my math a little bit easier. This will trend the HPS of CS higher. I'm also joining terms

@ 3HP, HPS = (0.935*1.1*K*2133*H) / (1.3333 * 1.5) + (0.935*1.1*K*0.198*H) * SP / (1.3333 * 1.5)
@ 2HP, HPS = (0.935*1.1*K*2133*H) / (1.6667 * 1.5) + (0.935*1.1*K*0.198*H) * SP / (1.6667* 1.5)
@ 1HP, HPS = (0.935*1.1*K*2133*H) / (2.0 * 1.5) + (0.935*1.1*K*0.198*H) * SP / (2.0 * 1.5)

Cleaning up:
@ 3HP, HPS = 1096.89*K*H + 0.1018 *K*H*SP
@ 2HP, HPS = 877.516*K*H + 0.08146*K*H*SP
@ 1HP, HPS = 731.263*K*H + 0.06788*K*H*SP

Comparitive HPS:
HPS HL = [K*(4400+0.428*SP)]/[2.5/H]
HPS DL = [K*(11733+1.148*SP)]/[1.5/H]
HPS FoL=[K*(7328+0.865*SP)]/[2.5/H]

Cleaning up:
HPS HL = 1760*K*H + 0.1712*K*H*SP
HPS DL = 4693.2*K*H + 0.4592*K*H*SP
HPS FoL=4885.3*K*H+0.5767*K*H*SP

Notice if you compare term by term, HPS of straight casting is ALWAYS higher. Hence, using Crusader strike is a net LOSS in HPS.

I did not know this. I used to use CS to generate HP, and the math shows this is wrong. My play style will change accordingly.

Haven't tested in game, so I'll have to get back to you on it, but the last time I checked, Crusader Strike does NOT proc seals.
Then that makes the mana cost of CS go up, and gives us even less reason to use it.
GAH! JUST ARGH!!!!!!!!!!!!!!!!!!!!!!!!!!!

Every holy paladin needs to go read this post here:

It's a well argued post that paladins should NEVER heal their beacon (exception being if someone's going to die). It's not intuitive, but all the math shows this to be true.

With this in mind, I have to completely rethink and remath spell selection. Right now, if I had to guess, we're going to change healing completely. Beacon will no longer go on the tank. The tank will be getting direct heals and we're going to beacon swap to the targets that need heals. This is so completely against everything I know as a holy paladin.

I'll repost when I figure out more.
Now that I've calmed down a bit, with regards to the hotfix:

Light of Dawn got nerfed. I don't know exactly how much, but we're going to be using it a lot less. I'd have to back derive the SP Coefficients and Base heals by sitting in front of a dummy, and I frankly don't have the time for it. By my rough estimates, it seems to be about 2/3 as weak, so you want to hit 3-4 targets instead of 2 for it to be as good as WoG. To be an effective beacon heal, it looks like you need to hit 5-6 to bring it past WoG. It will NOT surpass DL. This is NOT mathed out, so take it with a grain of salt.

Now, regarding the ToR nerf. For now, I'm going to ignore changing the spec. It MIGHT be viable to switch the points into Blessed Life, if BL generates more HP than ToR now. I don't know.

Anyways, my initial thoughts are that we should change our heal "rotation" to include rapid beacon swaps. I THINK this is the way to go. I've done initial play testing in heroics, but I need to figure out a good way to model it. Right now the problem is I don't have good specific and "fair" situations to compare.

Join the Conversation

Return to Forum