A Dwarf Priest

I’ve been thinking a lot about how we are going to be valuing stats post-2.4. The intertwining of intellect and character level into our mana regeneration formula throws a very large mana-filled wrench into our equations. Don’t get me wrong; the wrench is delicious. It’s just a little hard to digest.

(Warning: Long rambling math ahead! Scroll to the bottom for the quick-and-dirty summary. You have been warned.)

From finding upgrades, to getting enchants, to gemming gear, to elixirs and flasks: if I don’t know what is best, I will not be making informed decisions. I need to know how much intellect is worth 1 mp5, is worth 1 spirit, is worth +1 healing, and so forth. I need to be able to weigh stats.

Previously, I was using these stat weights for holy priests (in patch 2.3):

.15 stam / .2 int / .35 spirit / 1 mp5 /
.03 spell crit / .25 +heal / .1 spell haste

Example: A piece of gear with 20 +heal and 2 mp5 would be valued as a 7 by those numbers. A second piece of gear with 7 mp5 and 8+ heal would be rated as a 9. The second piece would be the better item.

Come patch 2.4, those numbers just aren’t going to fly. The hard part is how we value stats is going to depend on what our stats currently are. So, we are going to have to set up…

Some Test Cases

Priest Healer:

• Low Content
  430 int, 320 spirit, 1400 +heal

• Mid Content
  470 int, 470 spirit, 1750 +heal

• High Content
  540 int, 590 spirit, 2100 +heal

Stats are all unbuffed, but with assumed talents taken into account (Spirit of Redemption, Meditation, Spiritual Healing). All characters assumed to be level 70 and not human (and so not receiving the racial spirit bonus).

Roughly speaking:

• Low content -> Working through Kara, T4
• Mid content -> Working though ZA, T5, entry T6
• High content -> Working through later T6, Sunwell

Mana Regen

The most accurate model I can find of the PTR data for level 70 characters’ base regen is:

5* [.001 + 0.0093270000070334 * spirit * √(int)]
= Mana regen as a function of spirit and intellect

This formula comes from extensive research on the test realms, and is subject to change at any time.

What we want to do is run our test cases through the formula and find out how much they all gain from an increase of 1 spirit or 1 intellect (out of the five second rule).

Summary of results:

• Holy priest, low content
  • +1 spirit = 0.97 mp5
  • +1 intellect = 0.36 mp5

• Holy priest, mid content
  • +1 spirit = 1.01 mp5
  • +1 intellect = 0.51 mp5

• Holy priest, high content
  • +1 spirit = 1.08 mp5
  • +1 intellect = 0.59 mp5

This doesn’t take raid buffs into account. Let’s pretend you had mage, druid, priest, and paladin buffs on you, along with elixirs (Draenic Wisdom & Mastery), and a well fed buff. Here are the same results, but adjusted for proper buffage.

• Holy priest, low content
  • +1 spirit = 1.09 mp5
  • +1 intellect = 0.46 mp5

• Holy priest, mid content
  • +1 spirit = 1.14 mp5
  • +1 intellect = 0.60 mp5

• Holy priest, high content
  • +1 spirit = 1.21 mp5
  • +1 intellect = 0.69 mp5

Now these numbers are for outside the five second rule only. Obviously, the majority of our time is not spent standing around doing nothing, so we need to continue to modify these numbers to account for a certain amount time inside the FSR.

Healing priests will spend about 70% of their time in the five second rule. Obviously this varies from priest to priest, and depends on your raid make-up. It is fairly consistent across all levels of content (assuming you are the proper level for the content).

We will use this formula to find the “average” mp5 you get. This is just a weighted average of mana gain from within the FSR

Average regen in Mp5 :
(1-C)*R + C*M*R

Where:
C = Time spent inside the 5 second rule (70% is 0.7, etc.)
R = Regeneration outside the 5 second rule
M = % mana regen while casting (30% (or 0.3) with full meditation)

So, with 70% of the time in FSR and 30% regen:
(1-C)*R + C*M*R = .3*R + .7*.3*R = .51* R
(If you have the Primal Mooncloth set bonus, this would be: .595 * R)

So, with 30% mana regen while casting (meditation) and 70% of the time spent in the FSR:

• Holy priest, low content
  • +1 spirit = 0.56 mp5
  • +1 intellect = 0.23 mp5

• Holy priest, mid content
  • +1 spirit = 0.58 mp5
  • +1 intellect = 0.30 mp5

• Holy priest, high content
  • +1 spirit = 0.62 mp5
  • +1 intellect = 0.35 mp5

Now you should be able to see how easy it is to equate mp5, spirit, and intellect. But there is a catch! Intellect also increases your spell crit and mana pool. Spirit also increases your healing. Let’s account for those!

Mana from Intellect and Spell Crit

When adding intellect, we are extending our mana pool. Each point of intellect is worth 15 mana (16.5 if you have Blessing of Kings). However, the value of that extra mana to us depends on the length of the encounters. For low level content, we’ll say 5 minutes; mid, 7 minutes; high, 9 minutes.

(Assuming Blessing of Kings)
Mp5 equivalent from Int’s Mana =
5*(16.5* IntGain)/FightSeconds

Low content: +1 int = .2750 mp5
Mid content: +1 int = .1964 mp5
High content: +1 int = .1528 mp5

When added to our previous mp5 values for intellect, this gives us:

• Holy priest, low content
  • +1 spirit = 0.56 mp5
  • +1 intellect = 0.505 mp5

• Holy priest, mid content
  • +1 spirit = 0.58 mp5
  • +1 intellect = 0.496 mp5

• Holy priest, high content
  • +1 spirit = 0.62 mp5
  • +1 intellect = 0.503 mp5

Intellect also gives you spell crit. Now, this is difficult to weigh. Spell crit can give you some extra healing, and possibly save you mana from having to cast less due to Inspiration.

Quick run down of how Inspiration can work for a priest (with an example):

Let’s say a boss hits for 20k base damage and your tank has 16k armor. An Inspiration proc will bring that armor up to 20k. With 16k armor that tank is taking 8.5k damage; with 20k armor the tank takes 7.5k (57% damage reduction vs. 63%). That is 1/8th less damage the tank is taking, which means 1/8th less healing that has to be done … for 15 whole seconds.

Now, that was just a specific example, but it is easy to see how Inspiration can save you a couple hundred mana in the right situation. Its value obviously decreases when you are not healing a tank, or are in a magic-heavy fight. Previously I used the value of 33 spell crit rating to equal 1 mp5 for a priest (value of 1 point of spell crit rating is 0.03 mp5). That is seriously underestimating its value. Running test cases through Healcrafter and seeing how crit gear changes things, the value of 1 point of spell crit is close to 0.06 mp5.

So, what gain in efficiency does our intellect give us from spell crit?

Long Answer:
For priests, 80 intellect gives +1% spell crit. It takes 22.08 spell crit rating to equal +1% spell crit. I had previously surmised that 1 point of spell crit rating should be weighted at 0.06 mp5 for priests. 80 intellect = 22.08 crit rating = 0.06*22.08 = 1.3248 mp5 (for 80 int). So, 1 intellect also gives an effective bonus of 0.01656 mp5 from its increase to spell crit. The benefits of the bonus spell crit from intellect are very small in the larger picture. Even with the roughness of my approximation, it is easy to see how little the stat weights would change based on fluctuation one way or another.

Short Answer:
Spell crit is a very weak stat compared to mana regen.

Practical Answer:
What priest healing gear has spell crit? Spell crit is for paladins.

Let’s update the totals anyways:

When added to our previous values for intellect, this gives us:

• Holy priest, low content
  • +1 spirit = 0.560 mp5
  • +1 intellect = 0.521 mp5

• Holy priest, mid content
  • +1 spirit = 0.580 mp5
  • +1 intellect = 0.512 mp5

• Holy priest, high content
  • +1 spirit = 0.620 mp5
  • +1 intellect = 0.519 mp5

+Heal and Downranking

Now, to find the value of downranking, we need to just look at the mana saved when we drop down from one rank of a spell to the next.

Example:

Greater Heal rank 7 is 701 mana (with talents). Greater Heal rank 6 is 638 mana. The difference is 63 mana. The amount of +heal you need to grab to make rank 6 equal to what rank 7 was healing initially is 340. That equates to 5.396825 +heal per point of mana, or, weighted .1853 +heal per mp5.

The ratio in the example holds pretty close over many situations of downranking. Examining different scenarios of downranking and its effect on mana efficiency gives +heal a weight of about 0.19.

Spirit, with talents, gives 1/4 heal per point. So, spirit needs to be granted an additional 0.0475 to compensate for its +heal.

Updated weights:

• Holy priest, low content
  • +1 spirit = 0.608 mp5
  • +1 intellect = 0.521 mp5

• Holy priest, mid content
  • +1 spirit = 0.628 mp5
  • +1 intellect = 0.512 mp5

• Holy priest, high content
  • +1 spirit = 0.668 mp5
  • +1 intellect = 0.519 mp5

Shadowfiend, Innervate, and Other Mana

Here we are just going to note how these other mana gains could affect your stats. Due to inconsistencies in their usage, it is difficult to take them into account. Therefore, their effects on stats will NOT be taken into consideration for the final stat weights (below).

A shadowfiend gains .57 attack power for every point of shadow damage you have. It returns 2.5 mana for every point of damage it does. It takes 14 attack power for it to do another 1 damage each hit. It makes 10 hits in the 15 seconds it’s up.

Shadowfiend’s bonus mana gain from spell damage:
{2.5 * [(.57 * D) /14]} * 10 = 1.0179 * D
Where D = shadow spell damage

So, an increase in 1 shadow spell damage gives you 1.0179 more mana back from shadowfiend (not counting misses, crits, etc.).

Typically, your healing gear will also give +dmg in an amount equal to ~1/3 of your +heal. Also, your spirit increases your spell damage as well (spell damage bonus is equal to 1/4 of your spirit).

A priest will typically only use the shadowfiend once a fight, twice on long fights. Again, it is hard to say…

So, for every one use of a shadowfiend:

Mana gain from shadowfiend = 1.0179 * (.33*H + .25 * S)
Where H = +heal from gear, S = spirit

Generally speaking, from 1 use of your shadowfiend:
1 + heal on gear is a bonus .335 mana
1 spirit is a bonus .254 mana

This is mana, not mp5. Similar to intellect, it just extends your mana pool.

From the single use of 1 shadowfiend:

• 1 spirit = 0.0154 intellect
• 1 +heal on gear = 0.0203 intellect.

These are small values and are not taken into account in our numbers below. They should just be noted. You would have to convert from int to effective mp5 to take these into account in the stat weights (and don’t forget to multiply for how many times the shadowfiend is used!).

Innervate is a whole different story. Let’s examine innervate’s benefit in order to note how it would affect our stat values.

Innervate:
Increases the target’s mana regeneration by 400% and allows 100% of the target’s mana regeneration to continue while casting. Lasts 20 sec.

With innervate, you are effectively sitting at 500% of your OOFSR regen (regen from “pure” MP5 on gear will not count towards this). Now, pre-patch 2.4, an innervate would net you a couple thousand mana. In patch 2.4, the mana return from innervate will, in most cases, exceed your mana pool. That is provided Innervate doesn’t get changed on the current PTR (it’s likely it could see a change … just look at these numbers below!).

From 1 innervate, assuming full raid buffs:

• Holy priest, low content: 10,928 mana
• Holy priest, mid content: 15,199 mana
• Holy priest, high content: 19,376 mana

Now, if an innervate can completely fill your mana pool in most scenarios, it doesn’t matter what effect an increase to spirit or intellect will have on the mana gain from it. A full mana pool is a full mana pool!

There are other sneaky ways to get better regen that would be imprudent to account for in our stat values. Trinkets are a good example of this. Shadowpriests are another. It is too difficult to account for every single possible factor.

Spell Haste

Spell haste? Just add to taste!

Yes, you may think of spell haste as a flavor of the week. But bring up your talent trees and look at the description for “Divine Fury,” which you likely have points in, and I believe you will see the benefit. However, itemizing for spell haste as a priest involves dropping some mana regen or +heal on your gear to get it. It comes at a cost.

The “best benefit” scenario with spell haste and a healing priest is CoH spam in high raid content. Some priests keep “haste” sets for situational use. Because of its sometimes-a-benefit-sometimes-a-sacrifice nature, I am going to reduce my previous value of spell haste value from 0.1 down to 0.05 for high-level raid content. Yes, half of what it was before.

If haste healing gear drops in a raid, collect what you can, but don’t go out of your way. It’s nice to have a haste set assembled for encounters that benefit from it.

Stamina

This bring me to stamina. How much stamina you want really depends on your situation. In early raiding, stamina is a bit tougher to come by on gear, making it a more valuable stat. Once you get into higher raids, it is present on most items. Accordingly, I give it a higher value in lower raiding when it is more scarce. Before, I weighed stamina in as 0.15, but I believe I will change that value to be 0.2 in low raiding, 0.18 in mid raiding, and 0.15 in high raid content.

Obviously, use common sense and itemize more for it if you need it. PvP items are a great source.

Always, always, always keep extra stamina gear in your bags. Always. In raiding, many encounters will require you to up your stamina, while others will require you to maximize your healing potential. Let your healing gear be a swiss army knife; one blade just won’t cut it all of the time.

Spell Hit

… is extremely situational. For raiding, only grab this for when you need to CC. The general rule of thumb is to get 38 spell hit rating (3% spell hit) or 26/2% if you’re a draenei. Don’t go out of your way to itemize this. Just use some old gloves, gem them for spell hit, and put the spell hit enchant on them. Grab a trinket that gives you spell hit if that doesn’t work. We aren’t even going to factor spell hit in to our weights.

Summary

Here are our final stat weights:

Low Raid Priest (T4ish Content):

0.61 spirit /0.52 int / 0.2 stam / 1 mp5
0.19 + heal / 0.06 spell crit / 0.02 spell haste

Mid-Raid Priest (T5ish Content):

0.63 spirit /0.51 int / 0.18 stam / 1 mp5
0.19 + heal / 0.06 spell crit / 0.03 spell haste

High-Raid Priest (T6ish Content):

0.67 spirit /0.52 int / 0.15 stam / 1 mp5
0.19 + heal / 0.06 spell crit / 0.05 spell haste

Note: Human priests should multiply the spirit weight by 1.05 due to their racial bonus (0.64 for low; 0.66 for mid; 0.70 for high).

Use these numbers when looking for gear upgrades. Great websites like lootrank.com can help you use stat weights to determine “best” gear.

Important Note:
A person could easily surmise that the simplest way to get the best gear is to stack spirit (since spirit is cheaply itemized and highly valued). That is wrong! Balance your stats. Generally speaking, +heal gives you throughput, and mana regen stats gives you efficiency. All efficiency with no throughput makes you weak, and vice-versa. If your HPS (healing per second) can’t keep a tank alive, then you lose. To succeed, you need both strength of heal and longevity of mana pool. Focusing too much in any given area defeats the whole goal of “improvement.”

If we want to be lazy take a singular approach for all raiding priests, we can generalize across the values above to come up with:

0.64 spirit / 0.51 int / 0.18 stamina / 1 mp5
0.19 +heal / 0.06 spell crit / 0.03 spell haste

Note: human priests should use 0.67 for spirit.

It is important to note that the value of spirit seems to vary the most depending on your gear’s stats. Priests in great gear get better benefit from one point of spirit than a priest in mediocre gear. Spirit’s value can easily be anything from 0.55 to 0.70 depending your stats (the better your gear, the higher the value).

This isn’t an elegant approach, I know. In fact, it is down right brutal. There is no way we can take our collectively square pegs and force them into ever-changing round holes in any simple manner. I had to make some rough assumptions. As always, your results can and will vary. These are generalizations, not gospel.