Plus Heal|Site Info
of 4

Healycrafting for MoP

39 replies
427 posts
Untagged users
I decided to start a series of posts on our spells and their status, and stats etc, in MoP. Well it was going to be just one post but kind of got longer and longer until it turned into a series of posts. Here's the first two, more to come shortly. I'll probably try to keep a list of links at the beginning to all the posts. Even if you are math averse, I hope you can get something out of all this. I promise that the fourth post will have cool graphs. Maybe even kittens, we'll see about the kittens.
This will be from the Priest point of view first, but I'll try to expand to the other healers as time permits and depending on interest.

Alrighty, let's talk a bit about current status of the effects of our stats on our spells, in the beta. But before we do that, there is one cool mathematical idea that I need to get across, that will make understanding everything that will follow much easier. And the idea is simply this: Relative increases are easy to work out on multiplicative effects formulas.

That sounded gibberish didn't it? Well let's take a look at the formula that gives us the average amount of healing produced by a holy priest casting a Heal:
H = (1+MP)*(1+CP)*(B+c*SP)
Here MP is the mastery percentage, CP is the crit percentage, SP is our spellpower, and B,c are two constants that depend on the spell under discussion. This is what I call a "multiplicative effects formula", as the effects from each stat come in these distinct terms that all have a similar form, and are all multiplied together.

Now how do we work with such a complicated formula? If we knew someone's mastery, crit and spellpower, we could put those in and get a number. If we now wanted to see what effect adding say 1% mastery has, we could put those new numbers in, take the difference, and see the result. But there is a better way, a lighter way, a simpler way, a way that requires less computation, and one that carries a pretty fancy mathematical name, which I shall not utter at the moment, or at any time in this post.

The idea is this: Let's not measure how much more our cast will heal for, but instead how much more it will heal relative to its current effect. In other words we won't care if Heal heals for 200 more, instead we'll care that it heals 1% more.

Sounds more complicated you say? Maybe at first, but let's see how the formulas play out: Say I decide to increase my mastery percent by an amount DM, say DM=1%, so to take it from MP to MP+DM. Then my healing has changed by:
DH = DM*(1+CP)*(B+c*SP)
The other terms remain unaffected!!! That's cool, and awesome, and useful. If we now look at how much this DH is out of our H, we get a lot of like terms canceling:
DH / H = DM / (1+MP)
That's right, no more crit showing up, no more spellpower showing up. The effect of an increase in one stat on the relative amount of healing only depends only on that stat. For instance, if we have a current mastery percentage of 25%, then we would compute 1/(1+0.25) = 0.8. So our formula becomes:
DH / H = 0.8 * DM
So for instance an increase of our mastery to 1% more will increase our healing amount by 0.08%. No matter what our crit and spellpower are, an increase in our mastery percent of 1% will always increase our healing by 0.08% when our mastery is 25%. And the more mastery we have, the smaller this percentage increase would be.

What's that you say, you don't care about the relative increase? That may be the case, but that increase allows us to easily compare stats to each other. For instance we can find how much of a relative healing increase 1% more in our mastery does, then do the same for a 1% increase in our crit percentage. This would easily allow us to directly compare mastery in crit. The same is true for spellpower. Well, almost. There is one more thing we got to do there.

What we want to do is rewrite (B+c*SP) as:
c*(B/c + SP)
Then our overall formula for healing for holy priests would have been:
c*(1+MP)*(1+CP)*(B/c + SP)
Then if we want to look at the effect of spellpower, we would be looking at the quantity 1/(B/c + SP). It is this B/c quantity that plays a pivotal role in how spellpower affects our spells. Let's have a look at that quantity for the various spells disc has at their disposal, I will follow up later on holy. Here are the B, c, and B/c values for our main heals in MoP (someone tell me how to make tables in this forum):

Spell B c B/c Heal 9599.513 1.023 9383.688 Flash Heal 15354.635 1.642 9351.35 Greater Heal 20469.42 2.19 9346.3 Prayer of Healing 6267.5 0.671 9342.21 Prayer of Mending 5332 0.571 9338 Penance Tick 9369.03 0.635 9369 Holy Fire base 1024 1.11 922.63 Smite 2126.212 0.856 2483.8 Spirit Shell 20469.747 2.19 9346.916 Power Word Shield 17498.41 1.87 9357.439 Holy Nova 2309.832 0.247 9351.547

Don't get lost in the numbers, the key is that most of our spells have a B/c of around 9350, with 2 notable exceptions that I we'll play with in a moment. What this means is that all these spells gain the same relative increase from spellpower. That increase depends to some extent on our current spellower. For concreteness, suppose we are at spellpower 10000. Then:
1/(B/c + SP) = 1/19350 = 0.00005168
What this means is that 100 more spellpower would increase these spells' healing amount by 0.005168, or about 0.5%. When we reach 15000 spellpower, this term becomes 1/24350 = 0.000041, so at that point 100 more spellpower will increase the spells' healing amount by about 0.4%. And the more spellpower we have the smaller this increase will be. The more of a stat you got, the smaller the relative increase you get by increasing that stat.

Now to our exceptions! There are 2 of them, and they really, really stand out.

Take PoM for instance. Its B/c factor is 16771. When we compute 1/(B/c + SP) for it at 10k spellpower, we find a much smaller number 1/26771 = 0.000037. So 100 spellpower only increases our PoM by 0.37% at 10k SP, and by even less at 15k SP. PoM scales slower than other spells from spellpower.
PoM behaves like the other spells.

On the other hand, Holy Fire and Smite love spellpower. For instance B/c = 922.63 for Holy Fire, which brings its factor to 1/10922.63, or 0.0001, at 10k spellpower. So 100 more spellpower at that point increases the effect of Holy Fire by 1%, almost double that of other stats. Similarly for Smite, that factor is 1/12483 = 0.00008 at 10k spellpower, and so 100 more spellpower increases Smite's effect by 0.8%.

You should think of these results in a relative sense: What they are saying is how spellpower compares to other stats in its effect on spells. If 100 spellpower is the same as say 80 mastery for a spell like Flash Heal, then it is twice as good for Holy Fire: 100 sp would have the same effect as 160 mastery. And they are less good for the other stats.

This is important to keep in mind, because we usually balance stat weights around the effects on main spells.

On our next episode, also known as the next post in this thread, we'll take a closer look at the formulas for Spirit Shell and PWS.

EDIT: Values for PoM fixed.
EDIT 2: Updated values for Spirit Shell. Looks more like a table now.
EDIT 3: Added Holy Nova Coefficients
HealCalc MoP Healing calculator.
Heal Notes Blog
Posted Apr 19, 12 · OP · Last edited May 2, 12
427 posts
Untagged users
Here we are again! Let's talk about our two main "shield spells", Spirit Shell and Power Word: Shield. They are both affected by the Disc Mastery, affectionately called "Shield Discipline", in a very direct way: A percent of mastery of say MP increases those two spells by that amount. So if normally SS would have healed for an amount H, it will now heal for an extra MP*H, for a total healing of (1+MP)*H. Keeping in mind our previous formulas for spellpower, the total "healing" effect for those two spells would be:
TH = c * (1+MP) * (B/c + SP)
I put healing in quotes because it's a shield, and I have to work under the assumption that it will be consumed. I will not take into account the 80% that it will actually heal if it's not used by expiration. This would in fact play little role in our analysis, which is relative anyway and does not care about such constant multiplicative factors.

So now, the relative effect of an increase in mastery would be:
while the relative effect of an increase in spellpower would be:
D(SP)/(B/c + SP)

To complicate matters however, when we think of increasing our mastery, we do so by increasing our mastery rating, so we need to know how that affects the percentage. At level 85 the conversion rates are still the ones we are used to from cataclysm:
1 mastery rating --> 0.00558 mastery
179 mastery rating --> 1 mastery
1 mastery --> 2.5% shield buff (what we call MP)
1 mastery rating --> 0.00014 MP
100 mastery rating --> 0.014 = 1.4% MP

One note of advice when looking at mastery in your sheet. The percent of effect you are seeing is rounded up, but the system actually uses the true value corresponding to it. For instance I am at 17.78 mastery. This would correspond to a shield buff of 44.45%. The tooltip only shows 44%, but the actual effect in spells is that of 44.45%.

Since 100 spellpower is the same item budget as 100 mastery rating, we will compare the effects of those two. I will again use 10k spellpower as baseline, as it is about where my character is at. Then we already saw previously that this corresponds to a 0.5% increase in our overall healing. Let's see how that stacks against mastery. 100 mastery rating corresponds to an increase of 1.4% in MP. My character is at a 0.4445 MP, so, our percent increase in the effect of SS would be:
0.014 / (1+0.4445) = 0.0097
or about 1%. In other words, at my current level of mastery, mastery increases my shields by twice as much as an equivalent amount of spellpower.

Let us suppose I were able to get a lot more mastery, say another 5, or 12.5%. Then the effect of 100 mastery on SS would be:
0.014 / (1+0.5695) = 0.0089, or about 0.9%. Still a lot better than spellpower, it fact, we will likely need quite a lot more mastery to make the two equal.

The effects on PWS are exactly the same, since it has the same B/c ratio.

Now for the effect on regular heals. Divine Aegis makes things a fair bit more complicated (curse you blizz!), so I will spare you the painful bits, presenting just the factors we need to consider for each stat. For spellpower things are the same, the factor is still 1/(B/c+SP), which we worked out to mean that 100 spellpower gives us 0.5% increase in healing at 10k SP. Now on to Crit and Mastery:

Crit is the ugly one. The factor for the effect of crit is:
1/[ 1/(1.6+0.6*MP) + CP ]
oh boy, that looks complicated, but in effect it just involves our mastery a bit because of DA. And actually I have ignored the small effect that the meta gem provides, because it would have made things a lot uglier and I don't think it is such a considerable effect. Remember that for your crit percent you might want to add the extra 10% your basic heals get under Grace of Weakened Soul. I will do that for my character, who has a crit percent of 13.86%, so I will use 23.86% instead. Coupled with my 44.45% mastery, that gives:
1/ (0.5357047 + 0.2386) = 1.29
The rating conversion for crit rating is that 100 crit rating produces 0.5568% crit percentage, so multiplying that in we get that 100 crit rating increases our regular spells' healing by 0.72%.

Now for mastery. Again omitting painful details, the factor would be:
1/ [ (1+1.6*CP)/(0.6*CP) + MP ]
which for our numbers becomes:
1/ (9.651858 + 0.4445) = 0.099
so about a factor of 0.1. Those of you who have stuck with me will notice that the value of our mastery percent plays very little part in that calculation.
Now 100 mastery rating gives 1.4% more MP, so our percent increase is:
0.014*0.1 = 0.0014
or just 0.14%

So let's summarize what we got on the increase of each factor on our main spells:

100 spellpower --> 0.5%
100 mastery rating --> 1%
100 crit rating --> 0%

Single Target Heals:
100 spellpower --> 0.5%
100 mastery rating --> 0.14%
100 crit rating --> 0.72%

Now in order to compare these two each other, we sort of need to know how much of our healing comes from these two types of spells, since one type really favors mastery and the other really favors crit rating. Not having actually healed anything yet in MoP, let me make numbers up, to provide a framework for others to do their computations. Suppose for instance that 40% of our healing comes from shields, the other 60% from direct heals. Then 100 more mastery points will increase our healing by:
0.01 * 0.4 + 0.0014 * 0.6 = 0.00484, or 0.484%
This means mastery almost equivalent to spellpower, just a bit below. As far as crit rating is concerned, the benefit it provides would be:
0.0072 * 0.6 = 0.00432, or 0.432%
which would make crit inferior to both spellpower and mastery. You can adjust with your own percentages to find your own balance.

In our next installment, what does haste have to do with it all? And how is our new mana regen situation looking?
HealCalc MoP Healing calculator.
Heal Notes Blog
Posted Apr 19, 12 · OP
427 posts
Untagged users
Alrighty and we're back to look at the effects of haste. First of all, a brief recap: Our computations were based on the multiplicative nature of most healing formulas. Our basic formulas looked something like this:
Disc Shields: TotalShield = (constant) (1+MP) (B/c + SP)
Holy Heals: TotalHeal = (constant) (1+MP) (1+CP) (B/c + SP)
Disc Heals: TotalHeal = (constant) (1+1.6*CP + 0.6*MP*CP) (B/c + SP)

So ok, the disc formula one is a mess, but the main idea for most of them is that each stat effect enters the formula in a precise multiplicative form. This has a very important consequence: If we measure the relative percent increase resulting by increasing one of the stats, this only depends on the form of the factor for that stat, and is completely independent of the other multiplicative effects in the formula.

Say for instance you had a talent that increases all your healing by 10%. That would just result in an extra factor of 1.1 outside (again omitting details on how it interacts with DA but I'm trying to keep it simple). It would not affect at all the relative contribution of each of the other factors.

And now, for the Time Effect! (drumroll please)

In order to look at the effects of haste, we need to consider the HPS, healing per second, formulas. This simply amounts to dividing each of the above by the cast time (CT). For instance, the formula for Disc Shield becomes:

(constant) (1+MP) (B/c + SP) / CT

Now, how does haste affect the cast time? That's right, multiplicatively! If we have a haste percent of HP (I know this is confusing notation, but we won't be needing Health Points any time soon), then the cast time becomes:

CT = BaseCT / (1+HP)

Entering this into our above formula, we find that it becomes:

(constant) (1+MP) (B/c + SP) (1+HP) / BCT

(the haste factor was in the denominator for CT, so it will go to the numerator, as we were dividing by CT). Now the base cast time is just another constant, so I'm going to roll it as part of the bizarre constant I have in the front:

ShieldPerSec = (constant) (1+MP) (B/c + SP) (1+HP)

The haste percent effect is multiplicative, exactly like the other effects! This means two things:

1. Our estimates for the value of mastery, spellpower and crit on the relative increase of our healing remain exactly the same when we look at Healing Per Second.
2. It further means that in order to see how haste compares, we have to do the exact same computation, so here goes:

If we were to increase our haste percent by an amount DH, then the relative HPS increase will be:


Let us now work out what 100 haste rating buys us. The conversion between haste rating and haste percent is that 100 haste rating increases haste percent by about 0.0078. Using my priest as a test case, she is currently at 23.11% haste. This would make the factor equal to:
0.0078/1.2311 = 0.0063, or 0.63%. So we can amend our tables for the effects of each stat, this was for Disc with Adinne's current stats of 12 mastery, 23.31% haste, about 10k spellpower and almost 20% crit chance:

100 spellpower --> 0.5%
100 mastery rating --> 1%
100 crit rating --> 0%
100 haste rating --> 0.63%

Single Target Heals:
100 spellpower --> 0.5%
100 mastery rating --> 0.14%
100 crit rating --> 0.72%
100 haste rating --> 0.63%

This value of haste is universal. No matter which healer we consider, or what spell we consider, the percent increase of an amount of haste DH will always be:

I have glossed over one important detail here, and actually throughout, and that is the question of buffs. For haste at least, that is relatively easy to settle. Consider for instance our standard MoP haste buff: 5% more haste. This takes effect as all haste buffs in a multiplicative way: If our old haste is OldHaste, then our new haste can be found by:
(1+NewHaste) = 1.05 * (1+OldHaste)
So if you look back at our main formula for HPS above, nothing changes, we just get a different constant up front. But the relative value of a given amount of haste rating remains the same. 100 Haste Rating will increase your relative HPS by the same percentage, whether you have a haste buff or not. This fact takes a while to digest, so take a moment to think about it. Even though your increase in HPS from that 100 Haste Rating is bigger if you have the haste buff, the relative increase remains the same. And it is this relative increase that we use to compare to the other stats.

By now you are probably wondering how the other buffs we get, like +5% crit or +5 mastery etc, play into this, and the answer is an interesting one and perhaps the subject of a future post, but for now I will turn my attention to mana and the value of Spirit. But it probably deserves its own post.
HealCalc MoP Healing calculator.
Heal Notes Blog
Posted Apr 19, 12 · OP · Last edited Apr 25, 12
427 posts
Untagged users
The Mana Games (aka the Spirit World)

Comparing Spirit to the other stats is never easy. If you run oom before the fight ends, then obviously you want spirit spirit spirit. But what if you run oom because you weren't conservative with your healing? And what is the value of spirit once you are at a comfortable level? But before we get into all that, let's talk about two things: Our mana sources in MoP, and the mana costs of our spells.

This first thing to decide is on our unit of measurement. I have chosen mp5 for that. Everything will be scaled to mp5. For example, take PW:S. It has a base cast time of 1.5 seconds, namely the global GCD, and a mana cost of 6200. This means that it costs 5*6200/1.5 = 20667 mp5. If we were able to keep spamming PW:S, we would be burning our mana at this rate of 20667 mp5. I will go a bit further and express all these in terms of our base mana amount, BM, which is now also our total mana amount, BM = 100,000. This is important to do, as this amount will change with our level, but the costs of spells relative to our base mana doesn't. For instance PW:S costs 6.2% of our BM, for an mp5 cost of: 5*0.062/1.5 = 0.207*BM.
Let's do the same work for all our spells, and obtain their mp5 costs are percentages out of the total mana 100,000:

Renew: 0.026 * 5 / 1.5 = 0.087
Smite: 0.04 * 5 / 1.5 = 0.133 (Under Evangelism: 0.093)
SS: 0.06 * 5 / 2 = 0.15
Penance: 0.031 * 5 / 2 = 0.0755 (Under Evangelism: 0.053)
PoM: 0.035 * 5 / 1.5 = 0.117
PWS: 0.062 * 5 / 1.5 = 0.207
PoH: 0.05 * 5 / 2.5 = 0.1
Holy Fire: 0.02 * 5 / 2 = 0.05 (Under Evangelism: 0.035)
FH: 0.06 * 5 / 1.5 = 0.2
GH: 0.06 * 5 / 2.5 = 0.12
Heal: 0.02 * 5 / 2.5 = 0.04

So our mana usage ranges from about 0.08*BM = 8000mp5 on extremely conservative healing (even less with Heal), to 0.2*BM = 20000mp5 on extremely aggressive, Flash Heal-spamming, healing. A happy medium is at 12000mp5, which would allow us to spam GH. Boy those costs sound high don't they? And keep in mind, that haste will increase those mp5 costs by a factor of (1+HP).

Now let's discuss our mana sources.
1. Passive in-combat regeneration: For Holy that's 1250 + Spirit / 2, for Disc it's 1250 + Spirit / 4 (Make up your mind blizz!!). at 2000 spirit, this is just 2250mp5 for Holy, 1750mp5 for Disc. at 4000 spirit, it is 3250 for Holy, 2250 for Disc.
2. Initial mana 100,000. The mp5 worth of this depends on the duration of the fight:
3 mins: 100000 * 5 / (3*60) = 2778mp5
5 mins: 100000 * 5 / (5*60) = 1667mp5
7 mins: 100000 * 5 / (7*60) = 1190mp5
10mins: 100000 * 5 / (10*60) = 833mp5
3. Shadowfiend: 4min CD. Assuming it gets the standard 10 attacks each time it is out, that is 30% of our mana every 4 mins, or: 100000 * 0.3 * 5 / (4*60) = 625 mp5 on average. Its best usage would be two uses in a 5 minute fight, which would be: 100000 * 0.6 * 5 / (5*60) = 1000mp5.
4. Hymn of Hope: About 9% every 6 minutes. That's on average 125mp5. At best, with 2 hymns in 7 minutes, it is 214mp5.
5. Rapture: It appears to be 150% of your spirit back, at best every 12 seconds. That's 62.5% of our spirit as mp5. If we assume a proc every 15 seconds, that's 45% of our spirit as mp5.
6. Mana Tide: 200% of the caster's spirit for 16 seconds every 3 minutes. Assuming the caster has spirit similar to ours, then we get 100% of our spirit as mp5, for 16 seconds every 3 minutes. This averages to 8.89% of our spirit as mp5. at 4000 spirit, this is about 355.6mp5. at 2000 spirit, it is about 178mp5. I will not include it in the following however.
7. Other sources: Weapon enchants: One is the intellect enchant, "Jade Spirit". When you are at below 25% mana, it grants you 750 spirit for its duration. Assuming a 33% uptime on this, and assuming further that it will be available for 25% of the fight (shouldn't hope for more really), this amounts to 0.25*0.33*750/2 = 31mp5. Hardly worth mentioning. There seems to be no straight spirit proc enchant at the moment, unless I missed it.
8. Flasks: Looks like it is 1000 spirit right now, which would be 500mp5.
9. Food buff: Looks like another 300 spirit there, so 150mp5.
10. Potions: Two options, a straight 31500 mana pot, or a 4500*10 = 45000 concentration type pot. I will go with the latter for the computations.

So let's put it all together. There will be 4 columns, one 2 at the 2000 mana level and fight durations 5 and 10 minutes respectively, the other at the 400 mana level. This will be from the point of view of a Holy Priest.

Source --- 2k+5mins 2k+10mins 4k+5mins 4k+10mins Stock mp5 --- 1250 1250 1250 1250 Spirit --- 1000 1000 2000 2000 Initial Mana --- 1667 833 1667 833 Shadowfiend --- 1000 750 1000 750 Potion --- 750 375 750 375 Flask --- 500 500 500 500 Food buff --- 150 150 150 150 Hymn of Hope --- 150 150 150 150 Weapon Enchant --- 31 31 31 31 Total --- 6498 5039 7598 6039 Total(no initial) --- 4831 4206 5831 5206

The reason I included a column showing the totals without taking the initial mana into account is so that we can easily compute time to oom in each case. Notice that the difference in our mp5 gains between 5 minute fights and 10 minutes fights is minimal if we discount that initial mana. I will therefore stick to an average value of 4k mp5 at 2k mana. In other words our effective mp5 in battle is basically 3000mp5 + Spirit/2.
This will change if they change Shadowfiend and/or HoH to work off Spirit instead of total mana. I am further assuming that they will tone Rapture down a bit to bring it in line with Holy's regen. As it stands, Disc can go with one Rapture every 30 seconds and still catch up to Holy's regen.

It's time to OOM!

Yep you guessed it, we can't go on forever, even at 4k spirit. Let's see how long we can last at the current values. Recall that our most conservative healing estimate was 8000mp5, our worst was 20k mp5. This did not factor in the effect of haste, but let's add it now. At my 23% haste, I will be burning mana at roughly a rate of 8000*1.23 = 9840mp5. Flash-Heal spamming costs 24600mp5 under the same haste threshold. So assuming I am at 2k spirit, then I lose 9840 - 4156 = 5684mp5 every 5 seconds, or 5684/5 = 1136.8 every second. Given our mana pool of 100k this would allow me to last 100000/1136.8 = 88 seconds of constant conservative healing. OUCH!

Every 100 points of spirit increase our mp5 by about 50. It's time for some more formulas:

Our TOOM (Time to OOM) would be given by:
TOOM = ManaPool / ((ManaUse - 3000 - 0.5*Spirit)*5) = 500000 / (6840 - 0.5*Spirit)

We could just input a target spirit value here, and compute and see. We could manipulate the formula a bit, but it gets a bit complicated and I think I'd rather skip it. Let us however to one simplification to it, namely to pull that 0.5 factor out. This gives us the following formula for our TOOM:
TOOM = 1000000 / (13680 - Spirit)

Even though we haven't seen formulas with the expression in the denominator, the math works out fairly similar: An increase of D(Spi) will increase our TOOM by:
[1 / (13680 - Spi) ] * D(Spi)

For instance at a Spirit amount of 2000, this factor is 1/11680 = 0.0000856. So 100 units of spirit will increase our time TOOM by 0.856%. When we are at 4000 spirit, this will go up to about 1% TOOM increase per 100 points of spirit. And as a point of reference, our TOOM at 4k spirit would be: 1000000 /9680 = 103 seconds. Here is a table of our TOOM under different spirit amounts, and under different estimated usage patterns.

Spirit 16k use 14k use 12k use 10k use 2000 42 50 63 83 3000 43 53 67 91 4000 45 56 71 100 5000 48 59 77 111 6000 50 63 83 125

So even with very conservative casting (meaning cast all the time but only our cheapest heals), and at 6k spirit, we can still only last about 2 minutes of constant casting.

Spirit and other stats

So how should we compare spirit to other stats? This is a tricky question. One answer is, given that we can't survive any encounter by casting all the time, that spirit trumps all. But there are some natural limits: Surely having your heals buffed from the other stats might be more beneficial than being able to cast one more heal, but with all your heals along the way being smaller. Let's see if we can quantify that: Suppose that we have a healing pattern that produces a given average HPS healing, but which we can sustain for time TOOM. Then the total healing accomplished would be:
Now if we want to increase this total healing by a given percent, say 1%, we have effectively two choices:
1. Increase our HPS by 1%, via increasing spellpower, or crit or mastery (or even haste)
2. Increase our TOOM by 1%, via increasing our spirit regen.

Whether we favor throughput or mana regen, these two quantities should stay somewhat balanced. This should allow us to establish a comparison, but I will leave it at that for now, since this post ends up getting longish.

Edit: Corrected the Value of the Mana Pot in the table. The pot is better than the flask now.
HealCalc MoP Healing calculator.
Heal Notes Blog
Posted Apr 19, 12 · OP · Last edited Apr 27, 12
427 posts
Untagged users
The Paladin Factor

Alrighty, time to show Paladins some love. The first section will deal with the mathy stuff, later on I will talk about the various talents/spells that are different from Cata, though that stuff is better kept in the main thread on Pally information for MoP. If you get tired of the math feel free to jump to the last section.

First off, the basic healing formulas for Paladins are as straightforward as they come: The Pally mastery gives us a percent of every heal (with few exceptions I will ignore) as a shield, capped at 1/3rd of the Pally's health. Assuming that gets fully used, our basic healing formula looks like this:
Average Heal = (constant)*(1+MP)*(1+C)*(B/c + SP)
and our HPS formula would be:
HPS = (constant)*(1+MP)*(1+CP)*(B/c + SP) / (1+HastP)

Where MP is our mastery percent, CP is our crit percent, SP is our spellpower and HastP is our haste percent. For reference here are the rating conversions again (this is level 85):
  • 100 Mastery Rating gives us 0.558 Mastery, which gives us 0.558*0.015 = 0.00837 Mastery percent.
  • 100 Crit Rating gives us 0.005568 crit percentage.
  • 100 Haste Rating gives us 0.0078 more haste percentage.
  • All these should be compared to the contribution from 100 Spellpower.

This makes the comparison between stats fairly easy:
  • Relative benefit of Mastery is given by 1/(1+MP)
  • Relative benefit of Crit is given by 1/(1+CP)
  • Relative benefit of Spellpower is given by 1/(B/c + SP)
  • Relative benefit of Haste is given by 1/(1 + HastP)
All we got to do now is multiply each of these by the corresponding contribution that 100 rating units on that stat give, to see how they compare to each other. For that we need to know our stats, and to know the amount of B/c.

I will list all the coefficients in a moment, but for now keep in mind that for Pallies B/c is about 10180. Compare that to the about 9340 of the Priests. What this means is that at the same amount of spellpower, Pallies get percentage-wise a bit less from 100 more spellpower than Priests do. Spellpower scales slower for Pallies, at least for a while. Add to this that Priests get more from their spellpower cause of the 10% from Inner Fire, and I think we can see some scaling issues here.

To do the comparison, I will use my toon's stats. She is at 14.4% haste, 15.35% crit, 9000 spellpower, and 13.3*0.015 = 0.1995 = 19.95% mastery. The relative weights would then be:

  • 100 Mastery Rating: 0.00837 / (1+0.1995) = 0.00698, about 0.7% relative healing increase.
  • 100 Crit Rating: 0.005568 / (1+0.1535) = 0.0048, about 0.48% relative healing increase.
  • 100 Haste Rating: 0.0078 / (1+0.144) = 0.0068, about 0.68% relative hps increase.
  • 100 Spellpower: 100 / (10180 + 9000) = 0.0052, about 0.52% relative healing increase.

Remember the general golden rule: The more of a stat you have, the less of a relative increase more points in it provide, and the stronger the other stats become relative to it.

So it looks like spellpower is amongst the worst stats for Pallies, together with crit, at least at my levels of the stats. Well all the stats are fairly close, but Mastery and Haste are a bit ahead. To be fair, I should take into account all the standard buffs and compare then (I should have done this for Priests as well, oh well). Let us see:
  • 10% increase in Intellect and 10% increase in SP. The way this would enter the computation is intricate. a 10% increase in Intellect does not amount to 10% increase in spellpower, because of things like weapon SP. But ignoring for the moment that the first 10 intellect points don't give us any SP, then our SP is equal to (using a 397 weapon's 2783 SP, this will differ with the weapon):
    SP = 2783 + Int.
    Therefore a 10% increase in intellect amounts to a 0.1 * Int / (2783 + Int) = 0.1/(2783/Int + 1) relative increase in SP. At my level of 9k int, this translates to: 0.1 / (0.31 + 1) = 7.63% increase in spellpower. Therefore our effective increase in spellpower from those two buffs is by a factor of 1.0763*1.1 = 1.184. The factor in our healing formula would then more accurately look like:
    (other factors) * (10180 + 1.184* SP)
    which we can rewrite as (other factors including 1.184) * (10180/1.184 + SP) =
    (other factors including 1.184) * (8598 + SP)
    therefore 100 Spellpower would give us:
    100/(8598 + SP) = 100/(8598 + 9000) = 0.00568, or 0.57% increase in overall healing.
    Phew, that one took longer than expected
  • 5 more mastery, translates to 7.5% more mastery. This would bring my mastery rating to 0.00837 / (1+0.1995 + 0.075) = 0.0066, or 0.66%.
  • +5% more crit. This one is fairly straightforward as well, it would put my crit at 20.35%, for a total percentage increase of 0.005568 / (1+0.2035) = 0.0046, or 0.46% relative healing increase.
  • 5% Haste is a multiplicative effect, will not alter the weight of haste.

Now here is the long awaited table of healing coefficients:

Spell Mana B c B/c Divine Light 7000 21058.553 2.069 10178.1 Holy Light 2800 11103.13 1.09 10186.36 Flash of Light 7260 15785.672 1.549 10190.88 Holy Shock 2800 14696.489 1.445 10170.58 Holy Radiance 7000 4727.637 0.465 10166.96 Word of Glory HP 5334.6341 0.5238 10184.49 Light of Dawn HP 2413.0576 0.2363 10211.84 Sacred Shield 3600 5298.7741 0.3661 14473.57 Eternal Flame HP 1744.6399 0.1713 10184.7

Mana Costs and Sources
Let's start with the standard mp5 costs of spells, assuming no haste, and then assuming my effective haste of (1.05)*(1.144) = 1.2, or 20%. To find this is easy: Divide the mana cost by the cast time, then multiply by 5. If you have a haste amount of HasteP, multiply by (1+HasteP) also.

Spell Mana Cast Time mp5 Hasted Divine Light 7000 2.5 14000 16800 Holy Light 2800 2.5 5600 6720 Flash of Light 7260 1.5 24200 29040 Holy Shock 2800 1.5 9333 11200 Holy Radiance 7000 2.5 14000 16800 Sacred Shield 3600 1.5 12000 14400

So Pallies are looking at some hefty mp5 costs from their spells, averaging around 14000mp5 give or take, or around 12000mp5 unhasted. They do get a number of free casts from Word of Glory and Light of Dawn, which can alleviate matters a bit. If we assume that about every 6th cast is free, this brings the above costs down to 11500 and 10000 instead. I will work with an average estimated mp5 cost of 11k for our spells in the rest.

Now let's look at mp5 gains:
  • Initial Mana: This is the same as in our Priest discussion. On a 5 minute fight this amounts to 1667 mp5, on a 10 minute fight it amounts to 833 mp5.
  • Stock mp5: Everyone gets 1250mp5 at all times at level 85.
  • Spirit regen: Pallies have Judgements of the Pure, which grants an extra 30% passive mana regen, bringing our regen at 80% of our Spirit. At 2000 spirit, this would be 1600mp5, at 4000 spirit it would be 3200mp5.
  • Divine Plea: A standard 12% of our mana every 2 minutes. That would be 12000 mana every 2 minutes, or on average 12000 *5 / (2*60) = 500mp5. In a 5 minute fight, I will assume 2 uses, for a 400mp5 contribution. On a 10 minute fight, I will assume 5 uses, for a 500 mp5 contribution.
  • Flask: 1000 Spirit, or 800mp5.
  • Food: 300 Spirit, or 240mp5.
  • Potion: I will use the 45000 from the concentration potion. 750mp5 on a 5 minute fight, half that on a 10 minute fight.
  • Weapon Enchant: 750 Spirit when under 25% Mana. Assuming uptime of 33% in that period, and assuming that period lasts for one quarter of the fight, this amounts to about 50mp5. A lot more if it has a better uptime.
  • Seal on Insight: Chance of 4% mana on melee strike. That's 400 mana every time it procs. Given we won't be able to swing much in the first place, and it is just a chance to proc on each swing, I will ignore this for now; it might end up being on the order of the weapon enchant.
  • Lay on Hands if glyphed. This would be a 10k mana contribution once in the fight, amounting to 167mp5 on a 5 minute fight, 83mp5 on a 10 minute fight. I will not add this to the table however.

In the following table are the contribution for 5 and 10 minute fights, and assuming 2k spirit and 4k spirit respectively.
Source --- 2k+5mins 2k+10mins 4k+5mins 4k+10mins Stock mp5 --- 1250 1250 1250 1250 Spirit --- 1600 1600 3200 3200 Initial Mana --- 1667 833 1667 833 Divine Plea --- 400 400 500 500 Flask --- 800 800 800 800 Potion --- 750 375 750 375 Food buff --- 240 240 240 240 Weapon Enchant --- 50 50 50 50 Total --- 6757 5548 8457 7248 Total(no initial) --- 5090 4715 6790 6415

We see that pallies are in a somewhat better shape than priests mana-wise, that 80% spirit regen really starts to add up. We'll see how they balance it.

Odds and Ends
Here I will discuss some of the talent and glyph tradeoffs, and noteworthy spell changes.
  1. Holy Radiance is no longer a dot, and it heals its main target for an amount and 6 targets within a range for 50% of that amount, reducing in effect for more than 6 targets. Also makes our next Holy Shock double its amount by effectively spreading a second Holy Shock evenly among nearby players. Given how much Holy Shock hits for, compared to Holy Radiance, this would seem to make HS part of our AoE rotation.
  2. level 45 talents are very interesting from a healing point of view. And it would be very tough to choose amongst them.
    • First off, Eternal Flame is absolutely amazing right now. A 3-HP Eternal Flame heals for about 10k every 3 seconds for 30 seconds, making it our most efficient heal. Add to this the fact that it behaves like a regular dot with 10 ticks, gaining extra ticks at 10%, 20%, 30% etc haste. The amount of its healing is mind-blowing, I expect it will be toned down a bit. For the cost one one GCD, we get 100k healing, more like 120k at the usual haste levels we will be running. Even though it would probably transfer to the beacon, it would be somewhat wasted in that sense since we are likely to have beaconed the tank (more on that later though).
    • Next we have Sacred Shield. Lasts for 30 seconds, and can proc an absorb on damage every 6 seconds. Assuming all 5 procs, this amounts to a hefty 60k healing or so. Add to this, that it increases the critical chance of Flash of Light by 30% on the target, increasing its average effect by about 25% (depending on your current crit). This would bring FL on a similar HPM level as DL, and will put it even further ahead in terms of HPS. I can see tank healers taking this talent on the cases where the tanks take sudden big damage, or constant heavy damage, and using FL instead of DL.
    • Finally, Selfless Healer. On successful Judgements, I mean, Judgments, (which our passive talents should allow us to do), makes FL 35% faster and 35% cheaper, and 35% stronger when used on others. This can stack up to 3 times. The first stack increase's FL's HPS by 80%, bringing it's average HPS to a whopping 83k, and it increases its HPM by 180%, making it even more efficient than a Holy Shock. I don't even want to speculate what a second stack would do. Add beacon transfer to that, and this would make us by far the best spot healers around. Now the cost of a Judgement (I mean Judgment) is fairly cheap, about 1500, and the increase in FL's HPM more than makes up for it as well as the lost gcd. I believe this would make FL and judging on lull moments regular part of our rotation. I believe this would be my talent of choice, unless something changes down the line.
  3. Beacon of Light, now there's some interesting changes here! First off, if it mana-free and it has a non-negotiable 3 second CD, which I can see them increasing because of what I am about to describe (yes, I wield that much power!). The new amazing thing about it is the glyph of beacon of light, which removes the global cooldown from BoL. What does this mean you say? It means that on every 3rd heal we do, we can also move the beacon at the same time, effectively healing 2 people, instead of having the majority of the beacon going towards overhealing the tank. Couple it with the overpowering FLs produced by Selfless Healer, and I see a strong pattern here. I see us literally looking for a person to switch the beacon to every 3 seconds, while also starting a cast on another person at the same time, which I feel would drive me crazy, not to mention running into the problems with queuing instant casts. The other use would be for beaconing the same person we are about to heal, for faster HP buildup. One thing is for certain, this would make us think a lot of who the beacon is on and whether we should move it. I can see them turning it into a 10 or 15 second CD instead.
  4. Guardian of the Ancient Kings. Now this one has a interesting change: Every time the Guardian heals, it increases your haste by 10% for a few seconds. The beta servers are down so I can't check, but I believe it was at least 4 seconds. It also stacks up to 5 times. So I see us weaving in HRs to try to capitalize on this and keep the haste buff for as long as possible, using HRs instead. I'll edit this part more once the servers are back online, where I can see which spells do and do not use up the Guardian.
  5. Glyph of Illumination: This is an interesting one. It makes your Holy Shock criticals return 1% mana, but your passive mana regen will go from 80% to 70%. This 10% loss will cost you first of all 130mp5 from the spirit on food and flask, and also further depending on your spirit. At 2k spirit, it would be an extra 200mp5 loss, at 4k spirit it is an extra 400mp5 loss. Each Holy Shock crit returns 1000 mana. Holy Shocks will be cast every 6 seconds, so that would be 833mp5 if it crit every time. At a 20% crit rate, this is 167mp5. At 30% crit rate it is 250mp5. It does not seem able to compete with passive regen. At 30% crit rate (very high) and 2k spirit (very low) they are about equal. Unless they adjust the numbers, I don't see any use for this except possibly in pvp (I wouldn't know about that). Even if we consider Holy Shock spamming during Avenging Wrath, under the effects of Sanctified Wrath, that still would add at most 111mp5 on average over the entire fight.

Well that's all for now folks, thanks for sticking with it. This was a long one.
HealCalc MoP Healing calculator.
Heal Notes Blog
Posted Apr 26, 12 · OP · Last edited Apr 27, 12
427 posts
Untagged users
Let's HoT Up!

Boy oh boy, druids were not an easy class to work with. But one thing is for certain, druids thrive on hots! Let's start with a discussion of how hots work (which has already been done in other places, not much new stuff here), when we get new ticks and so on. Then we'll discuss the peculiarities of the various druids spells, as well as look at the spell coefficients and discuss spell weights. Then we'll look at mana, and we'll close up with a discussion of the glyphs and talents. You ready? Let's do it!

Tick Tock, Tick Tock
HoTs work in a fairly simple way, once you get the hang of it. A hot has a given duration, and a given time in-between ticks. Take for instance Lifebloom, it ticks every second for 10 seconds. Each tick follows the standard rules for how much it will heal, so there are B and c coefficients related to it. The time inbetween ticks is affected by haste. So what happens is, that as we get more haste we start being able to fit in more ticks in the default duration, thus gaining extra ticks. The truth is just a tad more complicated than that, and it involves a bit of rounding:

Let us denote by T the time in-between ticks, and by N the number of ticks. Then the regular duration of the hot is D = N*T. As we get more haste, our hasted duration decreases. When we have haste H, then the time in between ticks becomes T/(1+H). At a given amount of haste, the number M of ticks we can fit in the duration D would then be:
D / [T/(1+H)]
which is equal to
D*(1+H) / T
Now, remembering that D is actually N*T, we find that the number of ticks we can fit in the duration D is given by
M = N*T*(1+H) / T = N*(1+H)
In other words, it's just the current number of ticks times the effective haste multiplier 1+H. We can also express this as:
M = N + N*H

In other words, the number of extra ticks we get is given by N*H, where N is the number of ticks. Also, in general, the number of ticks we end up with only depends on the number of ticks we have originally and our haste. The inter-tick interval and overall duration do not play a role.

Now you will say, N*H will almost certainly not be an integer. The way the system resolves that is by rounding. So we get our first extra tick when N*H = 0.5, then our second tick when N*H=1.5 etc. This gives us our desired breakpoints for haste. The first breakpoint is at H = 0.5/N, and the followup breakpoints appear every 1/N steps. Before we look at each of our hots, one thing to notice is this: When you reach that threshold, and for some time after that, your hot will appear to have a duration longer than it's supposed to. Rejuvenation for example will appear to be 13.3 seconds long instead of its base 12 second duration. Nothing unusual about, but it does throw you off the first time you notice.

Now on to our spells. Take for instance Rejuvenation. It has 4 ticks (there is an initial tick but that doesn't count as part of the hot, it's more like a direct heal) and they appear every 3 seconds, but we don't care about that. Then 0.5/N = 0.125, and 1/N = 0.25. So our first breakpoint is at 12.5%, our next is at 12.5+25 = 37.5%, our third is at 62.5% and so on. We probably can't hope to reach that breakpoint. Here's a table of the various spells, their number of ticks, and corresponding breakpoints. I will try to arrange them roughly in percentage order (I am rounding a bit to make them look nice, you can compute more precisely if you wish).

Spell Ticks Breakpoints RJ/Tranq 4 12.5% 37.5% Lifebloom 10 5% 15% 25% 35% Swift/WG 7 7.1% 21.43% 35.7% Regrowth 3 16.67% 31% 45.28%

Some people might argue, that it doesn't matter whether we reach a breakpoint or not. They are only partly right. From a "Healing per Second" point of view, any increase in haste benefits us just as much as an increase that gets us over a threshold. But from a "Healing per Cast Time", as well as a "Healing per Mana", point of view, reaching a threshold has tremendous impact. For instance crossing the Rejuv threshold bumps the HpCT and HpM amounts of Rejuv by a factor of 6/5 (don't forget the initial tick! We went from 5 to 6 overall ticks), which is a 20% increase in the spell's numbers. That also adds an extra tick to Tranquility's hot (the channeled part always ticks for only 4 times), but compared to the overall healing that Tranquility does, it's not that huge of a change.

For Lifebloom on the other hand, reaching a breakpoint does not matter all that much possibly, depending on usage. In most cases, Lifebloom is not allowed to bloom, and instead keeps getting refreshed by other spells. From that point of view, we really don't care at all how many times Lifebloom would tick if it was allowed to expire, we just care about how often the ticks are ticking. On the other hand, on a Tree of Life situation, we might want to blanket the raid in lifeblooms and then let them bloom. in that case, getting that second tick would increase the hot contribution by about 12/11, or what amounts to a 9% increase to the hot, about half that to the overall heal, since the bloom is a considerable part of the heal. We'll talk more about this when we discuss glyphs at the end.

The breakpoint for 7 ticks might be interesting to reach, since it involves two of our main aoes, the Swiftmend-AoE-Tick component, formerly known as Efflorescence, and the staple of druid healing, Wild Growth. Going from 8 to 9 ticks would increase our aoe capabilities by a respectable 12.5%.

The druid mastery and the druid spells
The druid mastery is pretty terrific, and relatively straightforward. It flat out increases you direct heals by 1.25% per mastery point. Also, as long as you cast a direct heal every 10 seconds, it also increases your hots by the same amount. The CD on swiftmend is 15 seconds and you will want to be using that pretty much whenever it's up, since the aoe it leaves is pretty amazing. This leaves you with just needing to cast one more direct heal inbetween. Given Omen of Clarity clearcasting procs that will undoubtedly occur from time to time, there is absolutely no reason not to have a very very high uptime on the Harmony mastery buff. Some things to note about mastery and our spells:
  • Hots are based off the Harmony effect present when they were placed or refreshed.
  • Efflorescence (YES I will still call it that, OK?) effectively double-dips on our mastery. First off, the Swiftmend that gives rise to it is a direct heal that benefits from the mastery, and since Efflorescence healing is proportional to the size of the Swiftmend, it too is getting the same increase. But Efflorescence is also a hot (and a very HOT hot at that!), so it also benefits from the Harmony buff, in fact you don't even have to worry about getting the buff up, your casting of the Swiftmend just did it.
  • Our Wild Mushrooms appear unaffected by our mastery. By the way, while we are on the subject, these little buggers are terrific. According to their tooltip, they heal every person within range for about 10k for each mushroom, scaling with spellpower, and possibly critting. If they are to hit 10 people, that would be a 100k heal minimum. Of course their usefulness is radically reduced when people are not stacked.
  • The Tranquility channel appears to behave like a dot.
  • The Lifebloom bloom part does not appear to be affected by either mastery.
  • Cenarion Ward is unaffected by mastery.

Now let's talk stat comparison. We will need to examine a couple of different formulas. Let's start with direct heals, which would include Healing Touch, the direct part of Regrowth, Nourish and Swiftmend. This is the standard healing formula with one exception: Crits will proc a 30% Living Seed effect, which I will assume is fully used. This changes the value of crit, and makes our healing formula look like this:
Average Healing = (1+MP) * (1+1.6*CP) * (B+ c*SP)
Or reworking it to pull factors out, and keeping in mind that 1/1.6 = 0.625, we have:
Average Healing = (constants) * (1+MP) * (0.625 + CP) * (B/c + SP)
and for HPS we would be further dividing by (1+HasteP), resulting in a similar factor from Haste.

So these are the quantities we would need to compute to see what relative healing increase a fixed amount of the stat would give us. I will use my druid's stats (Arenwen on US Exodar):
MP: 14.67*0.0125 = 18.34%, CP = 18.68%, SP = 8234, HasteP = 15.63%
  • Mastery: 1/(1+MP) = 0.845. 100 mastery rating is 0.558 mastery, or 0.558*0.0125 = 0.006975 in MP. It will therefore increase the healing by 0.006975 * 0.845 = 0.0059, or about 0.6%.
  • Crit: 1/(0.625 + CP) = 1/0.8118 = 1.232. 100 crit rating gives 0.005568 CP, and will increase the healing by 0.005568 * 1.232 = 0.00686, or 0.686%.
  • Spellpower: 1/(B/c + SP). We will see later on the table that the B/c factor for the direct spells is about 8460, so this becomes 1/ (8460 + 8234) = 0.000059. So 100 SP gives us 0.59% increase.
  • Haste: 1/(1+HasteP) = 0.865. 100 haste rating is 0.0078 haste percent, which in turns gives us a 0.0078*0.865 = 0.00675, or 0.675% increase in our HPS.

So all in all for direct heals, all our stats are fairly balanced, 100 units of each giving us between 0.6% and 0.68% relative healing increase.

In this computation I did not factor in Regrowth's built in +60% chance to crit. That would considerably diminish crit's usefulness for Regrowth.

Now on to a discussion of our hots. Here is how the formula for the healing effect of the ticks would play out:
(nticks) * (1+MP) * (1+CP) * (Btick + ctick * SP)
So pretty much the same as far as mastery and spellpower are concerned. The only difference is that for hots, the B/c factor seems to be different, around 9730 or so. This would change spellpower's factor a tad, to 1/ (9730 + 8234) = 0.00005567. So 100 SP gives us 0.5567% increase, it has therefore lagged a bit behind. Crit has lost its 1.6 factor, which was giving it an edge. Its formula factor would now be 1/1.1868 = 0.8426, so 100 more crit would give 0.005568 * 0.8426 = 0.0047, or 0.47%. This makes crit fall behind the other stats for hots. I will summarize these in a table once we talk about haste.

If we want to discuss the "healing per second" of hots, then haste would factor in just as before. However, if we want to discuss its "healing per cast time" effect, then as we saw earlier there are breakpoints to reach and cross, that determine haste's effect. Let's try to find a way to estimate its worth however, and here is how it would go: Suppose we have just crossed a breakpoint. How much haste rating would we need to get to the next breakpoint, and what healing increase would that achieve for us? Then average it out to what 100 haste rating would have given us if the increase was progressive instead of in a big jolt.

Let's see how to pull it off, skip the next 5 paragraphs if you are not interested. First off, the amount of haste needed to move from one haste breakpoint to the next is 1/N, where N is the initial number of ticks. Then we would ask how much rating is needed to achieve this, and also what percent healing increase this extra tick offers us. Finally we scale that down to 100 haste rating for comparison with the other stats. (We could have scaled the other stats up to that amount instead, with much the same effect). I'll do one or two by hand so that we can see the pattern, before making a nice big table. Let us for instance take rejuv and I will for the moment ignore the initial tick. Then N = 4, so 1/N = 0.25. Now 100 haste rating buy us 0.0078 increase in hasteP, so to reach that extra 0.25 we need about 0.25/0.0078 = 32 of those 100 haste ratings (i.e. a total of 3200 HR). This is not entirely correct however, since for this part any multiplicative effects cannot be ignored as before. The effect of these "effects", is that of multiplying the haste increase from HR by a factor. For instance take the 5% raidwide haste effect. Then a given increase of HasteP from HR would in fact give us 1.05 times that. So to reach the 0.25, we only need to increase our haste to 0.25/1.05 = 0.238. This requires only 0.238/0.0078 = 30.5 of our 100-hasteR 'packages'.

Now how much of a healing increase does that provide. If we used that to go from the first extra tick to the second, we just went from 5 to 6 ticks, or from 6 to 7 counting the initial tick. that's a relative increase of 7/6 = 1.1667, or 16.67%. So the amount of increase 'per 100 haste rating packet' is 0.1667/30.5 = 0.00546, or 0.546%. This contribution will be smaller if we used it to go from 7 to 8 overall ticks, it would then be (1/7) / 30.5 = 0.0047, or 0.47%. So we can see that adding that second extra tick is barely worth it, a third extra tick certainly isn't; buffing other stats would be better.

In general if a dot has an initial tick of N times, then going from the first extra tick to the second extra tick is a relative increase of (N+2) / (N+1), or a 1/(N+1) converted to percentage. Similarly going for the third tick over the second would be a relative increase of 1/(N+2) and so on. Since those increases would require 1/N haste percent each to accomplish, this means the value of a 1 haste percent increase would be [1/(N+1)] / [1/N] = N/(N+1), or N/(N+1)/(1+HasteFactors) if we want to account for raid buffs and other haste factors. Finally, since 1 haste percent is achieved by 100 / 0.0078 haste rating, then 100 haste rating would give us a (god I wish I had a way to nicely show math equations) [N/(N+1)] * 0.0078/(1+HasteFactors) effect. So this at the end is the formula that would tell us how much of an increase 100 haste rating amounts to, when it is used to go from the first breakpoint to the second:
[N/M] * 0.0078/(1+HasteFactors)
Here N is the number of initial ticks, M is the number of actual ticks.

Assuming the haste factor of 1.05, here's what we got for various numbers of ticks:

Initial ticks 1st->2nd 2nd->3rd 3rd->4th 4 0.624% 0.52% 0.445% 7 0.683% 0.607% 0.546% 10 0.71% 0.65% 0.6%

So we can see here that past the first extra ticks, the value of the haste effort involved in getting more and more ticks reduces fast. Going from that 3rd extra tick of LB to the 4th extra is only a 0.6% increase, and we can achieve similar or better results from mastery instead. Add to this, that these breakpoints are not achieved at the same time, hence the above values cannot all be achieved at the same time, and we find what has been standard common knowledge on druid hots: Past the first extra hot, it is best to devote our resources towards other stats like mastery. The first tick in every hot is always worth it, because we need only half the resources to reach it (because of the rounding that was going on). This is why a lot of people stop after 12.5%, which gives them the extra tick for rejuv.

So to sum up, as far as hots are concerned: Reaching the first extra tick is the best use of our stats resources, past that point haste cannot keep up with mastery. The rest of our stats compare as follows:
  • 100 Mastery Rating gives about 0.6% relative increase
  • 100 Crit rating: 0.47%
  • 100 Spellpower: 0.5567%

So for our hots, mastery wins hands down. Notice how for all our hots, the B/c factor is considerably larger than that for our direct heals. This means that our hots get relatively less from our spellpower than our direct heals.

Before we move to mana questions, here's a table of the coefficients for the various spells:

Spell B c B/c HT 17296.39 2.045 8457.892 Reg 8921.982 1.053 8472.917 RegDot 710.26038 0.07272 9767.057 Nour 5714.128 0.675 8465.375 Swiftmend 11999.103 1.418 8461.991 WG.Tick 892.23831 0.09203 9695.081 RJ Tick 3815.961 0.392 9734.594 LB Tick 559.50051 0.05694 9826.142 LB Bloom 7345.8239 0.7519 9769.682 Mush 6048.0431 0.6199 9756.482 Tranq 8145.8502 0.8349 9756.678 Tranq Tick 1389.5453 0.1421 9778.644 Cenarion Ward 5564.95 0.57 9763.07

These coefficients all include the 25% buffed healing that speccing into the Resto tree gives us, but do not include the harmony buff.

Phew that was a lot, I think it's time for a short break, so go refill your coffee or grab a new pot of tea, cause there's more on the way!

Alright, everyone back? Let's do a ready check /readycheck. Darn I am not assist, oh well, let's carry on and take a light walk through the easiness that is the druid's mana sources. That is, until we get to clearcasting and the hellish nightmare that is a properly computed Revitalize effect. In fact the full effect of Revitalize could almost be considered a research paper, so I probably won't discuss it all here.

But first things first, the mp5 cost of our spells. At zero haste, they range from 5665 mp5 for a swiftmend, to 10200 for a Healing Touch, to 16900 for Rejuv. Remember that you need to multiply these costs by (1+H) to find out what they are like at a given amount of haste H.
Enter clearcasting:
Lifebloom ticks have a 4% chance to proc a free cast. In other words we expect a free proc every 25 ticks. Since ticks are every 1/(1+Haste) seconds, and the GCD is every 1.5/(1+Haste) seconds, those 25 ticks correspond to 25/1.5 = 16.67 GCDs, so we get one free GCD every 16-17 GCDS. This would amount to an about 6-7% decrease in the mp5 cost of casting. Not groundbreaking, but it all adds up.

Now let's go to the other side, and look at mana sources. The easy ones first:
  • Base mp5: 1250 as all lvl 85 toons.
  • Spirit based regen: The standard 50% of our spirit. 1000mp5 at 2000 spirit, 2000mp5 at 4000 spirit.
  • Initial mana. Just as in the other classes, depends on fight duration.
  • Flask: 1000 Spirit, or 500mp5.
  • Food: 300 Spirit, or 150mp5.
  • Potion: I will use the 45000 from the concentration potion. 750mp5 on a 5 minute fight, half that on a 10 minute fight.
  • Weapon Enchant: 750 Spirit when under 25% Mana. Assuming uptime of 33% in that period, and assuming that period lasts for one quarter of the fight, this amounts to about 31mp5. A lot more if it has a better uptime.
  • Innervate. It now gives us 20% of our mana, and we can glyph it so that we keep that 20%, and also give another target a 10%. 20% of our mana is 20000 mana, every 3 minutes. This amounts to 555.5 mp5. Pretty small compared to our other sources, but about on par with Pallies' Divine Plea, without its nasty side-effects. In the following, I will assume 2 innervates in 5 minute fights, 4 innervates in 10 minute fights. This brings it up to 667mp5.
  • Revitalize: Lifebloom and Rejuv ticks have a 20% chance to give us back 2% of our mana, but this effect can only happen every 12 seconds. Computing how often this would proc is mathematically very complex, but let's see if we can make a rough estimate of it. For that we would need to estimate how many ticks it would for a proc to occur, given its 20% chance. This is essentially a geometric distribution, meaning that on average we would expect it to happen at 1/0.2 = 5 ticks. Now, assuming we got 3 rejuvenates and a lifebloom going on, this should then happen within 2 seconds. To play it a bit more safe, let us go with 3 seconds, for an average time of 15 seconds inbetween Revitalize procs. Getting 2000 mana every 15 seconds amounts to 666.7mp5.

Alright let's put it all together in a table:
In the following table are the mana contributions for 5 and 10 minute fights, and assuming 2k spirit and 4k spirit respectively.
Source --- 2k+5mins 2k+10mins 4k+5mins 4k+10mins Stock mp5 --- 1250 1250 1250 1250 Spirit --- 1000 1000 2000 2000 Initial Mana --- 1667 833 1667 833 Revitalize --- 667 667 667 667 Innervate --- 667 667 667 667 Flask --- 500 500 500 500 Potion --- 750 375 750 375 Food buff --- 150 150 150 150 Weapon Enchant --- 31 31 31 31 Total --- 6682 5473 7682 6473 Total(no initial) --- 5015 4640 6015 5640

Compared to the other healers, druids seem to have comparative regeneration, which however will not scale well with spirit. I expect a change in how Revitalize and Innervate work.

Odds and ends, glyphs, talents, and all that jazz

Let's start by looking at new spells we are getting.
  • Ironbark: We now have a way to shield the tanks, at the cost of losing our Barkskin. This would produce a 20% damage reduction for 12 seconds, and is on a 2 minute CD shared with Barkskin. A very welcome addition.
  • Wild Mushrooms. They work like the boomy mushrooms, except that they heal. Yummy, very yummy. If you have time to put those down, their healing is considerable. According to the tooltip, the number of people in the affected area does not alter how much they each get healed for. This would become very OP in 25mans if it is true.

  • Glyph of Innervate: Makes it so that you still get 20% mana when you cast it on someone else. I imagine this would become fairly mandatory.
  • Glyph of Rejuvenation: When 3 Rejuvenates are active, it reduces Nourish cast time by 30%. Might make Nourish almost worth casting.
  • Glyph of Lifebloom: It transfers the full stacks when used on a new target. I'm not really sure I see much use of it in general, except for specialized situations. It can be helpful on tank swaps when you need to catch up on the new tank real fast.
  • Glyph of Healing Touch: Each HT reduces the remaining CD of Swiftmend by 1 second. When on tank duty, this would be quite nice.
  • Glyph of Blooming: The first of our 'tradeoff' talents. It increases LB's blooming effect by 20%, but LB can no longer be renewed by direct heals. I'm not sure how I feel about this one. It increases LB's HpCT and HpM by about 10%. Unfortunately the blooming effect does not scale with mastery. I suspect this glyph won't scale very well. The 20% increase really doesn't feel enough to justify losing the renewing ability.
  • Glyph of Regrowth: Turns regrowth into a guaranteed crit, but removes its hot component. I should probably work out more specific maths on this, but given what small a portion of Regrowth the hot is, this might very well be worth it if you are using Regrowth a lot. The guaranteed Living Seed is very nice.
  • Glyph of Wild Growth: It will heal one more target, but its CD is increased by 2 seconds. Let's try to evaluate this glyph's worth. Suppose that casting WG heals a person for an amount H. As it stands, we can accomplish this once every 8 seconds, resulting in an amount of 6*H/8 healing per second. With the glyph, this would become 7*H/10. This brings it from a 0.75*H healing per second to a 0.7*H healing per second. So assuming that WG is used on CD, the glyph is actually negative. However, if we were not going to use WG other than every 10 seconds, then the glyph is obviously an increase of about 16%.
To sum up, it does not really seem that any of our tradeoff glyphs are really worth it. I suppose that is the point of tradeoff glyphs.

The second and fourth tiers contain healing related talents.
  • Tier 2:
  • Nature's Swiftness: Every one minute, give instant cast. Also buffs the effect and duration of those spells by 50%. This might actually make it worth using on Regrowth instead of HT. In any case, this is a very nice and strong talent.
  • Renewal: A 20% instant self-heal on a 2min CD. Feels more PVPsh to me.
  • Cenarion Ward: In theory this sounds nice. Every 30 seconds you can put a "Ward" on someone, so that when they take damage a hot is placed on them with 3 ticks. It is however not affected by our mastery, and it currently doesn't seem to heal enough to make it worth it. Not when compared to Nature's Swiftness.
  • Tier 4:
  • Soul of the Forest: Provides 50% haste on the next spell used after Swiftmend. This is pretty sweet. I can see it working miracles when coupled with Wild Growth, which went from 8 to 12 ticks. Lifebloom also shot up to 17 ticks.
  • Incarnation: Supposedly buffs our ToL form. The tooltips did not provide any indication how it buffs it. My form did in fact seem "bigger" however.
  • Force of Nature: We get 3 Treants on a 1 minute CD. They cast 6 "Healing Touches" each. Each of their Healing Touch heals for about 900. That's a total of 900*6*3 = 16200 total healing every minute. This feels extremely poor. Unless they scale them up a bit, I can't see this competing with Soul of the Forest.

For me right now, Nature's Swiftness and Soul of the Forest are clear winners. So much for choices.
HealCalc MoP Healing calculator.
Heal Notes Blog
Posted Apr 26, 12 · OP · Last edited Apr 30, 12
427 posts
Untagged users
Shaman Time

Apologies for the bland title, but I couldn't come up with anything better. Anyway, this post will be all about them Shamans. Luckily for us, most of the hard work was already done in the previous posts, we'll just reap the rewards.

We'll start with a discussion of our mastery, and attempt to get some stat coefficients. Then we'll talk about mana, and close up with a discussion of the various spells and their interactions, as well as talents and glyphs. Even if you skip through most of the post, make sure to look at the Odds and Ends section.

Deep Healing
The Shammy mastery is perhaps the most interesting of the masteries. Its main principle is that it increases the effectiveness of our heals based on the health percent of the target. Each point of mastery gives a 3% amount. For example my shaman is at 1043 rating, which amounts to 1043*0.00558 = 5.82 mastery. This is in addition to the base 8 mastery, for a total of 13.82, which translates to a percentage of 41.46%. This would mean that if our target was at 0% health, our heals would be 41.46% stronger. When on the other hand our target is at 50% health, our heals would be only 0.5*41.46 = 20.73% stronger. In general, if we denote by HD the health deficit percent of our heal target (1 when target at 0% health, 0.7 when target at 30% health etc), then our heals would be multiplied by a factor of:
Our overall healing effect would then become:
This is almost correct. For direct heals, Ancestral Awakening kicks in, which makes it so that on a crit we also heal a nearby target for 30% of the heal. I cannot easily test if this new healing would also be further affected by mastery, which would cause it to double-dip on mastery. Further testing would be required on that, but for now we will assume it does not. The effect however is that for direct heals, the formula would become:
By pulling out the factors from those terms, we arrive at:
(constants)*(1/HD + MP)*(1/1.6 + CP)*(B/c + SP)
with (1/1.6 + CP) replaced by (1+CP) for non-single-target heals.

Recall that the basic idea of these multiplicative formulas, is that we can measure what relative increase to healing a given stat increase provides by considering "1 over the stat's factor". We can then compare these to each other to see how the various stats compare.
  • Spellpower: For almost all shaman heals, B/c is around 9940. At a spellpower of around 9000, this factor would be about 1/(9940+9000) = 0.0000528. So 100 spellpower increases the heals by around 0.53%. A notable exception is Healing Stream Totem, with a B/c factor of only 56.86. For HST, 100 spellpower increases healing by about 1.06%.
  • Crit: Assuming a crit percentage of say 20%, we would have a factor of 1/(1/1.6 + CP) = 1/(0.625 + 0.2) = 1.2121. Now, 100 crit rating would produce a 0.005568 increase in CP, hence the relative increase in healing we are after would be 0.005568*1.2121= 0.675% for single target spells. For non-single target the corresponding factor would be smaller, 0.005568 /(1+0.2) = 0.464%.
  • Haste: Haste has its usual effect on HPS, as was computed in earlier examples, the effect would be equal to 0.0078 * 1/(1+HastP). Assuming a haste percent of 18%, this would be equal to 0.0065, or a 0.65% increase in HPS. Haste however has one other important aspect to consider, and that is its effect on the number of ticks on our various HoT effects, such as they are. There are 3 HoT effects worth noting.
    • Riptide: It has 6 ticks initially, meaning it will get its 7th tick when 6*HasteP = 0.5, or a haste percent of 8.33%, which should be easy to achieve. The second tick would be at 25% haste, and is likely not worth reaching.
    • Healing Rain: It has 5 ticks initially. It will get its 6th tick at 0.5/5 = 10% haste, which should be reachable. The next breakpoint at 30% is probably not within reach any time soon.
    • Healing Tide Totem: Even though Healing Stream Totem is not affected by haste, the talent HTT is. It also has 5 ticks initially, so same as Healing Rain. (NOTE: the duration for HTT is claimed to be 11 seconds, which is not a multiple of the tick interval of 2 seconds. Need more testing to see how it actually works out.)
    • The Earthliving hot: It has 4 ticks initially, will get its 5th tick at 0.5/4 = 12.5% haste, which will likely be a reasonable haste goal.
  • Mastery: This is the hard one. Remember that its factor is 1/(1/HD + MP), where HD is the health deficit of our intended heal target. I will assume a mastery amount of 14, or a mastery percentage of 14*0.03 = 42%.
    • At a 20% health deficit, this factor is 1/(1/.2 + 0.42) = 1/(5.42) = 0.1845. 100 points of mastery give 0.558 mastery, or 0.01674 MP. This would give us a healing increase of 0.01674 * 0.1845 = 0.003, or 0.3%.
    • At a 50% health deficit, the healing increase would be 0.01674 * 1/2.42 = 0.00692, or 0.692%.
    • at a 80% health deficit, the healing increase would be 0.01674 * 1/1.67 = 1%
    Our actual mastery percent in fact plays little role in estimating the worth of more mastery when we heal someone at 20% deficit: We will be gaining around 0.3% increase for 100 points of mastery almost no matter our current mastery level. At the 50% deficit level, it starts playing a bit more of a role.

    If you spend the majority of the time healing people at around 70% health and above, mastery is probably somewhat weak as a stat compared to others. When we need it though, when we are going to heal someone who is at 50% health or less, mastery is the best of our stats there.

Let's summarize all those numbers for quicker rough comparison:
Stat Healing increase for 100 units Spellpower 0.53% Crit (Single target) 0.675% Crit (Non-single target) 0.464% Haste 0.65% Mastery at 20% Deficit 0.3% Mastery at 50% Deficit 0.692% Mastery at 80% Deficit 1% Haste threshold for 1st extra hot tick 12.5%

For completeness, here's a table of the spell coefficients.
Coefs: B c B/c Mana Mp5 Cost RT 4208.6544 0.4236 9935.445 3200 10667 (10351) RTDot 1988.0 0.2 9940 GHW 17105.137 1.721 9939 7260 20743 (20216) HW 9401.2188 0.9449 9949.433 1980 5657 (5130) HS 14107.336 1.418 9948.756 6320 21067 (20277) CH 6196.5533 0.6225 9954.302 4500 15000 (14298) HRTick 1743.8393 0.1765 9880.109 10780 26950 Unleash 3552.3067 0.3577 9930.966 1640 5467 (5151) Earth Shield 2975.2981 0.1892 15725.68 3800 12667 Earthliving Tick 5145.3638 0.5179 9935.053 Healing Stream 47.4565 0.8346 56.86137 4700 15667 Healing Tide 8589.954 0.962 8929.266

Mana Sources
Shamans have some interesting mana mechanics that we should discuss. The table above includes the mp5 costs of our various spells under zero haste, except that for HW and GHW I take into account the effect of Tidal Waves (The parenthesized values are those including Resurgence, as discussed below). In general, we need to multiply by 1+HasteP to find the actual costs given your haste percentage. So we see a starting mp5 cost of 5600 on our cheapest rotation, to 20000 on our expensive rotation, with a lot of our standard casts in the 10-15k range. Remember however, that at a haste rate of say 20%, these numbers need to be increased by 20%.
  • Resurgence
    Before we go on to normal mana sources, we need to discuss Resurgence. This makes it so that whenever your direct healing spells crit, you gain mana. The numbers for that are as follows:
    2634 mana on HW, GHW crits 1580 mana on HS, Riptide, Unleash Elements crits 877 mana on CH crits

    Instead of adding these into the mana resources, we will average them and remove this average from the spell costs. This average will of course depend on our crit. I will assume a crit percentage of 20%, meaning we expect every 5th heal to crit. In that eventuality, from HW and GHW we get 2634*0.2 = 527 mana on average per cast. Riptide and Unleash elements would average to 1580*0.2 = 316 mana on average per cast, which Healing Surge with its increased by 30% crit chance would average 1580*0.5 = 790 per cast. Finally, for Chain Heal, assuming all 4 heals will take place, this amounts to 877*4*0.2 = 702 mana per cast.

    To finish up on the topic of Resurgence, let us consider the effect of an increase in our crit chance. 100 more crit rating would increase our crit percentage by 0.005568, and therefore the average mana back from HW for instance would increase by 0.005568*2634 = 14.67 mana per cast. If we imagine that we were for instance spamming HW, with its reduced cast time of 1.75 seconds, this would amount to 14.67/1.75 * 5 = 42mp5. Given that 100 spirit would have given us 50 mp5, we find that Crit rating is almost as good as Spirit in terms of mana regeneration, and taking into account crit's other benefits it would probably make it a better stat to invest points in. Almost, we still have to talk about the effect of our spirit on Spirit Link for instance. (NOTE: This assumes the tooltip for Resurgence is correct, unfortunately the in game combat log was less than helpful in determining the mana gains.)
  • Telluric Currents
    This glyph requires special consideration. The glyph returns to us mana equal to 50% of the damage of our Lightning Bolt, but reduces that damage by 30%. This means that it returns 0.5*0.7 = 35% of the damage of the original LB as mana. the Given that we receive a passive 15% chance to hit on Lightning Bolt, we can practically assume that all our Lightning bolts will hit and not have to worry about the chance to miss. At my level of spellpower, Lightning Bolt hits for 9600, and costs 1420 to cast, which is fairly cheap. 35% of the damage is 3360, and assuming a 20% crit chance it averages to 3360*1.2 = 4032, meaning that LB gives us about 2612 mana back per cast. If we were to chain spam LB, this amounts to 2612/2.5*5 = 5224 mp5, scaled further up from our haste. This is absolutely massive as far as mana regen sources go. How much we actually get from it depends on how many casts we can squeeze in. If we can get one LB every 30 seconds, i.e. 1/12th of our time spend on LBs, this averages out to 435 mp5. This scales very nicely with our haste and crit.
  • Mana Tide
    The effect of Mana Tide is fairly straightforward. For 16 seconds every 3 minutes, we increase our spirit, and the spirit of everyone within 40 yards by 200% of OUR spirit. As far as our personal gains are concerned, this amounts to on average a 17.78% increase in our spirit, i.e. if we are at 2000 spirit, this amounts to roughly another 356 spirit, or 178mp5. If we are at 4000 spirit, it's twice that, or 356mp5. 100 more spirit would give us an extra 8.9mp5, bringing our mp5 gains from spirit to about 59% of our spirit. This devalues crit's mp5 benefit via Resurgence a tad.
  • Water Shield
    Water Shield offer a constant passive 407mp5 when on, and gives us 871 mana when hit (it tends to not proc from aoe effects however). I will assume the "hit part" does not happen. Water Shield can be glyphed so that it doest not give mana on hit, but so that the passive mp5 provided is increased by 10%, to 448 mp5. I will assume this glyphed value.
  • Other sources
    Our other mana sources are standard: 50% of our spirit, flask, food buff, pot and enchant behave the same as for most other classes.

Here's our standard mana table, with columns for 2k mana and 4k mana:

Source --- 2k+5mins 2k+10mins 4k+5mins 4k+10mins Stock mp5 --- 1250 1250 1250 1250 Spirit --- 1000 1000 2000 2000 Initial Mana --- 1667 833 1667 833 Flask --- 500 500 500 500 Water Shield (glyphed) --- 448 448 448 448 Potion --- 750 375 750 375 Mana Tide Totem --- 178 178 356 356 Food buff --- 150 150 150 150 Weapon Enchant --- 31 31 31 31 (Not added) Tel. Curr. (1 in 30sec) --- 435 435 435 435 Tel. Curr. (1 in 1min) --- 217 217 217 217 Total --- 5974 4765 7152 5943 Total(no initial) --- 4307 3932 5485 5110

The problem that Shamans might face, is that our mana sources don't scale at all well with Spirit. Resurgence and Telluric Currents scale well with the other stats, and will likely both be considerable sources of mana back as we gain more intellect, crit and haste. Water Shield appears to only scale with our level, which leaves us with only Spirit link to scale with our Spirit. Given the benefit to the other healers from Spirit Link, I foresee a risk of shamans having to sacrifice their own healing in order to bring stronger mana tides to the raid. Overall, excluding Telluric Currents, shamans seem to fall a bit behind the other healers in terms of mana regen.

Odds and Ends, Talents and Glyphs
Now that we got the ugly nitty gritty stuff out of the way, let's enjoy the beauty that is the shaman's available talents and glyphs. But first off, let's talk about our new or changed spells.

  • Burning Wrath. Passive, 10% extra spellpower to raid.
  • Grade of Air. Passive, +5 mastery to raid.
  • Instant Ghost Wolf. Passive. Yey!
  • Built in 20% extra healing on Earth-shielded targets.
  • Chain heals that start from a Riptided person (yes, riptided is a word, totally!) are 25% stronger WITHOUT removing the Riptide.
  • Air Totems: Two worth noting:
  • Spirit Link: 3minutes, 10 yard range. Reduces damage by 10% and redistributes health.
  • Stormlash Totem: 5minutes CD, 10 second duration, empowers everyone to do extra nature damage with their abilities. Mini-heroism.
  • Water Totems:
  • Healing Stream: Heals only one target now, but for a fair bit (8k or so at 9k sp). 15 seconds, 30 second CD. Ticks every 2 secs, with no way to reduce it. Does not crit. Costs 4700 mana. It can be glyphed to offer its target 10% magical damage reduction for 6 seconds.
  • Mana Tide: 200% of the caster's spirit for 16 seconds every 2 minutes. It amounts to a 17.78% average increase in our spirit. Discussed more at length in the Mana section above.

  • Tier 3
  • Call of the Elements: A clear winner on this Tier. On a 8 min CD, finishes the CDs of all totems that have a CD of less than 5mins (i.e. all but itself and stormlash). This means an extra Mana Tide, Healing Tide and Spirit Link Totem per fight, possibly 2 extra on long fights.
  • Tier 4: Lots of interesting talents here.
  • Elemental Mastery: 30% haste for 20 seconds every 2 mins. Not just for elemental shamans!
  • Ancestral Swiftness: Passive 5% haste all the time, and on a 1 min CD allows an instant cast. This offers the same as Elemental Mastery on average, plus the extra instant casts. Unless I know that the fight has very specific portions where I'll need to go "all out", I can't see myself picking EM instead of AS.
  • Echo of the Elements: Chance to duplicate the effect of any direct heal. The chance seems to be around 6%, translating to approximately a 6% healing increase if we assume no overhealing. It is unlikely there will be no overhealing however. AS is probably superior.
  • Tier 5: Oh boy so many wonderful options here.
  • Healing Tide Totem (1600 mana): 3 min cd, 10 second duration, pulses on 5 people every 2 secs affected by haste, healing on each tick for 17500 or so. 40 yard range. Quite powerful.
  • Ancestral Guidance: 2 min CD, 10 second duration, all your actions duplicated as healing for 40% of their effect on some nearby target. I can see this being quite nice in situations where we know for instance that 2 tanks will be taking a lot of damage at some point.
  • Conductivity: HW, GHW, HS, LB, when used on a target inside HR, make anyone under HR heal for 20%. The effectiveness of this depends of course on how many targets are inside HR, but it does not seem to be diminishing by distance. If there are 2 people inside, this already becomes better than AG, while when there are 5 or more it seems this could change our stacking AoE rotation, essentially making any single target heal superior to Chain Heal. It also makes it a lot easier for us to cast Lightning bolt for Telluric Currents, as each LB will heal each person within HR for about 1344. I can see this, combined with the Chaining glyph, radically change how we use our CH.

    Choosing between Conductivity and HTT will not be very easy. In cases where damage is fairly regular, I think Conductivity will edge ahead. It certainly will edge ahead on a fight like Ultraxion.
  • Tier 6: Shamans seem stock full of interesting talents, and the lvl 90 talents are no exception.
  • Unleashed Fury: Effectively buffs the Unleash Elements effects. For resto, this means that the increased effect on the next heal that UE provides is increased by 50%. I presume this turns the effect into 20%*1.5 = 30%, which does not sound all that. If it turns it into 20+50=70% however, this might be more interesting.
  • Primal Elementalist: Elemental Totems act as pets, are more powerful and get extra abilities. If they get some healing abilities, might be worth it, otherwise it's useless for Resto.
  • Elemental Blast: 2 sec cast damaging spell, 12 second cd, gives 1042 of a secondary stat (crit, haste or mastery) for 8 seconds. This one is interesting. It is about 5-6% increase on crit (or haste or mastery). It is on a very short CD, but takes up a whole non-healing GCD to cast. If the buff was bigger, it might be more worth it. It still sounds better than Unleashed Fury.

[*] Glyph of Chain Heal: Increases the healing done to targets after the first by 15%, but reduces the healing on the first target by 10%. Let us try to work out what this all adds up to. Let's denote by H the healing on the first target. Then as it stands now, each next target gets 30% less healing, so the second target gets 0.7*H, the third gets 0.7*0.7*H, the third 0.7*0.7*0.7*H. This adds to 2.533*H healing. Now, this glyph reduces the initial heal by 10%, to 0.9*H. It is not clear from the glyph if subsequent heals will be based on this value or on H, I will assume they are based on this value. The second heal will have a 15% increase on the 30% decreased value of that 10% decreased heal (you got all that?), for: 1.15*0.7*0.9*H = 0.805*0.9*H. The third heal would be 0.805*0.805*0.9*H, while the fourth 0.805*0.805*0.805*0.9*H. This adds up to 2.677*H. Our overall CH healing therefore goes from 2.533*H to 2.677*H. A small increase, but I'm not sure it warrants a glyph spot. Its main effect is that it evens out a bit more the healing on the various targets, with healing effectively decreasing by only about 20% from target to target. So if your first heal would have overhealed a lot, this might be worth it.
[*] Glyph of Chaining: Increases Chain Heal's jump distance by 100%, but CH now gains a 4 second CD. In cases where we spend the majority of the fight spread out, this is totally worth it. In cases where we are stacked a lot, the Conductivity talent might be better for AoE healing anyway, so we might not be using CH in the first place. In other words, Conductivity + Glyph of Chaining make a very interesting combo, with CH used to heal people who are spread out, Conductivity used to heal stacked people.
[*] Glyph of Cleansing Waters: Effective cleansing also heals your target for 4% of your health. Too little to be worth mentioning.
[*] Glyph of Healing Stream Totem: HST's heals also reduce Fire, Frost and Nature damage taken by their target by 10%. This can effectively provide this buff on about 3 people at any given time, if it keeps switching heal targets. Interesting, but keep in mind HST only has a 50% uptime at best.
[*] Glyph of Healing Wave: Heals you for 20% of the heal it does on its target. It is on par with the healing from Cleansing Waters, about 4k. Depending on how much use you get out of HW, might be worth it. There are probably better choices.
[*] Glyph of Telluric Currents: Discussed earlier. Mandatory as far as I'm concerned.
[*] Glyph of Water Shield: Discussed in the Water Shield section. Increases the passive mp5 from WS by 10%, but you no longer get mana on hit. Amounts to about 45mp5 if you never get hit.
[*] Glyph of Riptide: Now we're talking tradeoffs: Riptide no longer has a CD, but it also has no initial heal. The initial heal is about 25% of Riptide's worth, so it is a considerable reduction there. On the other hand, it will allow us to have Riptide ticking on effectively as many targets as we like, it gives us more targets for Chain Heals, and allows us to get Tidal Waves up more easily. Overall a glyph that is worth some thinking, depending also on your healing pattern.

Myself I am leaning towards Telluric Currents and Water Shield for sure, and will probably swap the third glyph depending on fight, with glyph of Chain Heal being the default.

Well there you have it folks! Thanks for sticking with it, I hope some of this information was useful.
HealCalc MoP Healing calculator.
Heal Notes Blog
Posted Apr 26, 12 · OP · Last edited May 1, 12
427 posts
Untagged users
The Monk Equation
The Monk class has been in a lot of flux, as was to be expected, so I am making this post with some trepidation, as it might all be quite out of date by tomorrow. I will try to keep things up to date. This will be a close look at the Mistweaver class and its abilities. This information is up to date based on today's beta patch.

Since this is a new class, we will start with a brief overview of the various spells and their performance at starting 85 lvls. We'll then discuss the Mistweaver mastery, and the Chi resource and mana mechanics. Finally we'll take a look at talents and glyphs.

Basic spells
The Monk's healing stance is called "Stance of the Wise Serpent". It replaces your energy bar with a standard 100k mana bar, and increases all healing by 20%. It also grants hit and expertise equal to your spirit, and attack power equal to half your spellpower. It also provides a passive ability, called Eminence, which makes all attacks heal the lowest health target in range for 50% of the damage dealt. We will discuss at some point how much healing attacking produces.

On top of this, the Monk can put down a Jade Serpent Statue. This lasts for 15 minutes, and heals targets for another 50% of any damage done, providing 100% transfer of any damage done as healing. It has a secondary 'Lightwell' effect, with 10 charges and healing whoever clicks on it for an amount, though there is discussion of removing this effect. The statue has a 3 minute CD.

Here's a list of the main spells available at the moment. The numbers are based on 6404 spellpower (The coefficients have changed, use the table below to compute), for other values check the HealCalc website.
  • Expel Harm this is solely a instant self-heal, on an 8 second CD for Mistweavers, healing for about 9-10k. It costs 2500 mana, bringing it at around 4 HPM, and generates 1 Chi. It further damages a nearby target for 100%, thereby causing another heal on someone of equal amount because of the stance and statue. Its effective HPM would therefore be around 8.
  • Soothing Mist A channeled ability, costing 1k mana per second and healing for around 16.5k per second, making it cost an efficient 16.5 HPM. Each tick also has a 20% chance to generate 1 Chi.
  • Renewing Mist A powerful dot, ticking for around 11k every 3 seconds for 21 seconds, duration renewable through Uplift further on. Caps at a maximum of 10 people. Costs 4200 mana, bringing it at around 18.33 HPM not counting the extra ticks it will be getting from haste. It currently does not crit. It has one extra important property however. It has the ability to spread to nearby targets up to 3 times. Assuming this happens once at every tick, this means one application will produce a total of 7+6+5+4 = 4*7 - (1+2+3) = 22 ticks, 26 if we have reached the first haste threshold. This makes Renewing Mist an extremely efficient heal, at a whopping 57.62 HPM, and a 161k HPCT, without even counting the Uplift effect yet. It should be expected that a Monk will be able to get these rolling on a whole 10 man group fairly easily. In terms of HPS, they amount to 3667 HPS per target they are in, 36667 HPS if ticking on a whole 10man group.
  • Uplift One of our two main Chi dumps. Costs 2 Chi, extends all Renewing Mists effects by 6 seconds, and also heals anyone with Renewing Mists on them for 14.3k. The Renewing Mists effect amounts to two extra ticks at least, (TODO: Find haste needed for this to add 3 ticks) or an extra 22k healing, per target. That's a total of effectively 36.3k healing on each target with Renewing Mists, 363k healing on a 10man raid with full mists running. On a full Renewing Mists 10man it would do 143k healing on the raid, bringing it at at 95.5k HPCT at zero haste, and without counting its mists extending effects.
  • Surging Mist This is in essence our only spell with a cast time, which is actually the same as the GCD anyway. Further, if it is cast during a Soothing Mist channel, it will be instant cast on the target of the Soothing Mist channel, and it will further not interrupt the channel. It costs 5500 mana, and heals for about 54.5k, bringing it at 10 HPM and 36.5k HPCT, it will be effectively our somewhat inefficient emergency heal. It also generates 1 Chi.
  • Enveloping Mist The other Chi dump. Costs 3 Chi, and has a 2 second cast time, but instant cast if done during Soothing Mist. Places a dot that is unaffected by haste, and ticks every second for 6 ticks. It increases the healing of Soothing Mist during its duration. (TODO: Compute its coefficients).
  • Life Cocoon A powerful shield on a 2minute CD. It shields for around 60k, and increases periodic healing by 50%, lasting for 12 seconds. Costs 4500 mana, bringing its shield effect at 13.3 HPM. As far as longish tank CDs go it's probably not the best one out there, it's nonetheless quite powerful.
  • Revival Another extremely powerful CD. It has a 3 minute CD, and heals everyone for 30k, as well as removing all magical, poison and disease effects.
  • Healing Sphere On top of being triggered by the Monk's mastery, this can be activated as an instant cast spell that can effectively be spammed. Spheres trigger when someone steps on them, and heal them for about 30.8k. At a 2k mana cost, this amounts to 15.4 HPM and 20.6k HPCT without haste.
  • Thunder Focus Tea Another Chi dump, on a 45 second CD. Though not a heal in and of itself, it doubles the healing done by the next Surging Mist or Uplift.
  • Zen Meditation A powerful raid CD, on a 3 min CD. A channeled ability that lasts 8 seconds. Reduces damage you take by 99%, and also redirects all the raid's damage to you. Quite, quite powerful.
  • Gift of the Serpent. The Mistweaver's mastery. Each of your heals has a chance to proc a Healing Sphere on a nearby injured player. More on this later.
  • Fortifying Brew A 3 min self CD. Increases health by 10%, reduces damage by 20%. Lasts 20 seconds.
  • Spinning Crane Kick Heals all nearby targets for 3712 plus a factor affected by spellpower, supposedly. Currently on beta is not affected by Spinning Crane Kick. Otherwise, this could prove to be a nice AoE heal.

Following is a table of the spell coefficients. You can compute the effect of the spells at a given spellpower by computing B+c*SP.
Note that Soothing Mist and Renewing Mist get less out of spellpower relative to what the other spells get (Higher B/c coefficient) while Expel Harm and Revival scale extremely well.

Spell B c B/c Mana Soothing Mists: 9981.687 1.016 9824.495 1000 Expel Harm 2801.923 1.912 1465.441 2500 Healing Sphere 17999.59 2.01 8955.02 2000 Life Cocoon 36011.57 3.74 9628.783 4500 Renewing Mist 7409.7350 0.5565 13314.89 4200 Surging Mist 30780.910 3.733 8245.623 5500 Revival 10064 4 2516 7000 Uplift 8305.2089 0.9389 8845.68 2chi

A Serpent's Gift

The Mistweaver's mastery gives all a Monk's heals a chance to proc a Healing Sphere. This chance is increased by 1.2% per point of mastery. It remains to be seen if all heals can proc this effect, and whether it has an internal CD or not, and whether a proc can produce its own proc, but assuming there is no CD and that all heals can proc this effect, then this can add up very very quickly. As an example, a stock mastery of 8+5 = 13 amounts to 15.6% chance, meaning this should trigger on about every 6-7 heals. Just having Renewing Mists on 10 people already guarantees a proc every 1.5 seconds or so. Channeling Soothing Mists should proc another one in its duration. I have not been able to do extensive testing on this yet, hopefully will be able to update. So far it does not seem to proc on the person receiving the heal. Will update when I know more. If anyone has combat logs they can link me from beta, along with their mastery stat at the time, that would be awesome.

Monks have an interesting "combo points" mechanic on top of their mana, called Chi. You can have up to 4 Chi. There are 3 basic abilities generating Chi:
  • Expel Harm: 2500 mana, produces 1 Chi. 1 Chi every 8 seconds, 2500 mana per Chi.
  • Soothing Mist: 1000 mana per tick, 20% chance to proc a Chi. At zero haste, 1 Chi every 7.5 seconds. 5000 mana per Chi.
  • Surging Mist: 5500 mana, produces 1 Chi. Gives 1 Chi every 1.5 seconds at zero haste, 5500 mana cost per Chi.

Given Expel Harm is very situational, our main way of producing Chi will be Soothing Mist and Surging Mist, or Jab that we will discuss later. This means that each Chi costs about 5000 mana to produce. As far as how often we can get them, under reasonable haste and counting on only Soothing Mist, we would be getting a Chi every 5-6 seconds, taking between 20 to 30 seconds to produce 4 Chi.

Chi is used mainly by Uplift, which consumes 2 Chi, and refreshes Renewing Mists by 6 seconds. It takes about 10-12 seconds to produce the necessary Chi. The key interesting mechanic of Chi is its interaction with mana. This is based on Mana Tea:

Brewing Mana Tea: Every time you consume 4 Chi, you generate one stack of Mana Tea, which lasts for 2 minutes and stacks up to 20.
Mana Tea channel: Consume one stack per second to receive 4% of your mana. Unaffected by haste. This essentially means that every Chi will eventually give us back 1% of our mana, or 1000 mana. We will treat this as a discount on the spells producing Chi:
  • Expel Harm: Effectively costs 1500 mana
  • Soothing Mist: Effective cost about 800 mana per tick
  • Surging Mist: Effective cost about 4500 mana

In terms of the effective MP5 obtained this way, if we stick to the estimate of 1 Chi every 5-6 seconds, we are looking at around 1000mp5 or so.

Mana sources
Apart from the Chi mechanic discussed above, and the standard 50% passive mana regen, Monks don't have any other special regeneration mechanics. What they do have is a lot of free healing from the refreshing of Renewing mists, Uplift and other Chi dumps. They also get double the benefit from Well Fed buffs, which is a small 150mp5. Here's our standard table, assuming as usual 2k and 4k mana.

Source --- 2k+5mins 2k+10mins 4k+5mins 4k+10mins Stock mp5 --- 1250 1250 1250 1250 Spirit --- 1000 1000 2000 2000 Initial Mana --- 1667 833 1667 833 Flask --- 500 500 500 500 Potion --- 750 375 750 375 Food buff (+extra) --- 300 300 300 300 Weapon Enchant --- 31 31 31 31 Chi (better as --- 1000 1000 1000 1000 cost reduction) Total --- 5498 4289 6498 5289 Total(no initial) --- 3831 3456 4831 4456

Odds and Ends
Here's our main healing-related talents:
  • Tier 2 (all cost 2 chi):
    • Chi Wave: bounce from the target to others within 20 yards, up to 5 targets. Heals friendly targets, damages enemy targets. 8 second CD. About 12k per target.
    • Zen Sphere: Heals target for an amount plus a tick every 2 seconds for 16 seconds. If cast again, previous one explodes healing everyone in range for an amount. Was not working on Beta so I wasn't able to get exact numbers.
    • Chi Burst: Damages the target and also heals a nearby target. Did not seem to do enough healing to justify over the alternatives.
  • Tier 3:
    • Power Strikes: Jab produces an extra Chi every 15 seconds. Jab does not heal for enough to make this worthwhile.
    • Ascension: Increases maximum Chi by 1. I can't imagine us sitting on our Chi, so not much benefit from this.
    • Chi Brew: Restores all Chi, 1.5 min CD. This amounts to 4% mana every 1.5 minutes, or 500 mp5, on top of some nice extra healing from the Chi dumps. Almost mandatory considering the alternatives.

All in all, the Monk healing talents don't feel as awesome as what some of the other classes get. However, the Monk heals feel pretty awesome in the first place.

Glyphs for Monks are not implemented yet. Will update when there is more information.

A few glyphs are ready however:
  • Glyph of Mana Tea: It makes Mana Tea instant and getting the mana from 2 stacks, but puts it on a 10 second CD. Given that it takes us longer than that to generate the necessary 8 Chis used, I imagine this would be almost mandatory, given then number of gcds saved.
  • Glyph of Renewing Mists: Renewing Mists now spread to furthest target within 40 yards instead of closest within 20 yards. Potentially interesting on fights where people are forced to be very spread out, probably not too useful otherwise.
  • Glyph of Surging Mist: Surging Mist no longer requires a target, but instead heals lowest health target within 40 yards. Not particularly appealing.

HealCalc MoP Healing calculator.
Heal Notes Blog
Posted Apr 26, 12 · OP · Last edited May 9, 12
262 posts
god I wish I was smarter.
Posted Apr 27, 12
76 posts
Untagged users
Im thinking haste is going to be ranked lower and lower after each patch in MoP. With a static mana pool, you will simply run out of heals quicker and quicker with more haste. I think this xpac and static mana pool the entire expansion, secondary stats will be about getting more heal per mana (read: crit and mastery).

Spirit should still be #1 secondary stat. Im talking crit vs mast vs haste.

Am interested to see how this turns out.
Posted Apr 27, 12
of 4
Announcement of