Good point, you have it !

It is completly different from the others spells. it is interresting to see that the *2 coef on critical spell is not anymore applied on SS and that the RNG behavior on critical does not exist on SS.

Mastery appears as a very strong stat as it is on PW:S (higher than intel) and Critical stat is more or less the same.
I have estimated +1% Hps for each 100 mastery point added and +0,57% Hps for 100 critical point
(of course, it depends on clothes and lvl)

For haste, first approach is to say that the duration of the stacking is too short to win a spell cast, So haste is 0.
But depending on what spell is used to cast, haste can help to add a spell in stacking just before the end of SS.
So, depending on which spell are used during SS, we might have haste value to reach (cap).
In these circonstance, haste appears as a empowered stat ! (except if the maximum of 60% HP stack is reached before SS fade)

I must say, that I love the idea of PW:Solace.

As a thought experiment, I was considering whether it was possible to use this (instead of spirit) as one's primary form of mana regen (over the base 2% mana/5 seconds). The basic idea is to stack haste (instead of spirit) so that you had enough extra throughput to weave PW:Solace regularly into your rotation without losing much healing throughput.

Here are some (probably wrong) numbers:

According to Derevka (talesofapriest), combat regen is:

Combat Regen = Total Mana * 0.02 + (1.1287 * SPI * Meditation%)

where meditation is 50% for holy and 40% for disc right now. Just to pick nice round numbers, in order to get 4% MP5 (twice the base), holy needs 10,631 spirit.

To get the same regen using PW:Solace, one would need to hit with it once every 5 seconds (on average); I'm going to assume that Blizzard will make Divine Fury apply to PW:Solace as well (otherwise we'll have to worry about misses...).

If you convert that 10,631 spirit into 10,631 haste rating, that translates into just over 21% haste (assuming the 502 haste rating/1% haste I grabbed off wowhead; I'm not on the beta...), which is not quite enough to fit in a PW:Solace every 5 seconds without losing throughput.

Assuming 0 baseline haste (unrealistic) then 21% additional haste would let you cast 6.05 seconds worth of unhasted casts in 5 seconds. Taking off the 1.5 second PW:Solace, leaves one with 4.55 seconds of casts for heals, resulting in ~9% reduction in cast time for heals, but results in some damage on the boss.

If the extra haste pushes us over a threshold for an additional renew tick (and renew is used), this would mitigate the loss in throughput. Same with Divine Hymn.

Solace regenerates mana regardless of whether or not it hits, so no worries on that front.

In absolute, your assumption is correct but you will need both.
You cannot substitute spirit for haste completly because :

- You will have always spirit on clothes even if you reforge. So haste will not be very high.
it will be impossible to reforge when spirit and haste are both on the clothes.

- More you will have convert spirit into haste, more you will have to use PW:Solace and heal part will not be suffisant.
You fight for lives and HpS not for mana regen.

- Take also into account that when you increase haste by 1%, you increase also your mana use by 1%

- impossible de keep this cycle rgarding boss event and when you will lose some PW:Solace beacause of event or players low life, it will be very dangerous for your mana.


Anyway, the concept is fine

Probably few people are working on simulators just yet, since numbers and even formulas are subject to change before release (and after release too I suppose, but more likely until then). Based on philosophy I can make some good guesses though:

They're aiming to make tanking gameplay focus on mitigation. Specifically that good tanks require less healing. Unless that means bad tanks will simply be unhealable two-shot deaths, it should mean healing focus will be about mana conservation.

As you note, intellect will likely be very strong. In early gear where the socket bonus tends to be some low amount of a secondary stat it seems likely we'll be gemming straight int. If I were a designer, I'd make the bonus almost worth it on a few pieces (e.g. +int bonus on a piece with one red and one yellow socket) and then set up metas to require a few (e.g. x% mana pool increase requiring 2 either yellow or blue). If they don't, straight int is fine for the first tier.

I anticipate int >> spi > crit >= mastery > haste for first tier (or two). Other than int for efficiency, spirit will dominate as the lone regeneration stat (which is fine, if healers aren't hungry for spirit then who is going to use all the spirit gear that drops?). Crit should outshine mastery as long as you're avoiding overheal (which I expect to be the case early on), although disc may favor mastery to shine in burst periods. Haste will fall behind if conservation trumps burst, though holy will likely want to push renew to 5 ticks to make that more HpM (and technically make it a longer buff, widening the window it can be refreshed via chakra).

Thanks for your comments.
It is clear that spirit will be probably our priority to be abble to burst.

Due to Spirit Shell, i have noticed in my simulation that mastery will be a litle bit higher than crit.
in fact it is depending on the gameplay.
The overheal , even if it is low at the first tier will always be higher than the overabsorb, so Mastery is additionnaly re enforce.

My findings from playing dungeons and a bit of raids:

All our spells are extremely expensive compared to our mana pool at 90 as disc. Our "expensive spells" cost ~20k per pop and our mana pool is 306k. In other words you will expend100% of your mana pool by casting just 15 greater heals With full 463 blue gear you are looking at maximum 8k or so spirit and a combat regen of about 10k, which basically means intensive healing will use a large chunk of your mana within a minute or so. Mana regen is going to be our primary concern, so gemming/gearing for anything else than spirit looks tough atm.

On the other hand healing is a bit weak relative to health pools compared to cataclysm. A greater heal heals for 100-120k with full 463 gear and all buffs (grace+archangel), where health pools are roughly 300-500k. As a result crit is better than it was in TBC. However just like in TBC crit is much worse than it appears to be on paper, because overhealing is still a major problem for crits, but its practically a non issue for normal heals.

The strongest throughput stat at the moment for disc appears to be haste by far. The reason is that strangely all haste buffs stack multiplicatively. So if your haste is 10% and you get borrowed time your overall haste will be 1.1*1.15 = 1.265 or 26.5% not 25%. If you cast powerinfusion and get borrowed time with 10% haste your overall haste will be 1.1*1.15*1.2 = 51.8% not 45%. So in other words our haste buffs also benefit the extra casts we get from haste on gear. Unfortunately since mana is so tight, you cant really afford much haste even if you pick solace. The other two secondary stats also boost HPM, but its not trivial to determine which one is better. My gut instinct tells me mastery, but I think that may depend a lot on playstyle.

When there is a choice right now we will pick spirit over intellect, though to be honest intellect comes automatically on all caster gear so you don't really have to choose between spirit and intellect that much. The choice is mostly between spirit, haste, crit and mastery for the most part and spirit is definitely the best secondary stat atm.

Our regen options are interesting now. Atm the formula for fiend and mindbender appears to be

number of attacks = 1+(duration)/1.5.

This is not modified by haste as borrowed time and borrowed time+PI do not increase the number of attacks. This may be a bug that will be fixed in the final version.

As thing stand mindbender returns 11 *1.3 = 14.3% of your mana every minute. Normal fiend returns 9% of your mana every minute. Thus mindbender gives you an additional 5.3% of your mana every minute. That means you need to cast 3 solances every minute to beat mindbender. That is equivalent to 10% of your cast time, compared to just 2.5% of your cast time for mindbender but solace benefits from haste. I think solace is the best regen talent in terms of available mana, but it can be very awkward to use. In general mindbender appears to be a bit shit right now compared to solace. If mindbender starts benefiting from haste things might change due to the availability of borrowed time/PI.

Using just your ballpark numbers, I come up with this comparison:
Mana pool 306k
Mana regen ~10k (mp5?)
Fight Length 6 minutes (360 seconds, 72 mp5 regen blocks)
Mana regen during fight 10k * 72 = 720k mana
Total mana for fight = 720k + 306K = 1020k
GHeal - 20k mana cost, heals 120k average (with crits?), 2.2s cast (estimating a little more haste than enough for 5th renew tick)
GHeal casts in fight 1020K / 20k = 51
Time spent casting GHeal = 51* 2.2s = 112 seconds (only spending 1/3 of the fight casting).

Clearly there's no trouble running out of mana. Time between GHeals can be filled with Solace for bonus mana to support extra casts, and our other sundry heals (using that Solace mana). I have zero support for this being a viable metric so I'm already off in wild speculation territory. Anyway, continuing...

Given an empty gem slot, is 320 spirit better than 160 int for throughput here?
1.1287 * 320 * 50% (for holy) ~= 180 mp5 --> 180 * 72 = 12,960 mana over the fight, or about 2/3 of a GHeal. I could try matching it up with another spell you could actually cast for that much, but since my numbers above really only have GHeal call it 80k healing-worth of mana?

For 160 int to do better it would have to provide over 80k more healing, or roughly 80k/160 = 500 more healing per int. Given that you're casting 51 GHeals, and probably around that many other heals. By the time you average in buffs and crits, I don't think that's going to go over about 3.5 (Gheal coeff of 219%, 10% stats, 5% mysticism, 15% twist, mastery, 15% Chakra, 25% serenity, etc). Looks like on a 6 minute fight spirit should beat int on a budget comparison.

On a side note, if anyone has a priest calculator and would like to support/shoot down these wild guesses go for it

For now looking more like Spirit > Int (~= SP) >> Mastery >= Crit > Haste for first MoP tier.

Edit: Mathed. Dropped my frame from 10 minutes to 6 minutes earlier and forgot to update the mp5 blocks from 120 down to 72 (as noted by Jeges). After reading above and noting it takes a lot of Solace to support even one heal (3 casts to get 10k back) even though I'd now have more time being filled with non-GHeal casts, turns out I'd get fewer other heals on the bonus mana than I'd imagined. That left the number of other heals around the number of GHeals cast, making the final analysis similar to my earlier estimate.

Rerun your numbers with 360/5=72 (not 120) mana regenerations and see how they compare?

In my admittedly little experience with the 1st tier of MoP raids we are going to chain casting for most of the encounter. So there is very little free time. We need to do most of our healing with our cheap spells and stack spirit to keep going.

I estimate you would need 10-12k spirit before mana the limitation stops being so harsh. Assuming things stay like they are now 10k spirit is not going to be easy to reach. You are probably looking at 7.5k spirit or so in full 463 blues and maybe 8.5-9k in full 478 epics.

Even in full epic gear its going to be a problem to chain cast with moderate usage of expensive spells.

Edit: It seems that I underestimated the amount of spirit you can get. It looks like 10k+ spirit won't be that much of a problem once you have raid gear.

I posted a calculation for this on the beta forums (eu) for PoH spam.

This is a summary:

Formula for shadowfiend and mindbender: Number of attacks = 1+(duration)/1.5, for now this is not affected by haste.

So shadowfiend attacks 9 times returning 3*9/3 = 9% of max mana per minute, while mindbender returns 1.3*11 = 14.3% of base mana per minute. Mindbender also requires 1 casts per minute while shadowfiend requires 0.33 casts per minute. The difference is 5.3% max mana per minute at a cost of 1sec of casting time per minute.

Solace adds 0.7/1.5 = 0.467% of max mana for every second you spend casting it.

A very rough way to calculate it is that solace needs to provide 5.3-0.467 = 4.833% of max mana per minute to equal the return from mindbender. This is because you get 0.467% per minute automatically by using 1s per minute needed to cast mindbender for solace. Therefore you would need to cast solace for 4.833/0.467 = 10.34s every minute to equal mindbender. However that is not really a very good way to compare the spells, because you also need to have enough time to use the mana you get from solace. Here is the detailed calculation I did for PoH:

Let assume that your mana regeneration allows you to cast a maximum of n PoHs per minute. It takes n*2.5 seconds to cast those PoHs. Since you are casting only n PoHs per minute (that is the maximum your mana regen allows you to without solace) then you have (60-n*2.5) seconds of free time per minute. In order to get additional PoHs using solace in that time, you not only have to get enough mana to cast them. There should also be enough time remaining to cast them. If you solace too much and use up all your free time, you will end the fight with excess mana without actually casting anything extra. To gain enough mana for a PoH you need to cast solace for 4.5/0.467 = 9.64s you also need 2.5s to cast the PoH, so for every 9.64+2.5 = 12.14s of free time you can cast an additional PoH, by using solace. To find the total number of PoHs you can cast given a certain amount of free time you need to divide that free time by 12.14

In contrast mindbender provides roughly 1 extra PoH per minute.

To sum it all up Solace provides (60-n*2.5)/12.14 extra PoHs, while mindbender provides 4.833/4.5 = 1.074 extra PoH.

In order for solace to be better we require that solace provides more extra PoHs (60-n*2.5)/12.14 > 1.074 => n < 18.78 PoHs

This means that solace is better than mindbender for PoH spam if your mana regeneration allows you to cast less than 18.78 PoHs every minute, which is roughly 1 PoH every 3.19 seconds.

We can make a simple approximation for a mixture of spells. If the average cost of your heals is M% of base mana and the average cast time is T and the maximum number of spells per minute is N then for solace to be better than mindbender you need:

which leads to

This means solace is only better if your mana regeneration allows you to cast a less than per minute.

Lets stick some numbers. Lets say average cost of spells is 4% and average cast time is 2.1 seconds. You need to be able to cast 22.43 or less spells per minute without solace in order for it to be better than mindbender. That is equivalent to 1 spell every 2.67 seconds.

To factor in haste, since mindbender and shadowfiend don't benefit we simply need to modify the return from solace to 0.467*(1+haste) and the cast time to T/(1+haste)

Thus the relationship becomes

the final formula is then

Basically haste is effectively equivalent to having an increased amount of free time as one might expect.

What you did with the fight duration is kinda meaningless. Yes shadowfiend return goes up and down quite a lot depending on the fight time, while mindbender return is smoother, but not all fights always last the same amount of time. And things can change quite a lot even with a 30s difference, because it can easily mean an extra minbender, which could change the relationship dramatically. Its not really possible to predict the fight time down to 30s intervals, so if you include the fight time into the calculation you can end up being spectacularly wrong.

Thus to obtain a useful comparison one needs to average out these differences. Then the relationship between mindbender and solace depends only on the frequency of spell casts (i.e. cast time and mana cost of the spells you are using and your mana regeneration).

Edited in response to posts below