I'm not sure what you mean here. You can do that, and it doesn't matter that it procs off the total heal. The formula you used:

(Base + spellpower*3/3.5*1.88)*(1 + 0.5*crit + crit*0.45)

is equal to (Base + spellpower*3/3.5*1.88)*(1 + 0.95*crit) for all crit, spellpower, and base heal values. You get exactly the same answer either way. Since nothing in your HPS calculations took overhealing into consideration, there was never any need to separate the crit heal term from the divine aegis term in this particular case.

Exactly -- discontinuous is the proper term. In practice the casting times and cooldowns are usually not in synchronicity, so you'll end up with discontinuous HPS gains from haste.

I'm assuming you'd lose a Flash Heal cast every 30 seconds without Enlightenment haste (although that's just a guess). Borrowing your casting sequence and numbers, that's a % HPS gain of:

5270 / (135718 - 5270) = 4.04%

A crude estimate for 5% haste would be:

(1 / 0.95) - 1 = (20 / 19) - 1 = 1/19 = 5.26%

I suppose "a lot" is a relative term, but that difference could matter in a tight race. Due to cooldown restraints, you're only getting about 77% of your potential haste bonus. By contrast, a holy priest just spamming GH may get nearly 100%.

Note, however, that the 4.04% above does not apply to all haste values even though it may be true for 5% total haste, which is a bit dissatifying. So the only really accurate answer to how much does the 5% haste from Enlightenment gain you in terms of HPS is "It depends". If you also have spell haste from gear, it may not be 4.04% anymore. As you increase spell haste from 5%, at some point you'll be able to squeeze in a Greater Heal in place of a Flash Heal and your HPS will jump up a little. With more haste, maybe you can squeeze in another Flash Heal, etc. And what about when you pop Power Infusion?

A more interesting question, though, might be what's the maximum deviation from a crude % estimate (for all positive values of haste)? And at what haste values does that worst case scenario occur?

In Vanilla most casters had to drink after every pull, particularly beacause a lot of the fights were mana limited. It wasn't very fun, but people got used to it. In TBC it was initially the same, but eventually the gear caught up. I had to drink all the time when healing in my quest greens. Now I drink before and after bosses.

If someone who's got good gear needs to drink after every pull then we're back to Vanilla. That would be a shame. If someone in bad gear needs to drink after every pull then eventually the gear will prevent that.

This assumes that you're casting Flash Heal to proc the SoL. I suspect the more common use case you're mostly casting GHeal/CoH/Holy Nova. When Surge of Light procs, you're not going to accidentally consume the SoL and the Clearcast. More often that not, you'll probably lose the SoL if you proc SoL and HC while you wait for a good time to drop your big GHeal.

So the statement "there is zero change of managing to react to an SoL proc" is only true in the cases where you're chain casting Flash Heal.

Serendipity refunds 25%. SoL refunds 100%. You can't lose mana to SoL. And you'll lose approximately 9% effectiveness from test of faith, but only when targets are below 50%.

"very bad" is an awfully subjective phrase. Lemme see if I can come up with a metric. You're basically swapping HPM for HPS which is the root of downranking so it isn't fundamentally a bad tradeoff.

(edit) Ok, now I know why I wasn't a math major. But I would like to note that SoL procs of any spell critical, whereas HC only procs off Flash/Binding/GHeals. So depending on what you're casting, I would imagine the impact of simultaneous SoL+HC is variable. Add on top of that, SoL is 50% of your criticals while HC is, at best, 45%. Makes for a mathematical model that is beyond my ability to compute. But bottom line, having both (SoL/IHC) feels like it should increase your "next heal is free" uptime but it is beyond my ability to model.

Personally, the biggest usefulness killer for me (in raid settings) is the smaller window of opportunity to use it versus HC/IHC and the fact that you may have to change what your next spell is while HC/IHC is more like a passive buff. You don't necessarily have to even notice it proc to get some benefit from it since you can be chaining the same spell.

I'm not ready for a conclusion regarding SoL yet, but your assumption doesn't hold, I think,

Prepatch this would be true. In 3.x, however, we will be using Flash Heal a lot, at least with the current balancing. Even when tank healing, we'll be switching between FH and GH instead of switching between GH ranks. GH is way too expensive to be used blindly - at least so far. FH with the FH glyph and Serendipity is decent, so we'll use it a lot.

Even if true, you only "accidentally" lose HC/IHC procs when you're chaining FHeal and only FHeal. For example, if you're managing raid-wide damage with CoH, you'll never proc HC/IHC, but you can proc SoL. Whether or not you want to use the FHeal is really dependent on the context but I think it serves to illustrate that there are a lot of variables at play.

You fire a gheal at the tank, then you notice someone at low health and flash them. A crit on the gheal would be a disaster. You will lose your clearcast proc for nothing as well as a hasted charge and possibly some mana from a clearcast gheal overhealing and you will lose the test of faith crit. If that happens in 5% of cases its a big dent in the effectiveness of SoL. If you are chaining FH you could lose up to 50% of your clearcast procs.

Basically SoL trades off HPS, casting time and some clearcasting/test of faith/serendipity procs for some free healing. I currently estimate it to be worse than any talent it competes for points with.

As for test of faith I suggest you redo your calculations.

The 6% more healing amplifies your existing crit too. With the crit test of faith is on average a 9.6% boost. With out the crit its a 6% boost. Thus a non crit FH on a sub50% target loses 37.5% of the effectiveness of test of faith AND an extra 2.7% of procing clearcasting, since increase in crit = increase in clearcasting chance.

Basically if you use that FH on a sub50% target your gain is really quite small.


SoL is not just bad, its flipping awful, because instead of synergising with our other talents it clashes with them.

Come on... if you were relying on the FH to crit to save the sub-50% person, they've got a pretty good chance to die anyways. A SoL FH is faster and it's free. In the case where you procced IHC at the same time, you lose the chance of a crit. In the case (55% of the time) where you didn't, you gain the entire cost of a FH.

Saying that you lose "37.5% of test of faith" is needlessly obscuring the issue. Test of faith only adds what, 10-11ish% on average to your heal? So 37.5% of test of faith is about 4%?

So measure the loss of 4% of your FH against the chance that the person would have died in the next 1.3 seconds before your un-SoLed flash heal would have landed.

I agree that the talent isn't great, due to the way it mucks up some of our other procs, but it's hardly the worst two talent points in the holy tree.

Ok I made a little spreadsheet to calculate the ammount of mp5 that crit rating gives (Surge of Light and Improved Holy Concentration). I did this for Flash Heal and Greater Heal spam and included the 30% haste effect for IHC, this were the results:

IHC:
-Spamming Flash Heal gives 7,30mp5 per %crit (0,16mp5 per crit rating)
-Spamming Greater Heal gives 7,58mp5 per %crit (0,17mp5 per crit rating)

SoL:
-Spamming Flash Heal gives 11,59mp5 per %crit (0,25mp5 per crit rating)
-Spamming Greater Heal gives 6,95mp5 per %crit (0,15mp5 per crit rating)

In total:
-Spamming Flash Heal gives 18,89mp5 per %crit (0,41mp5 per crit rating)
-Spamming Greater Heal gives 14,53mp5 per %crit (0,32mp5 per crit rating)


The sheet (hope you can see it, I made it in Excel 2007 and converted it)
Free file hosting by Savefile.com

I dont see where you are getting at. You cant know ahead of time that SoL will proc, so its not a question of saving people If I was not sure the FH would land, then I would not queue a flash after the GH since I have no guarnatee that my gheal will crit AND proc SoL. I would queue an instant instead. The question is in this senario am I well served by a free instant heal? The answer is no, because I am trying to take advantage of test of faith. Basically in this senario I lose a clearcast a haste proc and the extra healing, extra chance of a proc from test of faith. Basically I made a net loss, if I did not have SoL I would get more return from that particular heal. By taking SoL I may have gotten free heals after some of my crits, but I have lost some of the effectiveness of my other talents.

It may not be the worst two talent points in the holy tree, but compared with every other talent it competes with in the holy tree its worse, when you take everything into account.

=================================

If I spam Flash and have x% crit then SoL should proc on 0.5x of my non SoL flashes.

Thus if I cast f flashes then my total number of flash heals will be f*(1+0.5x)

That means the percentage of FHs that are free is 0.5x*f/f*(1+0.5x) = 0.5x/(1+0.5x)

In contrast if I am spamming FH with IHC I have a 45% chance to proc a free heal on crit, so it means that if I cast f flashes then 0.45*x*f will be free so the percentage of free heals is 0.45xf/f = 0.45x

With x= 20%

SoL ==> 9.1% of flashes are free
IHC ==> 9% of flashes are free.

With x = 30%

SoL ==> 13% of flashes are free
IHC ==> 13.5% of flashes are free.

If I am spamming FH with both, then clearcasting will only apply to 1/(1+0.5x) of my heals. So the value of clearcasting gets reduced to 0.45x/(1+0.5x)

In addition I have a 50% chance of SoL overwriting an IHC proc, so the value of clearcasting becomes 0.225x/(1+0.5x)

In total with clearcasting and SoL together I have 0.725x/(1+0.5x) of my heals being free.

With x=20%

SoL ==> 9.1% of flashes are free
IHC ==> 9% of flashes are free.
SoL+IHC ==> 13.2% of flashes are free

By taking SoL with IHC I lose 53% of the mana regen from clearcasting. Its a net gain ofc, but much lower than the benefit of SoL on its own (47% of its value).


Lets look at gheal/fheal spam.

SoL: If I cast g% gheals and f% fheals then I will gain (f+g)*0.5x free flashes from SoL. Thus my total number of heals will be g+f', where f' = f*(1+0.5x). Thus (f+g)*0.5x/(f'+g) of my spellcasts will be a free flash heal. For an equivalence k = g/f, the proportion of my casts that will be a free flash heal is (1+k)*f*0.5*x/((k+1+0.5x)*f) = (1+k)*0.5*x/(k+1+0.5x)

IHC: Assuming I use a gheal on clearcasts then IHC will return 0.45x of my spellcasts will be free gheals.

IHC+SoL: Assuming I use a gheal on clearcasts then IHC will give me (f+g)/(f'+g)*0.45 = (1+k)*0.45*x/(k+1+0.5x) of casts will be free gheals.

I will waste a clearacsting if I follow with an FH immediately after a gheal crit (I cannot react if I get a double proc). Since I will cast f' flashes in total, the chance that I will follow a gheal with a fheal is f'/(f'+g), thus 0.5*f'/(f'+g) of my clearcasts will be wasted which is equal to 0.5*(1+0.5x)/(k+1+0.5x)

I.e. IHC will give me (1+0.5x)*(1+k)*0.225x/((k+1+0.5x)^2)

at x= 20% and k=2 (arbitrary values)

SoL ==> 9.7% of casts will be free FHs

IHC ==> 9% of casts will be free gheals

SoL+IHC ==> 9.7% of casts will be free FHs and 7.5% of casts will be free gheals.

By taking SoL I lose 17% of the mana regen from clearcasting. But clearcasting can be used with gheals giving it more power, overall I only benefit from 60% of the return SoL has on its own.

When you factor in the loss of HPS and casting time, you will find that SoL is actually a pretty awful talent.

To me your view is a bit too narrow. You do not consider at all, that a person might die while you are casting FH. Moreover, you take Test of Faith for granted, which is reasonable and nice if you want to do it but you can not generalize it, I do not think that Test of Faith is worth taking in its current state and I do not think that I´m the only one. I agree that basically FoL is quite screwed, especially concerning the anti-synergy with IHC. Nevertheless I think that FoL has more potential to save random ppl than Test of Faith does (although, of course, none of the two is anything near to what I would call reliable since both are based on crit which makes them based on rng).

Anyways, making FoL and IHC synergize with each other would require no more than taking the crit-dependency from one the talents or at least implementing a query that would prevent the FoL-FH from consuming the Clearcast.

A more general note: I am kind of afraid that there are less than 2 months to go until release. Disc gets some shape and sounds more reasonable as the last couple of builds progress, but Holy gets messed up more and more, they really should bring in some baseline, pillars of some sort, so that we at least know what can be expected from us instead of juggling with talents that have been changed fundamentally several times already.

In order to understand the power of SoL, you first need to grasp that throughput isn't your main concern. It certainly isn't your concern if you're considering casting Flash Heal.

Holy Priests natively out-throughput almost everyone else - they're at worst 2nd-3rd at any given healing task and they're ridiculously out-in-front in multi-target healing. So Blizzard can't realistically design content on the premise of stressing Holy Priest throughput without precluding other healers from even being able to heal it in content-appropriate gear.

The consequence of this is that, at any given gear level, you're almost certainly going to want to exchange that excess throughput for mana efficiency.

Now, since you're making this exchange why would you then be panic-clicking your spells to maximize the throughput you just traded away?

Inarguably, Surge of Light would be better if it could crit. But even without critting, it opens up a method of play that grants you enormous healing longevity if you're willing to approach healing as a task involving matching a fluid set of conditions rather than simply an exercise in pounding buttons.

I wish something like that happens, but with current balancing, sorry, I just don't see it.

Instead of looking at numbers and statistics (which will change anyway), let's look at what SoL is from its nature:

- Is it a baseline healing spell/mechanism? No.
- Is it an emergency facility? Would be, if you could invoke it.
- Is it a throughput increasing mechanism? No, only by chance, and only for a short interval, not on average.
- Is it a mana efficiency mechanism? Well, that depends, as already has been outlined in detail.

My reading of SoL is that right now it's a gadget, like many of these engineering items - only you can't control it. It may function, or it may explode. And you cannot even control when you invoke it.

It may save mana or cost mana. It may save people or let people die. And it's not even in your hand, except when you decide not to take the talent.

To make it a tool instead of a toy, it doesn't need to get crit capability. Getting rid of the negative interaction with HC is what's needed. Or allow Greater and Binding Heal in addition to Flash heal. Or make it an on-use feature with a cooldown.

They could also remove Flash Heal from HC/IHC and buff the proc rate. It doesn't HAVE to be a change to Surge.

SoL only functions with Flash Heal, which is the lowest throughput of your single target heals. Since SoL isn't consumed except if you cast Flash Heal, and you wouldn't be casting it during a time of stress on your throughput, there isn't actually any advantage or disadvantage to it.

In terms of throughput, Holy Priests easily out-throughput everyone on multiple targets. Priests have the four highest throughput multi-target heals - Glyph'd Holy Nova, Prayer of Healing, Circle of Healing and Binding Heal.
In terms of single targets, it doesn't take much math to demonstrate that Priests out-throughput Druids and Shaman on single targets under virtually any reasonable set of assumptions.

I don't 'button mash' currently, and that's with only the incentive of being able to heal more efficiently. With the added incentive of managing SoL, I'm not sure why I'd suddenly start button-mashing.

Without downranking if you have a target with a 4k deficit that needs to be healed quickly FH is your highest throughput spell and in wotlk it will make sense to use it.

In moments of stress you chain cast just like everyone else. If you dont have to chain cast the HPS requirement of hte encounter is trivial. So I don't buy the "I don't chain cast arguemnt" for 1 instant.

Priests always had the highest throughput aoe heals. It was not sufficient to make them the best aoe healers in TBC. Stacking specialists proved equally effective. If you think that encounters will not challenge your HPS you are mistaken.

1. Your '4k target needs to be healed quickly' argument requires someone who is down 4k - and will be dead within 2.5s. I'd suggest this scenario happens seldom-to-never. And if you're clamping the amount of healing a target can receive, you can't really claim any spell that gets clamped is higher 'throughput' than another. You can make the claim that some are more efficient than others.

2. 'Moments of stress'. Playing WoW isn't 'stress'. It's relaxation - and an almost exclusively intellectual ability. The job of a healer is to predict how they can best map a set of differently shaped components onto a fluid structure. It's basically Tetris with slightly different rules and the ability to get 'better' shapes as you play more.

3. No raid in WoW - now or ever - has challenged single target healing throughput. Not one. In fact, there isn't even any raid that's close. The highest single target dps you'll see from a raid boss is around 3k. This is only slightly more than a single Holy Paladin can muster via Holy Light.

I suspect you're confused because we're using dps metrics where they're not really appropriate. The reason we use 'throughput' and 'efficiency' is because they're enormous simplifications of some very complex data analysis that would (a) necessarily exclude 99% of the readers who don't understand statistics at any deep level and (b) require very large sets of accessible data.

Except we're talking about WotLK content, and the people are standing around with 16k health.

Some arbitrary WWS parses:

Brutallus kill by Devierum

Brutallus' single target dps is 4775 dps for this encounter.

Nalorakk death by Pugnas Rache

Nalorakk deals 465 dps to the main tank - about a fifth of the sustained hps of Greater Heal with T5-equivalent gear. The total dps for the entire encounter is around 1295 dps - less than what a T5 Holy Paladin could generate via Flash of Light.

The problem isn't that you have too little healing. The problem is being able to place it correctly and deal with lulls and bursts of damage.

This is getting ridiculous. Please have a close look at encounters before posting.

Regarding the Pugnas Nalorakk kill. Keep in mind that Pugnas go there with SWP decked tanks. Overall DPS is way higher with a tank not far overgearing the instance. Also, the critical point is not overall DPS - it's DPS in bear phase, which is a lot higher (extremely depending on avoidance luck, though). You need to be able to deliver in excess of 3000 HPS if you want to single heal, since you need to cover the short silence phases, of which you can have two in a single bear phase - depending on raid DPS. And this was an example for a really lowlevel encounter.

Regarding the Brutallus kill, two things. First, again you are pulling a log from a raid outgearing the instance, killing the boss a full minute before the enrage timer. Second, the critical phases are the stomp phases where the tanks armor is reduced to half its value, increasing incoming damage.

Even with this kind of equipment, watch this:

19:27'14.084 Brutallus melee swing hits Sinex for 6402 Physical. (649 Blocked)
19:27'14.084 Brutallus melee swing hits Sinex for 2507 Physical. (649 Blocked)
19:27'14.780 Brutallus melee swing hits Sinex for 4071 Physical.
19:27'14.780 Brutallus melee swing hits Sinex for 7815 Physical.

This is not very bad luck, this is rather typical, even for a perfectly geared tank with full support.

Addendum: This is not spike damage. It's regular damage during stomp.

Crit-healing the tank and free-instant-healing someone at low health is not a disaster. I'm not the grammar police, but I am a concerned citizen of the English-speaking world, and while many things can be described as disasters, your scenario is definitely not one of them.

I do agree that these (rather obvious) talent conflicts should be ironed out. Since it seems like the Flash Heal effect was added to SoL without much consideration of the consequences, I remain confident that at least some of these problems will be addressed, and once that happens we can complain about the remaining issues more fruitfully ^_^


A couple loosely related questions:

Twin Disciplines has been reworded to refer to spell power instead of damage and healing. In addition to making it even slightly more useless for leveling (10.5 spellpower, woohoo!), this would seem to weaken it for most spells, but make it include PW:S (which was my main complaint for why it was completely stupid). Is that accurate?

Also, has anyone tested whether it interacts with SoL Flash Heal? (or smite, for that matter)?

Holy fire on beta is critting for less than it was recently hitting for.

"Throughput" is a well-understood metric. It refers to how much raw effect you can pump out over time - heal/sec or damage/sec. For dps classes, it is a useful metric because that's really all they do. For healing classes, we use the metric because it is simple to calculate and explain. More importantly, it has classically mirrored two concepts we actually do care about - efficiency and burst - but which are incredibly difficult to calculate and require large data sets linked to specific encounters.

That is, if Player A has more throughput than Player B, it has always been true that Player A will generally be better at dealing with bursts of damage than player B. Even cross-class, this has been true - if purely by accident (Holy Light is faster than Greater Heal and coincidentally better throughput, Healing Touch is slower than Greater Heal and coincidentally lower throughput). About the only deviation would have Paladins using crit - breaking the relationship between throughput and efficiency (more critical lead to greater efficiency at the expense of throughput).

In WotLK, it is not true that throughput mirrors efficiency and burst response. With Holy Concentration and Surge of Light in particular, you can actually reduce throughput while increasing burst response.

So far everything you've said tends to lead me to believe you have a somewhat weak understanding of what I'm trying to explain - that throughput is not a reliable metric for more than rough purposes - but can't deal with the consequences of that understanding.

The current SoL is a flawed mechanic. Well, it does what it says but I think we can all agree that it's awkward and not nearly as good as it seems at first.

I have suggested the following change on eu beta forums:

Surge of Light:
Holy spells reduce the casting time and mana cost of Flash Heal and Smite by 10%. Stacks up to 10 times. Lasts 20 seconds.

I would be interested in hearing people's opinions about the above so I can continue to provide the best possible feedback and suggestions. Keep in mind that

1) I consider CoH spam with higher crit rates and Surge of Light to be too good, but also think that at the same time casting FH in a situation where CoH spam is needed is counterproductive
2) I think that due to more reasons that I want to list (it's all been said already) SoL is a bad idea that doesn't fit well into gameplay
3) I think giving us random procs (I'm also looking at you IHC) isn't really in line with our design philosphy, and is not overly beneficial to healers in the form that we have received them (HC excepted).