Topic Shaman: Deep Healing mechanic clarification
Edited by Jynus on 11/22/10 1:06 PM (PST)
I was one of the shaman like many others who said mastery was worthless a while back. But is was NOT because of how the mechanic worked. It was because mastery only effected about 25% of our total healing done at the time. AND on top of that, scaled WORSE as we gear due to increased mana allowing us to CH and HR more. So no matter how much we stacked, it was worthless as you lose out on stats that work on 100% of our heals in order to buff 25%. There was no point to it.
But then Mastery was buffed to make use of both our top 2 most common used spells. So rather than effecting only about 25% of our total healing done, it's now in the >75% range with only ELW and riptide hot being uneffected. So it's now potentially VERY viable. There is a definite use to mastery now where we can actually see benefit to having it unlike before. Now the only part to figure out is at what % hp does having mastery actually benefit us more than stats. (which I do not know at this point in time for certain)
Edited by Faveokatro on 11/22/10 3:41 PM (PST)
Happy to do so :P
First to value crit, we take the crit multiplier (150%), and the fact that AA procs on single target heals (130% * 150% = 195%). So for each 1% in crit, we either get a .95% throughput increase (on single target heals) or .5% throughput (on multi-target)
Crit Value: .5% or .95% per
To value haste, 1% haste = you can cast 1% more spells in the same window as before. In other words, 1% haste = 1% throughput.
Haste Value: 1% per
Mastery is actually not as difficult to calculate as people think it is, because it's linear. That means that at 75% health, you get healed for precisely (1-.75) * how much mastery you have. At 50% health, you get healed for (1-.5) * mastery and so on. Once you understand that, you just plug in both sides of the equation and solve.
Mastery Value: (1- Health % required for breakeven) * 2.5% is how much each % of mastery improves healing. If you don't believe the formula, just plug in 100% health: (1 - 100%) * 2.5% = 0. Or 0% health, effectively 1 hp (1-0%) * 2.5% = 2.5%.
So to calculate breakeven, it's just simply setting the equations equal to one another. At what level of health does the mastery formula equal .5%? .95%? 1%?
Haste: 1% = (1 - Health) * 2.5%
1 - Health = 1 / 2.5
Health = 1 - (1 / 2.5)
Health = 60%
Crit, Single Target: .95% = (1 - Health) * 2.5%
Health = 62%
Crit, Multi Target: .50% = (1 - Health) * 2.5%
Health = 80%
While these are the pure throughput breaking points, you have to remember there are other facts that make it excellent even above the break points.
Crit gives some amount of regen, but is RNG whereas Mastery is consistent. So even for some area above the 62% mark for single target heals, I would argue mastery holds value over crit for throughput.
Haste isn't RNG and has a 60% breaking point, but it also doesn't improve hpm except indirectly (if the 5SR still existed, you could argue that casting faster and standing still had value, but it doesn't anymore). Mastery improves efficiency, and in situations where you really need throughput badly it would beat out haste. On the other hand, faster casts does give you more casting flexibility and wiggle room for movement fights so I wouldn't say mastery is blindly better than haste.
Edit: Oh also, before people bring up stuff like Spark of Life or Purification, they wouldn't change the relative values between spells you cast (since they affect everything)
Edited by Faveokatro on 11/22/10 3:40 PM (PST)
A if you wanted a breakdown on the rating points themselves (instead of %) since it may be more relevant for gearing decisions: Crit and Mastery both take 179.28 points per % so nothing's different, but haste only requires 128.06 per %. The math to take that into account:
Haste: 1% / 128.06 = (1 - Health) * 2.5% / 179.28
Health = 44%
Crit, Single Target: .95% / 179.28 = (1 - Health) * 2.5% / 179.28
Health = 62%
Crit, Multi Target: .50% / 179.28 = (1 - Health) * 2.5% / 179.28
Health = 80%
So on a per point basis, Haste works out to be the dominant throughput stat for most situations which isn't a surprise really, since it always has been. The question then becomes how much haste your regen can sustain. Either way though mastery certainly has considerable value, and imo I would take it over the inconsistency and arguably lower throughput when needed of crit.
85 Orc Shaman
Edited by Azloz on 11/22/10 4:26 PM (PST)
I checked over Fave's post and his logic and numbers are accurate. Since mastery, crit, and haste are multiplicative their impact on each other is fairly minimal.
Edit: The above might happen in terms of ratings, since the % they yield diminishes are higher values.
Purification will not affect the calculations, since they are done relative to each other, not absolute. This is assuming healing bonuses stack multiplicatively...and they do.
Deep healing not being RNG is a legit selling point. Every tank already knows taking more damage constantly is better than less, but in huge spikes.
It's worth mentioning that if somebody is >80% and damage is low you don't care much about throughput, it's when they are low that it matters. Fave's post was accurate, but did not go deep enough to touch on healing habits and probability. The odds are, if you are using greater healing wave, your target is at low health(under the 62% breakpoint) already, or are about to be and if they aren't you're probably going to use healing wave or something where you already don't care much about the healing per time to care that much.
You simply don't heal people who are already at full health, so the deep healing bonus is almost always close to, or better than(chain heal) crit.
This is some great stuff.
The only point where this breaks down is tidal waves and healing surge, but healing surge is so expensive that I don't really see people using it-- and the additional crit would work in favor of mastery anyway.
So to summarize, assuming your mana regen is sufficient to complete the encounter,
Haste > Mastery > Crit
This applies to spamming CH into the melee, triage healing random DPS, and even hardcore tank healing. Have I read that correctly?
Edited by Faveokatro on 11/22/10 4:11 PM (PST)
Nah, diminishing returns only occurs for avoidance, not for throughput stats.
If you're referring to how stats can change in value depending on how much of another stat there is, that can actually happen. But it's a whole different analysis with a different goal (to maximize mathematical throughput). That's valid for dps, but for healing it takes a more holistic approach.
This is some great stuff.
And yeah, I considered Tidal Waves but it doesn't affect the breaking point and in terms of throughput it does give the advantage to mastery anyway as you said. I would lean away from saying Haste > Mastery > Crit definitively because certain fights favor some mechanics over others (Chimaeron would have Mastery > Haste > Crit for example), but as a general guideline yes.
60% over haste. Good enough for me. :D
I made a post a while back talking about stat priorities for cata, and it looks like my theorycrafting was right on the money after seeing your calculations. Actually here, i'll find it.
I'm unsure myself outside of how I _feel_ from my beta experiences.. My working theory is int will still be the stat of choice. Then you'll stack haste to the easiest reachable hot cap points. not sure what they are yet for 85, only 80. once you're at the haste hotcap point, you'll then stack either crit or mastery. I'll be going mastery. But it seems most everyone else will be going crit/spirit.
Looks like I was bang on. :D
Well, 44% if you're going by rating points over haste :P So, haste still has the advantage of being entirely consistent and of course, giving you more casts which is a benefit in itself (allows you to heal more targets in a shorter period of time). But at the very least mastery can compete with it and make it an argument.