As a Restoration Druid, there are several stats that must be considered when determining whether a given item is an upgrade: Intellect, Spirit, MP5, and +Healing. Each of these should be weighed relative to one of the others to get a good idea of how a given item "breaks down." MP5 is my favorite indexing stat since the other stats have values greater than 1 when making a comparison. Today I'd like to examine how much +Healing equates to MP5. But unlike the calculation for Spirit vs. MP5 (in which it is relatively simple to account for time spent within the 5SR), the value of +Healing is a much more complex calculation that must take into account the following variables:
- The percent of total healing comprised by each of our various Healing spells
- The total health restored and mana cost for each of the above spells
- The scaling of each of the above spells for each point of +Healing (article here)
From the above data, we can calculate the HPM for each spell (health restored per point of mana) and then weight this HPM by the percentage of total healing provided by each spell. Obviously, the way in which we heal (and thus the percent of our total healing that comes from each spell) is subject to the dynamics of a particular encounter. I will be using data from last night's Gruul kill to demonstrate how this can be calculated. I am using the excellent UI addon, Recount, to breakdown my healing by spell:
Breakdown of Total Healing by Spell
(Phaelia vs. Gruul the Dragonkiller)
As you can see above, Lifebloom was the source for 71% of my total healing. Unfortunately, Recount doesn't track the difference between a triple stack of Lifebloom and a single application. This is important, however, as the strength of 3x Lifebloom is its high HPM efficiency. To determine that, I used additional data from our guild's cumulative raid stats that indicates that 47% of my total healing was to our main tank (the most likely target for 3x Lifebloom):
% Total Healing 3 x LB = 0.71 * 0.47 = 33.37%
% Total Healing 1 x LB = 0.71 * 0.53 = 37.63%
Note: I recognize that this estimate does not account for the other forms of healing I applied to the tank, but this is an unfortunate limitation of the data I was able to mine.
Using these values and assuming an average of 1500 +Healing, I was able to generate the following chart (download the spreadsheet):
From the values above, I can conclude/calculate the following:
- On average, 1 point of mana converts to 11.74 points of health.
- Over the course of an 8 minute fight (estimated time for Gruul), 1 MP5 provides 96 mana or 1127 health (8 * 12 = 96; 96 * 11.74 = 1127).
- 1 MP5 converts a 0.46% boost to Healing (1127/244826 = 1127 / total healing).
- 1 point of +Healing represents a 0.0418% boost to healing.
- If 1 point of +Healing is valued at 0.0418% boost to healing and 1 MP5 is valued at a 0.46% boost to healing, then 1 point of MP5 is equivalent to 11.00 +Healing (0.0046/0.000418).
I recognize that I had to make a lot of assumptions to find this value and that there are a lot of variables that might change based on the encounter. I've provided the spreadsheet for download below, so feel free to adjust the figures for your own collected values. Note that there may be reason to further increase the relative value of MP5; after all, if at the end of the fight the tank dies because you are out of mana, having healed for a bit extra at the start will be of little comfort.
Note: This calculation ended up feeling less "solid" than calculations I've been able to perform previously, mostly because I lacked good data for the number of stacks of Lifebloom that were applied. Nonetheless, after plugging in different values for the "soft" figures that I've estimated (% of 3x Lifebloom vs. single Lifebloom), the numbers do not seem to change significantly.
Download the Spreadsheet
10 comments:
I like your number crunching very much. The calculation for spirit vs. MP5 was well done and your +Healing discussion was very, very insightful.
This one is the most difficult so far. I understand that data mining limits the determination of hard figures in this case. Your conclusions are pretty close to my experiences. My rule of thumb in this case was 1 MP5 equals 10 +Healing. But, however, this rule was completly "guessed" in lack of any good data. Perhaps you can imagine my astonishment as I read your last conclusion. :-)
By the way - there was I slight mistake in that last conclusion. Shouldn't it be:
"... then 1 point of MP5 is equivalent to 11.00 +Healing (0.0046/0.000418)."?
Your version goes:
"... then 1 point of +Healing is equivalent to 11.00 +Healing (0.0046/0.000418)."
I really, really appreciate your blog and check it at least two times a day. Many thanks for the tons of helpful information so far. Keep up the good work! :-)
Cheers,
Nebelmond, EU-Die silberne Hand
Oh, forgot one more thing...
I remember reading the official WoW forum the other day and some complaining about some buffs that overwrite others (like the armor bonus of MotW is overwritten by the paladins Devotion Aura).
Do you know if this applies for MP5 buffs, too? For example like the sporefish MP5 bonus being overwritten by the paladins Blessing of Wisdom?
So long,
Nebelmond.
I think I'll try out that Recount. I've always been partial to SW Stats, but Pie Charts are teh <3
We'll see how it stands up from a tanking point of view =)
@Nebelmond:
Consumable buffs are calculated separately from casted buffs - so BoW and sporefish do indeed stack. This was not always the case as Elixir of the Sages used to be overwritten by Arcane Brilliance/Intellect. Presently, however, you can have all manner of party/casted buffs and still gain more from consumables. The change was made quite awhile ago so I do not have a citation, but if you check your character sheet between having BoW and then having sporefish, you will see the increased MP5 listed.
@Phaelia
I very much liked the breakdown of how much +heal affects each of the different spells. I often hear about a rumored "soft cap" to healing, at which point (i.e., 2k +heal), its affects are diminished and stacking it further is pointless. I haven't ever read anything conclusive or official on this though. Do you know anything? As much as I may like math, I just don't have time to test such things. I just know that if I double trinket and get up to say 2400 +heal, my Lifeblooms tick for more and I am happy - whether this is a proportionate increase or not though I am unsure.
Ada
*whistles* All that math flew right over my head. So Spirit does absolutely jack for you guys right (with exception of innervate)?
@Nebelmond: Thanks for the correction. Who would have guessed that 1 point of +Healing isn't, in fact, equal in value to 11 points of +Healing? LOL. I'm really glad that you are enjoying reading. I hope I can continue to hold your interest.
Adarel does an excellent job of responding to your question regarding food buffs. To give a little more information, the rules for stat buffs are:
- Pure class stat buffs stack with mixed class stat buffs. Mixed class stat buffs stack with each other. This means that you can simultaneously have Stamina from Prayer of Fortitude as well as Stamina from Mark of the Wild and Blessing of Kings.
- Pure class stat buffs (such as Fortitude and Intellect) overwrite the effects of scrolls. Scrolls do, however, stack with the stat bonuses afforded by Mark of the Wild, and I assume Blessing of Kings. Scrolls are most useful when soloing or grouping without one of the classes who can cast the buff in question. This, of course, means that I have about 12 stacks of scrolls on alts. =)
- Food stacks with everything, but you can only have one food buff. This includes Rumsey Rum Black Label.
- You can have one Guardian and one Battle Elixir or one Flask and the effects stack with everything.
- The resists from Mark of the Wild don't stack with anyone else's resists. /cry
@Kai: I think you will LOVE Recount. I couldn't believe how much information it gave to me. And in such a readable format. I think my favorite stat that I haven't seen elsewhere was the ability to view my % HoT uptime (i.e., how much of the time I had a HoT active on someone). As a Restoration Druid, this is just awesome. Finally, a way to measure and improve my ability to maintain a rolling Lifebloom! I may have to showcase Recount as a featured addon soon.
@Adarel: You are correct that +Healing suffers Diminishing Returns. It's easiest to think of it in these terms: assuming a Heal spell that heals for 1 HP. Further assume that it has a 100% healing coefficient (so that each point of +Healing applied to it would grant 1 point of additional healing). The first point of +Healing applied to it would grant 100% additional effectiveness. The second point would grant an additional 50% and the third would 33% additional effectivness. Obviously, the numbers are more complicated than this "ideal" example, but you get the basic idea. I cannot tell you at what point you will begin to see significantly diminished returns, but it sounds like something would be fairly easy to extrapolate from the +Healing spreadsheet I put together. Sounds like a potential blog post. Oh, and it sounds like you have way too much +Healing. You should definitely send some of that my way! ;-)
@Matticus: Actually, Spirit is arguably better for Druids than it is for Priests. You might check out the article on Mystery No More: Spirit and MP5 for an explanation of what it does for us. I'd personally rather have more MP5 than Spirit, but it's a stat that shouldn't be discounted!
Wonderful break down, if I ever level my fat cow I think I know where to look for details.
The ratio of +heal to mana/5 shouldn't be static as it will depend on how much +heal and mana regen you have. As you approach your mana regen cap (the point where you won't run out of mana in a given period of time even when chain casting), +heal becomes more useful because it's applied across all of your heals during that time (for example, 9 minutes of +1 healed every spell adds up). An extra point of mana/5 would give you very little benefit at that point as you're already able to chain cast for so long. Conversely, if you have very little mana regen, +heal is much less useful to you if you have no mana after 30 seconds. Essentially, both have diminishing returns and both reach those diminishing returns faster if the other statistic is significantly lower.
Overall, the best answer is that you should try to strike a balance. Overload on mana/5 and you may only need 5 or 6 +heal to improve your total healing power to the same extent as 1 mana/5 would improve it (for my current gear, my ratio is 5.56 mp5 to 1 +heal, for example). Stack +heal exclusively, you may need 15 or 16 to equal the benefits of an additional 1 mana/5. (I'll also note that +heal is MUCH more powerful for shaman thanks to chain heal essentially having a coefficient of 129% as long as you average 2 players healed each casting.)
And the shorter answer: If your tank is dying because you ran out of mana, stack more mana/5 (or Spirit). If your tank is dying and you still have mana, stack more +heal. And if your tank ain't dead, stack a little of both!
Thanks for the comments, Draezele.
You are correct. As stated in the article, your personal value of MP5 vs. +Healing is likely to fluctuate based upon the encounter and how you heal as a result. Obviously, if you never run out of mana, you probably don't place a high premium on MP5. However, for many players (myself included), mana management is an integral part of any boss encounter lasting more than 4 minutes or so (re: most boss encounters in Karazhan and Gruul's Lair). As long as I need to use Super Mana Potions in addition to my own Innervate, I will continue to prioritize MP5, sometimes at the expense of +Healing. Naturally, you should never completely ignore one stat in favor of another (no, not even Stamina). This article was intended to demonstrate how the value of +Healing should be calculated for a long fight like Gruul (at least for now). Under such conditions, MP5 is much more valuable. Because our guild never wipes due to the fact that I lack +Healing but does on occasion because I and other healers have depleted our mana reserves, I will work to improve my gear for such fights over those "easier" fights where more +Healing would be nice but nonessential.
A couple of other addendums. I assume that you meant to say that you currently value 1 MP5 as equal to 5.56 Healing instead of the other way around. And a triple Lifebloom stack from a Druid has an even higher healing coefficient (1.86) than that of Chain Heal. At a conservative estimated 33.4% of my total healing, I still derive a much larger value for MP5 than 5.56. Maybe there is some caveat of your calculation that I haven't considered.
Thanks for the reply, Phaelia. I look forward to the post. :D I also wonder if this changed somewhat from pre-BC? Back then, 1k +heal was a high point and many healers thought it useless to stack beyond there. I can only imagine we'll be up to at least 3k by level 80.. maybe it is just too early in the morning for me and I'm likely missing something simple that makes this all work.
And I don't usually run with >2400 +heal haha. I have Essence of the Martyr though and the druid trinket from SSC which gives an even bigger boost to +heal when it procs, together for a sum of about 650 additional +heal for 15 secs. Of course, if Lifebloom is first applied while those trinkets are active, as long as it is rolled without allowing it to bloom, it will retain that amount of +heal for the duration. So I guess if I am rolling perfectly, I do work under those conditions... but fights do not always lend themselves to such order.
Anyway, the point is, with 2500 heal or however much it is, a 3-Lifebloom stack can tick for over 900, while a normal (untrinketed) stack for me ticks for around 710 (with around 1850 +heal). Rejuv ticks for about 1080 as opposed to a bit under 900. That much of an increase (3x and rejuv on two tanks, ~1600 extra healing every 3 seconds) seems to diminish the importance of the 'cap' but I'm usually the only one thinking that. ;) I do hope you research the topic - I'd love to see your calucations.
-Ada
Post a Comment