Mana Regeneration

This has been a booger of a post to write – please pardon any disjointedness as I paste cuts from multiple drafts. See, I think this is actually going to be fairly important, but I’m not quite grasping it my own self.

Here’s the bottom line: How much mana regen do you need – mp5, spirit, whatever? While there’s not a cap, I’m coming to believe there’s a point where there are better things to build. And I’m going to help you figure out YOUR need.

Now, I drafted up a model for this. Used standard values and everything, and it looked great. Except… it didn’t jibe with reality. Which means something in my model is wrong. It might be my model. Or maybe some of the standard values are mistaken. As you can guess, I’m going to get longwinded.

Let’s start with the standard model – and my basic rule for “how much is enough”. Mana pool divided by NET mana burn should last a minimum duration. That last – minimum duration – is up to me. I used to be fixated on 10 minutes, but am coming to believe 7 is sufficient for my purposes. Maybe 5 – we’ll see. Anyway, Net mana burn is mana burn (all spells cast divided by time of casting for a mana per second rate) adjusted by mana gain (mana regen plus mana restoration – and don’t forget to factor ‘free’ casts).

Now, I can tell you what I normally cast over a period (50% renew, 30% PoM, 15% GH, 5% PoH) while in Kara. And I can tell you the derived mps of these based on my level, gear and talents (335 mps).

I can also tell you my typical regen values. My 450 OFSR/150 IFSR, at a nominal 70% IFSR, gives me about 48 mps. Add in various trinkets and clearfocus generating talents (Holy Conc, IF), and I have another 6 or so mps – make it easy math and say I’ve got about 55 mps regen. (Note I’m intentionally excluding shaman and s-priest mana regen – I’ve got the numbers for that too, but it’s an unecessary complication.)

My buffed mana pool is close enough to 10,000 that we can use that for basis.

Nominally, I can use my IFSR as ‘time casting’ – that is, burning mps. I’m going to tell you up front that this is the flaw, but will run with it for a moment. 70% of 335 rounds off to 235. 235 (burn) – 55 (regen) is 180. It will take me 55 seconds to burn through my mana pool. At some point in there I will use a potion – 2400 mana – and my fiend – 3600 mana. I get another 33 seconds till I’m OOM. So by the model, I should be OOM even using my basic regen tools by the time I’ve healed for 88 seconds. Remember that cooldown on pots is 120 seconds, and on the fiend it’s 180 seconds, so I can’t even use another of those.

But – I’ve never run out of mana in any fight of less than 7 minutes. Not while at 70 IFSR, anyway. Which means SOMETHING is wrong. It might be my tools (yes, more than one) that are measuring how fast my dance is. Or, it’s a flaw in the model. As I already said, I suspect the latter.

Here’s the big kicker. OFSR kicks in after 5 seconds of no mana burn. And equally critical, we have the fact that if I go three seconds of no mana burn my burn rate is lower but the OFSR is still nothing.

Basically, the concept is right. But any tool that tries to determine your burn based on IFSR/OFSR for out-go of mana is going to be… flawed. There is a solution, but you’re not going to like it.

Not least, there’s the fact that it’s a bit of a chicken-and-egg situation. See, what you do is keep WWS reports for lots and lots and lots of battles. And from these battles, you pull a short list of values.

How much total time was spent in all the battles. How many of each spell did you cast.

That’s it – to begin. Now, for each of the spells you cast you need to note cast time and multiply by the number of casts. Sum these times, and determine what percentage of the total battle time you spent casting. Oh – while you’re at it you can get your average burn rate too. Determine the mana cost of each of the spells you cast (ignore clearcasts and such for now), multiply by number of casts, sum the mana and divide by time casting. TaDaaa, your nominal burn rate is at hand. Except you can’t use that. You need your combat burn rate, which is the total mana divided by total COMBAT time. Or you can multiply nominal burn rate by the cast time percentage. Either way works and should come out about right.

It turns out that my combat cast time – over several dozen boss battles – is in the vicinity of 40%. In other words, my combat burn rate isn’t 235 but ~135. 135 (burn) – 55 (regen) = 80 net combat mana burn rate. And THAT takes me 125 seconds to burn. My fiend gives me ~45 more seconds, and my pots give me ~30 seconds. Assuming I pop each at the last minute I still get to 3 minutes before I’m OOM. And if I’m smart….

Pop a mana pot at about the 30 second point and I’m back to 10K – a new 125 seconds. Wait 45 more seconds and pop the fiend and again I’m back to 10K (and am 75 seconds into the fight). In 75 more seconds I can pop the next pot – giving me ~85 more seconds to OOM or just under 4 minutes total combat time. And… only 20 seconds short of being able to use my fiend to get yet another 45 seconds.

This is a lot – a LOT – closer to my experience. Close enough, in fact, to suspect what’s wrong is I’m slightly underestimating the effective mana regen of various things like Holy Concentration. But this is for me, specifically.

For MY casting style, if I can get my net combat burn down to 70, and I’m smart and careful with my pots and fiend, then I can go over 6 minutes. And if it gets down to 60… I can go longer than any battle in which I’ve participated.

10 mps is 50 mp5. Of which 18 is coming Tuesday when meditation 3/3 goes from 15% to 30%.

So for me and my current style… I am very close to being able to go for 10 minutes “non-stop”.

Let’s review, shall we? So you can rough out your own numbers.

1 – use a stat program – I recommend WWS – to determine how long you’re in combat and what spells you’re casting. Remember the more combats you add together, the more likely you are to be ‘accurate’.

2 – using your mana time costs, determine how much mana and cast time you use in combat.

3 – using THESE, calculate your combat mana burn rate (in mana per second) and in combat rate.

4 – Now, determine your IFSR/OFSR mana regen proportions (your ‘dance time’), and multiply those against the respective rates for your base combat regen rate.

5 – To this, add estimated (or calculated) regens from talents and trinkets. If in doubt, it appears that a wisely used IF plus typical Holy Concentration are worth ~40 mp5.

6 – change the regen to mps, and subtract this from the combat burn rate to get your NET combat burn rate (NCBR).

7 – divide your mana pool by your NCBR to determine the number of seconds you can expect to last. Do the same with your mana potion and shadowfiend numbers for how many seconds they can add at appropriate intervals. Plot their use (use the instant they’ll restore to 100%, and used again as soon as cooldown is up AND they’ll not have any wasted mana – at least nominally) to determine what your total combat endurance is.

Decide how long you want to last, and then determine what you need to get there. Note there are three variables to be adjusted:

  • Increase the mana pool;
  • Decrease the combat burn (reduce frequency of casts or mana costs or time spent casting)
  • Increase mana regeneration.

Here’s the real bottom line. There is a point – where is very dependent on YOUR style of play – where more mana regen is “nice” but where you might be better off looking for other places to put it.

Of coure, you could also add more mana regen and increase your combat burn rate. It’s just… now you can at least begin thinking of how to plan and do.

Regardless, this will hopefully increase your FUN. Go, play, have fun. And if you think of a way to take this raw beginning and make it MORE useful… don’t forget to tell me so I can have more fun too.

~ by Kirk on November 9, 2007.

5 Responses to “Mana Regeneration”

  1. I think I would go about it by breaking it down a little differently. There are the easy things to know, such as how long your fights are, what you mana regen is, etc. Then there are the tougher things to know, such as your typical burn rate.

    I would approach it by first creating a model for your base “longevity”. The stickied post on the WoW-US priest forums already gives a way to link spirit, mp5, and intellect by only knowing the total time of the fight and average time in the FSR (and there are addons out there to track both of these). From there it should be straightforward, given the battle time, to factor in shadowfiend, potions, etc.

    The second part would be to create a comparison model for your longevity vs. your burn rate. If burn rate is greater than longevity, then increase longevity (or get more healers in the raid!).

    Calculating burn rate is going to be the trickiest thing, as there are many situational variables to take into account. Tools like Recount and WWS can track how much you are casting of what, but the variance between certain battles (Maiden vs. Curator), raids (Kara vs. Gruul’s), or even nights (2 healers in Kara vs. 3).

    The whole concept of measuring burn rate just begs for a spreadsheet. Plug in your stats, talents, battle time, avg time in-FSR, what ranks of spells your are using, and what percent of each spell you use, and VIOLA! Burn rate. A spreadsheet would make it easy to quickly switch around a few variables to account for the situational differences mentioned above.

    There is a similar spreadsheet like this out already, but for shadowpriests. It can tell you how long you can last based on what you are casting, what your gear is, and what your talents are. Calculating it for healing may be a bit trickier, but I believe it can be done.

    Sorry for the long reply. Good luck on your epic quest. 🙂 I have been working on a much simpler spreadsheet myself and haven’t gotten very far.

  2. Kirk ,

    Thanksfor the detailed explanation.

    I see exact the same (a spirit based build is *very* raid liable) my priest currently working to Curator (include) and I never have a problem with OOM.

    However, what I learned is, that I need to adjust my playstyle (and I now know why this is necessary 🙂 ). But even on curator I go out of the battle with 20% manapool (utilizing mana pots and the fiend during evocation). And this as the “keep the raid alive healer” – it is working very well.

    And this is before the insane buff of meditation…

  3. Something I think you may be forgetting is the 5 second rule. When you are healing in a raid or instance you are not constantly casting a heal or any other spell for that matter and can sometimes go for as long as 10-15 seconds before you cast another spell. That will cause you to mana regen at the non casting rate if only for a couple of seconds which can be a huge bonus.

    I know that when I was a healing priest that is a method that I would practice to get the absolute most benefit to mana regeneration.

  4. @MK, re spreadsheet – look at the links to the side. That said, the problem for a healer spreadsheet trying to determine burn rate is a constant problem for healstats – we do quality, not quantity. In the end, yes, the solution is to track what you do, look for generalities, and determine the frequency of any exceptions so as to determine if you need to adjust your planning.

    @lehmlove. Thank you, and good job. But I can’t help commenting on your comment. Allow me, please, a general point of comparison – when I’m raidhealer in a 2-heal (other healer gets the tanks), I have at least 50% mana at the end of Curator, usually more. In a lot of ways, I think he’s an ideal test of your mana management skills BECAUSE OF his evocation. That’s 25-30 seconds during which you should be NOT casting heals, and you’ve kicked it off with a shadowfiend. If you’re using a raid notification program (deadly boss mobs, bigwigs) you’ll be able to cast the fiend 2-3 seconds before he evocates, getting a touch MORE mana. And… you probably shouldn’t heal anybody during the last add kill either unless they’re under 50%. Bandage, sure. Spell, no. Just hit anybody still in need a couple of seconds before the evocation ends.

    @Mojo. No, I’ve got the 5SR in there. In fact the ‘first pass’ that doesn’t correctly show my mana burn rate is based on my time in and out of the 5SR.

  5. Kirk,

    no problem. Please keep in mind, that my gear is not full epic and the only epics are the craftables and the rest is filled up with blues.

    This will lead to 9K mana (buffed, unfortunately). Yes, this is poor but I try to compensate this with high regen rate which works for me.

    I foresee, that I can focus more on heal/+int in the feature when I have more room for that, currently this will not work because my +heal is only at 1390 and I’m already used gemslots for that.

    So, respeccing to bump my manapool will result in loss of +heal and regen. But if I lose +heal my lower rank heals will become ineffective and I’m not able to conserve mana effciently. For my current gear massive focus on regen does work astonishly good because I can heal long enough and I do not need a greater manapool. I hope this does not change until I’m able to catch up gearwise.

    The curator encounter is very exiting and your tips are coming from someing who mastered that some time ago 🙂 Thanks for the advising, I will try to get better.

Leave a reply to MK Cancel reply