Author’s note: I have quit playing World of Warcraft before release of patch 3.0.5. The content of this post is most likely out of date.
To follow up on my previous post, here’s another short series of comparisons in terms of item budget, using a more complex casting pattern.
Our fictional Theoradin (I think I need to trademark this one) is sporting this holy build:
Our good Theoradin, at level 70, will have a base 10% crit on Flash of Light, and 16% on Holy Light. He sports a whopping +33 healing from talents, but for the sake of this comparison we will again assume he has +1400 healing shown on his paper doll from other gear stats.
Instead of spamming Flash of Light 7, he will toss out emergency Holy Lights 11 conveniently every 14 seconds (so that he keeps Light’s Grace rolling, and the mobs are happy to oblige by avoiding any unforeseen spikes). In other words, a 14-seconds cycle starts off with a HL11, followed by 8 FL7s.
Without any further +crit% or MP5, and starting with a 10000 mana pool, our perfectly timing Theoradin will be able to keep this cycle up for about 64s before running mana dry, accounting for illumination returns. During this time, he will have provided an average of 73200 raw healing, accounting for crits.
Which, BTW, is a pretty lousy figure.
The calculations have been made so that when a 14-seconds cycle finishes, Theoradin will try to cast another HL11 followed by as many FL7s he can until he ends up mana dry. If he hasn’t enough mana for an HL11, he will cast several FL7s instead until dry (which will account for the non-linear nature of the results).
Please note this important caveat, though: For the purpose of keeping the calculations relatively simple (in other words simple enough for someone of limited intellect like me to manage), I made the following shortcut: for each spell, the mana cost I took into account in order to calculate the total casting duration was discounted by the statistical Illumination return. In other words, a FoL7 which normally costs 180 Mana has been calculated as costing only 170 mana with 10% base crit chance: 180 x 10% x 60% = 10.8, rounded down to 10. We are therefore firmly in the realm of statistical numbers, and the end of a casting series will warrant less remaining casts in reality for a simple reason: If Theoradin has, say, 178 mana left after his last full casting cycle, the spreadsheet will have granted him another FoL7 cast whereas in reality, he would have had to stop casting and worry about mana recovery (or LoH first followed by mana recovery, but I digress).
Again, I hope nobody takes these simulations for more than what they are: a purely mathematical simulation of +MP5 and +crit in order to give some insight about their respective efficiency in terms of item budgets. Reality will vastly differ, first and foremost because Theoradin has no lag, 0 latency and chains casts like a pro. In practice, less perfect execution, pauses in combat, lag and changing circumstances will give more ticks to MP5 to regenerate mana, thus extending casting time.
Without further ado, here are the graphs:
In terms of outlasting the challenge, 9.2 MP5 are always more interesting than 1% bonus crit, which is hardly surprising, all things considered. Let’s not dwell on this part any longer.
The healing ouput using this more complex casting scenario rests firmly on the side of +crit, contrary to the purely FoL7-spamming exercise in my previous post on this matter. Still, at the risk of restating the bleedingly obvious, a healer which is OoM produces 0 healing. The limitation of this kind of studies is its polarization – if I ever come up with a simulation combining both stats in a both realistic AND relevant manner, I’ll post it as well.
Last but not least, the summary chart displays the difference in favour of +crit in terms of raw healing output. As hinted above, the shape of the curve, in particular the drop at 184 MP5 / +20% crit is due to the way my little simlation handles the final and unfinished cycle.
As in my last post, all this little number crunching does is simply to provide generic pointers. Ngita, commenting on the previous study, made a couple of good points, one of them is that depending on your assignment in terms of raid groups (provided you play with more or less the same 4 other classes in your group), the relative worth of a percent of additional spell crit chance above MP5 can change a lot – if you have plenty of mana regen available from a shaman or a shadow priest, for instance, MP5 becomes less useful to you. If you’re healing without synergies, things look differently.
Last but not least, and outside the scope of this study, your base mana pool as well as +healing all contribute to fuel your healing output and your healing time. In the end, Theoradin can only do so much with number crunching. Skill, good timing, a good potting strategy, wise use of Divine Favour or Divine Illumination and class synergies leveraged by a smart raid leader will always turbocharge whatever stats your gear holds. Back in the days before TBC, my warlock in instance blues could out-DPS T1-clad locks simply with a better talent build and optimized spell rotations. The same holds true for healing today. A FoL held off half a second which results in 0 overheal instead of 30% will always be best for your team’s survival, and the experience factor makes a huge difference here – something a spreadsheet and a couple of charts will never be able to account for properly.