Friday, October 5, 2007

Comparing Healing Gear (Part 4)

So now you have a measure of how much each of the various healing stats will improve your overall healing output. Great! But what about other stats like Stamina, spell haste, or Armor? How important are they? Should you assign weights to them as well? And what other refinements can (or should) you make?

It is relatively simple to calculate the percent improvement for the majority of stats. The difficult part, if you want to directly compare these numbers to your healing stats, is determining how much weight you should give each one. For PvP, for example, a 1% increase in survivability (Stamina, Armor and Resilience) is probably as important as a 1% increase in healing output. While raiding, however, only Stamina is likely to be of real use to you, and almost certainly not at a 1:1 ratio.

The formula to calculate the percent benefit for increasing a point in Stamina (i.e. health) is equal:

10 / Buffed Health * 100
Including Blessing of Kings, Fortitude (Improved = 71 * 130%), Mark/Gift of the Wild (14) and [Blackened Sporefish], this becomes:
11 / ((Base Stamina + 102.7 + 14 + 20) * 11 + 2979) * 100
For PvE, I consider a 1% increase in healing stats to be about as important as a 5% increase in Stamina, so when assigning weights I divide the result by 5. For my PvP calculations, I use a 1:1 ratio.

The improvement to physical damage mitigation for each point of Armor is equal to:
1 / (Current Armor + 10557.5) * 100
Armor is relatively unimportant for PvE, so I assign it a weight of 0. For PvP, the increase in survivabilty is about ½ of the Armor's mitigation increase (spell damage is not affected by Armor). I therefore divide the percent improvement by 2 when calculating weights.

It takes 39.4 points of Resilience to reduce opponents' crit rate by 1% and their crit damage by 2% (for about a 1.5% reduction in damage taken). Resilience is a flat-rate improvement (assuming you're not already capped), which results in:
1/39.4 * 1.5 = 0.038%
Resilience is another survivability stat that I consider unimportant for PvE, so I assign a weight of zero. For PvP, I give survivability improvements a 1:1 ratio to healing output.

Spell Haste
Although Spell Haste doesn't affect your total healing output (assuming the fight is long enough that your limiting factor is mana reserves and not simply having enough time), it does affect your healing per second (HPS). As of patch 2.3, 15.76 spell haste rating will provide a 1% increase in casting speed (and therefore HPS):
1/15.76 = 0.063%
How much weight to give spell haste is a personal choice. I consider haste effects to be about half as important as output for PvE, and equally as important for PvP.

Other Considerations
There are various tweaks and improvements you can make to these calculations based upon your own preferences. Some examples are:
  • PvP/Arena: I've calculated a different set of values for my arena gear because I'm not able to count on the majority of my raiding buffs or mana regeneration sources. I consider improvements in survivability equally as important as improvements in healing output and have shortened my fight length to 5 minutes. If you're interested, here are the comparitive weights:

    Relative "value" of PvP healing stats for Draezele
    Spell crit0.640.80
    Spell haste0.972.42
    Note: I reduce the weight of mana/5 in PvP by ½.

  • Non-Healing Spells: I use a weighted average of HW, LHW, CH and ES based on my usual spell mix, but there's no reason you have to stop there. For example, totems, Earth Shock and Purge easily make up 10-15% of my total spells cast during arena fights. If I wanted to create truly accurate evaluations, I could include these as "zero health healed" spells. In practice, however, including these numbers will not greatly affect your overall results.
  • Damage Spells: If you play a dual-role as a healer and damage-dealer within the same fight, you could also include the improvement provided by DPS-oriented stats (+damage, hit rating, crit rating, etc.). Unless you're playing a hybrid role in arenas, however, I suspect that you'll have fairly defined roles for each individual fight, making their inclusion unwarranted.
I hope that this analysis was of use to you. Obviously you'll need to adjust the formulae to your own personal gear and playstyle, but the basic premise is still the same. Likewise, this methodology can also be adapted for other classes as well.

Go back to Part One (Average Heal Amt.)
Go back to Part Two (Maximum Spell Casts)
Go back to Part Three (Stat Valuations)

Also see Comparing Healing Gear (Revisited) for further discussion.


Finwe said...

Great write-up. I love getting into the theory of it all. I am just about to start some of this on my druid. The weights for different stats is most interesting.

FYI, I think you hit a typo on the MP5 for your PVP gear. The note says it should be half of PVE, but it's actually listed as higher. 2.56 instead of 5.56 maybe?

Draezele said...

Because of the lack of mana regeneration sources available to me in PvE (mana tide, mana spring, consumables), Mp5 would actually have a ratio of 11.12 for PvP if given equal weight to the rest of the healing stats. However, I feel this significantly overvalues its usefulness, at least based on my experiences with my current team, and I therefore multiply its weight by 1/2.

Ralph said...

I read these pages because of a link from ( ). I've been doing gear evaluation spreadsheets of my own, so I was very glad to see the reasoning behind your evaluation, not just the final scores. Many thanks!

I ended up choosing a different basis for my evaluation. I started with the same 'how much healing can you do in X time' logic you demonstrated, but there were a few things I didn't like about that:
- consider one gear set that lets you do 1000 HPS for 1 minute, and another that lets you do 500 HPS for 2 minutes (but you still go OOM before the end of your fight). The "total healing" metric would value both sets equally, but I think the first gear set is clearly better--the first gear set lets you do burst healing much better, and if 500 HPS over 2 minutes is what you really need, you can achieve that with the first set just by casting less often.
- suppose you have enough mana and mana regen that it takes you 9 minutes to run out of mana. Adding more mana regen doesn't increase your "healing in 7 1/2 minutes" value at all, but it clearly has some value, because very long fights do occur some times.

In order to deal with these issues, I ended up evaluating my healing as a weighted average of the healing I could do over time, with the weights decreasing exponentially out to infinity. The down side, of course, is that the math got much more complicated.

Draezele said...

Excellent point there, Ralph. I considered a more involved way of modelling my healing power (burst healing versus longevity), but ultimately I felt that it was needlessly complex, especially since the relative importance of each is heavily influenced by the other healers in the raid. Obviously a "one size fits all" mentality is inadequate for every fight, so I'll often carry a couple of alternate gear items for different situations. But in general, I find that a balanced approach to my healing stats provides the best overall utility.