WotLK Healing Compendium v3.3.5: Fun in the Twilight with Halion - Page 5

Frmercury · 04/22/10, 10:19 AM

Originally Posted by Hegan

XI. Gearing Questions (outdated, changes incoming)

That section is still being reworked as noted above.

Haste (to cap) can be considered viable for Holy. More on that in the On the Value of Haste thread.

Sgat8516 · 04/22/10, 1:55 PM

Originally Posted by Hegen

Sorry to say this, but some of the numbers in the current guide are actually incorrect (not the basic formulas nor the reasoning).

I would like to see that as well. I've never really taken the data given in those formulas as anything other that at face value. Reworking the math on my own (using rawr, as I'm surely horribad at anything other than simple math accomplishable with a calculator :P), they do seem quite off.

Holy after Kings+5% spirit
1 intellect = 0.165 Mp5 (Replenishment)
1 intellect = 0.275 Mp5 (5min mana pool size)
1 intellect = 0.0962 Mp5 (5min Shadowfiend, assuming 7 of 10 hits)
1 intellect = 0.145 Mp5 (Meditation, 1200 Spirit assumed, generated from disc)
1 spirit = 0.445 (Meditation, 2000 intellect assumed, generated from disc)

Disc after Kings+6% spirit+15% Int
1 intellect = 0.1898 Mp5 (Replenishment)
1 intellect = 0.3163 Mp5 (5 minute fight, mana pool size)
1 intellect = 0.1107 Mp5 (5 minute fight, shadowfiend)
1 intellect = 0.1667 Mp5 (Meditation, 1200 spirit assumed)
1 spirit = 0.45 Mp5 (Meditation, 2000 intellect assumed)

Notes:
Used Rawr to value Spr/int values with +/- buffs, to figure, used as many options as possible to get as close as possible.

My math for meditation is probably off to a +/- 10-20% value, due to limited sampling size, but I think those values for Int/Spr are more accurate for the current content (Disc Priest without Int/Spr gemming).

Starfire · 04/22/10, 2:22 PM

Originally Posted by Sgat8516

I just reviewed several logs for this, and I think your making an association where one doesn't exist.
I checked 6-7 logs, everyone had 1-2 Disc Priests. Every Disc Priest had 9-13 Rapture proc's, but 100% of all those those proc's, also proc'd the revitalize-style secondary effect of Rapture (IE, the priest's were shielding themselves, and when those shields broke, they were getting the secondary and primary effect of rapture, but only on themselves).
Seems to be just a flaw in the secondary effect of the talent, rather than a 'on/off switch that blizz can flip'.

Blizzard did NOT intend that everytime Lich King 25HM cast infest, that the Disc Priest shielding the raid would get 50% of his mana bar back instantly... (2.5% per x 20 shields, and its almost everyone of those, depending on resist rates.. minimum is usually around 16-17..)

Intentions or not, it's highly unlikely Blizzard is going to nerf it at this point. Every single kill so far has used a Disc Priest, nerfing the intended-or-not mechanic would require Infest or the fight itself to be substantially tweaked. Either that, or we go back to stacking druids to throw Innervates at the priest.

Hegen · 04/22/10, 5:27 PM

While checking the regen stuff in the guide, I stumbled over reports stating that shadowfiend:

- profits from spell haste on gear
- does not profit from spell haste on gear, but from heroism/bloodlust if fiend is popped before herois is cast
- does not profit from spell haste on gear, but from heroism/bloodlust if fiend is popped after heroism is cast

I tried to verify this using a heroic target dummy (as it's based on swings, shouldn't really differ from real bosses). Testing with 0 haste, with 28% haste, with heroism before fiend launch, heroism after fiend launch showed no difference. In each case I got between 8 and 10 swings return 5% of max mana each. Heroism was applied to the fiend according to combat log. I've seen 10 swings with 0 haste and no heroism, and I've seen 8 swings with heroism active before fiend was launched.

Naturally, due to heroism debuff, my test series regarding heroism wasn't really large. That said, given the extremes I've seen, it doesn't look like any of the above is true in 3.3.3. Looks more like assuming 9 swings on average is realistic.

If someone has done a larger test series that contradicts this or combat logs that show differently, please speak up.

Maybe the haste effect on the fiend is a side effect from shadow form (like the haste benefit to DP and VT).

tedv · 04/23/10, 1:03 PM

My log analysis showed that the shadow fiend double dips on bloodlust. He copies your spell haste as attack haste, and can also get the bloodlust buff himself. So you'll see logs where a lusted shadow fiend deals around 70% more damage than an unlusted fiend (1.3 * 1.3 = 1.69). For this reason, it's my understanding that shadowfiend also benefits from haste rating on your gear. He gets your spell haste as melee haste, plus additional haste from whatever buffs are on him.

Tersa · 04/23/10, 1:13 PM

Originally Posted by Shylena

I have recently started experimenting with moving away from use of haste, serendipity and PoH to greater use of SP and Renew spam. I still value haste, but I have in experimenting started regemming so as to replace haste with more SP.

I will be interested in any experience others have with such a modified playstyle in ICC as a holy priest.

As my "main" spec is disc (I have holy and disc specs and switch between the two on a regular basis due to mechanics or raid comp), I use this exact healing style when I play holy. My gear is gemmed with pure spellpower and even my meta is more for throughput since I typically do not have mana concerns. Using a priority of PoM>CoH>Renew I have much better throughput than using PoH. I have even dropped serendipity from my holy spec as I really never do use PoH or GHeal as a raid healer. My decision, about the time blood queen was released, was that it was almost always a loss of ehps.

Sgat8516 · 04/23/10, 1:41 PM

Originally Posted by Tersa

PoM>CoH>Renew

The problem I have with Renew Holy Builds, is they tend to hit on the same problems as Resto Druid HoT's (IE: They have no discrimination on who they heal, heal just as many 99% hp players, as 1%). Sure its probably padding the effective healing by sniping more heals from other healers, but really, how many of those players didn't have incoming heals that removed the need for the HoT?
Personally, thats why I favor Disc Bubble spam, it doesn't just snipe effective healing, it prevents the dmg before its caused. No damage caused on 10 players = 10 players less to heal directly.

I stumbled over reports stating that shadowfiend

Doesn't the character info tab show a pet tab while the shadowfiend is out? Wouldn't that be a more reliable method of checking Shadowfiend attack rate?

Starfire · 04/23/10, 2:41 PM

Originally Posted by tedv

My log analysis showed that the shadow fiend double dips on bloodlust. He copies your spell haste as attack haste, and can also get the bloodlust buff himself. So you'll see logs where a lusted shadow fiend deals around 70% more damage than an unlusted fiend (1.3 * 1.3 = 1.69). For this reason, it's my understanding that shadowfiend also benefits from haste rating on your gear. He gets your spell haste as melee haste, plus additional haste from whatever buffs are on him.

Interesting. Any idea if he benefits from spell-crit? For what it's worth, he also benefits from AP, in terms of damage he deals, not in mana regen.

Hmm, I suggest we should do some tests to see if he benefits from the 20% Power Infusion as well. And heck, this might potentially be a new vector for Haste trinkets/pots. Guess we need to do some research.

Tersa · 04/23/10, 5:22 PM

Originally Posted by Sgat8516

They have no discrimination on who they heal, heal just as many 99% hp players, as 1%

Then, to put it bluntly, the priest is doing it wrong. Take blood queen for example. I do not have time to put renews on everyone in the raid group. Instead I do with renew much as I would with shields as disc - I try to predict damage. I make sure there is a renew on each tank, on the people with shadows or pact, and on anyone close enough to the center of the room (our meetup point for pact) to take damage from people running to drop it.

The main reason I still choose holy vs disc on this fight is the stronger PoM (yes, mostly overheal, but occasionally worth the extra spellpower from the swap) and the effectiveness of CoH at targeting low hp players. Yes, smart heals are boring, but they are good at what they do.

Edit: Re-reading your post it seems that you are thinking of ten mans. In a ten man situation I might indeed be more temped by bubble-spam, however in a 25 I believe holy edges slightly ahead.

Hegen · 04/23/10, 5:57 PM

Regarding the overall behaviour of shadowfiend, I've done more testing. This time in shadow spec, for the shorter cooldown. Note we don't know if everything works exactly the same for holy/disc.

The following list of results is proven in my opinion:

Shadowfiend damage is melee damage (but shadow school)
Shadowfiend damage per swing scales with spell power.
Shadowfiend swings can glance and crit
Shadowfiend swings can be blocked and partial resisted

Test series for spell power:

37sp, 0 haste: 10 swings, dmg 225 before resist
216sp, 0 haste: 10 swings, dmg 314 before resist
681sp, 0 haste: 10 swings, dmg 448 before resist
1197sp, 0 haste: 10 swings, dmg 710 before resist
2949sp, 933 haste: 10 swings, dmg 1275 before resist

Question: does a shadowfiend profit from temporary spellpower gains?

It does, and it refreshes per swing. Tested by popping an on-use trinket after the first shadowfiend swing. Second swing has far higher damage.

Question: does a shadowfiend profit from haste on gear and on use trinkets?

Test (run 3 times): 1017 base haste, popped trinket giving haste for a total of 42,45% (including base haste) for 20s, launched shadowfiend

Result: still only 10 swings!

Regarding haste: at least regarding heroic targets dummies and regarding haste on gear: I highly doubt this. 42% of haste is too much not to see a single swing more. Now this is not fully conclusive, but I think you can see why I have serious doubts regarding the haste effect in 3.3.3.

More current test data, please.

Update: have received several interesting logs from current patch, sometimes showing a gain of swings for heroism, sometimes not. More details when I had a closer look. This is a mysterious pet.

Originally Posted by Sgat8516

Doesn't the character info tab show a pet tab while the shadowfiend is out? Wouldn't that be a more reliable method of checking Shadowfiend attack rate?

Good idea, but I don't get a tab for the fiend.

Sgat8516 · 04/23/10, 10:24 PM

Originally Posted by Hegen

Regarding the overall behaviour of shadowfiend,
Good idea, but I don't get a tab for the fiend.

Hmm, I didn't get one either..
I know for a fact I get a pet tab for my Enhance Shaman when summoning Spirit Wolves... As I can actually check their AP/Ect on the fly... That is odd..

Cryptrobber · 04/23/10, 10:31 PM

I was looking at this forum and realized that the coefficients were out of date. Since I was recently wondering if CoH could possibly heal more then PoH with Tier Bonus and ICC buff increases, I realized now that it can't possibly (thats another story).

But I did do some testing that resulted in an interesting find (I am not a math whiz but i attempted).

How I tested for coefficients was by:
Casting CoH and PoH 15 times with no gear and not speced. I casted it single target (on self) and ignored critical heals (i will log out, and keep the current test spec i had). Also Casted 15 PoH/CoH w/ talents, but without T10 bonus

*Note* This was without Buffs
Results:
Without Gear/talents/buffs(0 spell power)

Avg PoH:2171.2
Avg CoH:1011.86

With Gear but W/o Talents/buffs(3200 spell power)

Avg PoH:3847.1333R
Avg CoH:2300.2

i used the formula (Avg. Heal/Spellpower=Coefficient)
PoH Coefficient: 0.524
CoH Coefficient: 0.403

Tested the numbers (w/ talents using the coefficient)(not including T10 bonus)
Avg CoH casted: 3113.3
Calculated CoH Avg: 3086.2
Avg PoH casted: 4928.8
Calculated PoH: 4912.2

Again I am not a math whiz, I was simply playing with numbers and seeing if it worked or not. I was wondering if someone could check my methodology that i used, and let me know if its correct.

Sgat8516 · 04/23/10, 11:05 PM

Originally Posted by Cryptrobber

I was wondering if someone could check my methodology that i used, and let me know if its correct.

Coefficients don't change from tier to tier (with exceptions of when blizz actually changes the coefficients).
The problem with your method, is that you aren't getting a very large sample size. A PW:S doesnt have a variable amount of healing (base), so you'll generate the exact same numbers everytime with the same spellpower. A direct heal will fluctuate 5%? off the base amount of the heal, and would skew your numbers off dramatically.

From what I've seen, Rawr is almost exactly accurate for use in making comparisons between abilities (the only variable it can't seem to account for, are values that aren't specific (warlock % spellpower buff as example)).

Originally Posted by Tersa

Edit: Re-reading your post it seems that you are thinking of ten mans. In a ten man situation I might indeed be more temped by bubble-spam, however in a 25 I believe holy edges slightly ahead.

(edit: for reference)
WoW Meters Online - Combatlog Replay
This as an example. Heavy Raid AoE dmg, with at any and random points during the first, players drop below 50% HP (HM, the pulsing aura scales upwards with vamp numbers).
Rejuvenation(R15) = 61.3% of total healing
Glyph of Rejuv = 0.4%

So, that HoT (since its the only comparable mechanic ingame, that has any sort of hp % modifier) only healed someone below 50% hp 1.3% of the time (0.4 x 2 (since its 50% of the heal)) / 61.3). ToF works similar to this, but the effect doesnt apply unless the HoT is cast on the target while under that threshold (and then ticks for that amount for entire duration).

Elimbras · 04/24/10, 6:50 AM

The same applies to shields with exactly the same reasoning. You prevent damage that would have been healed by JoL / other heals.

Basically, any heal you can't "chain" on the same person suffers this limit: if you don't have enough "low" people, you have to use it on people out of danger, or use another heal.
For holy priest, I use a priority kind of following : POM (low people who will be taking damage, or tanks) > COH (group of low people, or 2 danger of death people) > flash heal (people in danger of death, if possibile instant) > renew (low / mid life people) > renew (preventive on high health people).

If you use PW:S instead of renew, that's exactly the same your're doing (except that you can PW:S the near-death people, if they don't have weakened soul debuff). I fail to see any difference.

Now, one could argue to do all the raid healing with "direct heals that can be chained", ie. flash / great / binding / penance / POH. That is still useful in some cases of burst damage (like marks on Saurfang, Razorscale back to Ulduar, Rotface / Putricide for sick people / targets of adds). But if no one is low, and if you expect damage on most of the raid, then it's better to put as much preventive healing as you can (ie. renew / shields).

Sgat8516 · 04/24/10, 1:50 PM

Originally Posted by Elimbras

You prevent damage that would have been healed by JoL / other heals.

That illustrates the larger difference between the 2 abilties. No one would heal a full health target with a PW:S on them, they would move on to the next players down the raid frames that were at 80%/20% hp, where as a person with a HoT would get just as many heals at 80%, as likely the person at 20% (at least the guy at 20%, that gets a shield applied to him, isn't effectively at 20% hp anymore, more like around 40-50%).