WotLK Healing Compendium v3.3.5: Fun in the Twilight with Halion - Page 9

Sargeth · 05/28/10, 11:35 AM

On the Tier 10 pieces
-snip-
I recognize that the 4 pieces is sexy as all get up for priests. However, I’m interested to see what people have in mind as their rotating/non tiered piece.
-snip-

When looking for non-set pieces, you can basically just go by ilvl. I have [Meteor Chaser's Raiment], since it's easy to get, and I don't really care about spirit these days. If you do, [Thaumaturge's Crackling Cowl] seems fine.

This is, of course, assuming you wont be seeing ICC25H anytime soon.

Elimbras · 05/28/10, 4:10 PM

I like empowered healing, though I spam pretty heavily renew, especially on the numerous aura fights of ICC.
But there are some fights where I feel renew just can't do it. These are the fights where damage is concentrated on few people (Saurfang is likely the best example). Then there is no point in blanketing the raid with renews, and I need some immediate healing tool, that is called flash heal.

Sgat8516 · 05/30/10, 7:53 PM

Originally Posted by Hegen

No. RecountGuessedAbsorbs - like WoL - adds up actual absorbs. For each individual absorb, it tries to find out which spell caused it. So the value from "Guessed Absorbs" is the same as effective heal.

Yea, the only deviation from the actual amounts of absorbs, are when there are multiple overlapping effects, and the most recent effect expired before being used. (Like Glyph of PW:S critting and proc'ing Divine Aegis, if Divine Aegis expires before PW:S does, WoL/Guessedabsorbs will no longer track that absorb value, WMO still does though).
Similar effect to when you're running multiple Val'anyr's in the raid, since that buff is a shorter duration. When the Protection buff expires if not used, any PW:S cast on the target before the proc bubble went up, will count as 0.

WTB combat log tracking for Absorbs >.<

metapseudo · 06/07/10, 12:03 PM

On the subject of best in slot gear and gearing in general, it seems that all items from ICC show unified pattern:

Sta+Int+SP = const for same item group, unless modified by gem slots
mp5 cost is 2x cost of spi, crit and haste. The three have same cost
Sta+Int+SP+2 stats of (spi, crit, haste, mp5) are constant unless modified by gem slots
gem slots are calculated for blue gems so it is alway better to have item of same type with more slots

Based on those 5 stats, items have these relative stat value:
Chest, head, legs = 1
shoulders, hand, feet, belt = 0.74
finger,back, neck, wrist = 0.56
wand = 0.31
2H = 2
1H+OH = 2.03 (but always more SP and less other stats then 2H)

It basically means that in SP+Int all items of same level are equal (even across armor type boundary)
only significant difference is selection of 2 from 4 optional stats (spi, crit, haste, mp5)

Typical iLvl 264 priest items have these values of the stats:

Slot	Sta 	Int	Spi	SP	haste	crit	mp5	armor	Y	R	B	M	Value
Head	123	123	92	162	0	100	0	287	0	0	1	1	591
Chest	123	123	84	162	0	108	0	354	1	1	1	0	587
Legs	123	123	108	171	92	0	0	309	1	0	1	0	585


shdrs	92	92	80	132	0	72	0	265	0	0	1	0	435
hand	92	92	80	132	72	0	0	221	0	0	1	0	435
feet	92	92	80	122	0	64	0	243	2	0	0	0	435
belt	92	92	64	122	80	0	0	199	0	1	1	0	436

finger	69	69	0	97	52	60	0	0	0	0	1	0	327
back	69	69	0	97	52	60	0	177	1	0	0	0	327
neck	69	69	60	97	52	0	0	0	0	0	1	0	327
wrist	69	69	58	97	0	53	0	155	0	0	1	0	326

2H	156	156	121	741	121	0	0	0	1	1	1	0	1190

wand	39	39	0	60	34	34	0	0	0	0	0	0	185

Remark: I am not sure of slot value for meta gem. Estimated value I have used was 38.

Hegen · 06/10/10, 10:02 AM

So it's done. I have rewritten the section regarding the value of int and spirit while recalculating all of the numbers. As I recently announced, the numbers differ from previous versions in at least one surprising way.

I will just quote the end results here. Please have a look at the guide for the full details, feedback regarding the numbers and the assumptions highly desired:

Originally Posted by Hegen

Ability	Stat	MP5 Holy	Full gains	MP5 Disc	Full gains
Total	Int	0.7321	0.7890	0.7574	0.9581
Total	Spi	0.5905	0.6356	0.3927	0.4579

Total for Humans	Spi	0.5905	0.6546	0.3927	0.4716

End results (for ICC gearing levels):

For holy priests, spirit may be preferable to int even when gearing for regeneration
For disc priests, spirit closes the gap to mp5 around ilvl 264, then surpasses mp5

Both results - which really differ from the previous guide versions - are due to changes in the game as well as far higher gearing levels. The result for holy, however, is also due to HC now properly factored in.

Hegen · 06/14/10, 4:53 AM

The section "XI. b) Gemming" in the compendium is now rewritten and contains a detailed stats breakdown based on the theorycrafting results from "XII. c) Value of Intellect and Spirit" - instead of just listing the gem names.

MP5 contribution in a raid buffed scenario is now directly comparable between gems, throughput is still divided into sp/crit/haste, but also including raid buffs and talents.

If somebody is interested to do a ballpark HPS conversion for holy and disc for sp/crit/haste, I will add that.

Elimbras · 06/14/10, 7:32 PM

Originally Posted by Hegen

So it's done. I have rewritten the section regarding the value of int and spirit while recalculating all of the numbers. [...]

End results (for ICC gearing levels):

For holy priests, spirit may be preferable to int even when gearing for regeneration
For disc priests, spirit closes the gap to mp5 around ilvl 264, then surpasses mp5

I don't understand your Full Gains column. However, if I'm correct in that it corresponds in "best raid buff scenario", then some lines are not correct.

For the holy table, you have correctly 0.15 * 1.1 = 0.165 and 0.25 * 1.1 = 0.275 for the resplenishment and mana pool lines. But for example, you have 0.1250 / 0.1235 ~ 1.01 (should be sqrt(1.1)~1.048 for the Meditation line, and 0.0743 / 0.0725 ~ 1.01 (should be sqrt(1.1)) for the HC line. I guess that the error is in the value of sqrt(1.1).

Worst, I don't understand spirit lines. Spirit effect is always linear, as far as I recall. But your lines ratio differs, from 1.08 to 1.15 about.

Last point, about the Shadowfied line (for holy at least):
Considering 1 Shadowfiend per fight, with 11 attacks at 5% mana each, that's 55% of your mana that is returned.
Increasing Int of 1 yields 15 mana. 55% of 15 is 8.25 mana per fight. Fight is 300s long, so that's 0.1375 MP5 per Int, non 0.1146.

Other lines seems correct for me in the holy part, at least for unbuffed priest without spirit talents.

Hegen · 06/15/10, 4:17 AM

Thanks a lot for checking - I'd really like to have values that we can apply some degree of trust to. Verifying the end result with simulation to see if the ballpark is right isn't the same as verifying each single value.

Originally Posted by Elimbras

I don't understand your Full Gains column. However, if I'm correct in that it corresponds in "best raid buff scenario"

Some explanations first - seems I should elaborate a bit in that section:

The disc and holy columns basically differ in the assumptions about the time spent OO5SR and the int and spi baselines.

The "full gains" columns for each include effects from talents (these are listed) and raid buffs that scale int and/or spirit. Usually these are applied first, taking the input value of 1 int or spirit, then scaling that to the amount of int or spirit after buffs and talents.

Meditation/spirit stuff: yes, a nasty error has crept in: in some of the cells for holy (not disc), OO5SR percentage was applied again in the full gains column. That's why I usually don't like doing this stuff in Excel - it's easy to miss mistakes that only show up when looking beyond the scope of a single formula.

I have corrected this in my sheet, but will do more checks for this kind of thing before updating the compendium later. Looks like the end result will be that spirit and int are even closer for holy.

Shadowfiend: my excel sheet had the old fight length hard coded, which was 6 minutes. It's now fixed, including the new totals. For now, I maintain my conclusion regarding int vs. spi for holy.

Update:
Checked the rest of the sheet: the mentioned error had an effect on the "full gains" cells for meditation (holy only), HC, and OO5SR regen (holy only). In each of the cases, both int and spi cells were affected. Fortunately, the practical implications haven't really changed, except making spirit even more worthwhile for holy. Also, added some documentation regarding the table columns and updated the gemming section with the new results.

Elimbras · 06/15/10, 9:22 AM

Your full-gain column is not correct yet for the Meditation and OOFSR lines, concerning Int gains.
Unless I'm mistaken, the only gain is BoK, which multiply Int by 1.1.
But due to the Spirit-regen formula, the gain of BoK for int impact on Spirit Regeneration is sqrt(1.1), not 1.1.

Hegen · 06/15/10, 10:50 AM

Originally Posted by Elimbras

But due to the Spirit-regen formula, the gain of BoK for int impact on Spirit Regeneration is sqrt(1.1), not 1.1.

It is, but all of the columns assume the same amount of base int and spirit after effects. So we're looking at a factor of sqrt(1900+1.1)-sqrt(1900) to find out what we gain if we socket 1 int at 1900 int fully raidbuffed.

I agree this may be a counterintuitive depending on what you want from the numbers. What you cannot do with the numbers is extract the overall effect from kings by dividing the cell values. In order to be able to do that, I'd need to assume a base int/spirit unbuffed/untalented and scale that accordingly.

Maybe it's better to remove the columns without talents/buffs altogether as they are not terribly useful.

Elimbras · 06/15/10, 2:45 PM

Originally Posted by Hegen

It is, but all of the columns assume the same amount of base int and spirit after effects. So we're looking at a factor of sqrt(1900+1.1)-sqrt(1900) to find out what we gain if we socket 1 int at 1900 int fully raidbuffed.

Yes, sorry. I've been trying to go too quickly in my computations. The derivative value doesn't change, and the gain in MP5 value is the derivative value, multiplied by real Int gains (ie. after buff int gains when gaining 1 unbuffed Int).

However, even if I had correctly guessed your model, I think you would be right to clearly specify that your base values are considered as during the fight (meaning after buff is there are relevant buff).

xnknown · 06/20/10, 7:26 PM

could you add a section on professions and which ones would be most beneficial to the priest?

cyberblitz · 06/21/10, 4:21 AM

Originally Posted by xnknown

could you add a section on professions and which ones would be most beneficial to the priest?

From a healer perspective I still believe that you don't want something proc related. So as always Jewel Crafting is (IIRC) a minimal bit better than Inscription / Enchanting / Alchemy / Blacksmithing / Leatherworking.

This are - if you just go for spellpower - the only viable options.

I currently got Tailoring since we need a tailor. But that's it. You have to put much effort in it if you want your SP proc at a certain time. If you go for mana you have to do nothing. Just be happy that you have 60% of your mana at the end of a fight instead of 55%.

If you believe you need crit maybe Skinning might be your profession.

Engineering is - as always - a bit tricky. Haste on use and a bit less haste but more SP compared to your regular enchants.

Herbalism and Mining are just useless.

Beside Skinning, Engineering, Tailoring, Herbalism and Mining it shouldn't make a big difference what profession you choose. Maybe your guild / raid needs a special professions since no one covers it or you just go for one you can make most profit with.

Hegen · 06/21/10, 4:28 AM

Originally Posted by xnknown

could you add a section on professions and which ones would be most beneficial to the priest?

Well, the patch with the cataclysm profession and mechanics changes is lurking somewhere close, so I prefer to correct things over adding new content.

The druid guide treats professions quite well: Resto PvE Compendium and General Discussion . The recommendations basically apply to priests, too. Shoot for best spellpower bonuses.

Amoenitas · 06/21/10, 7:47 AM

Originally Posted by Hegen

The druid guide treats professions quite well: Resto PvE Compendium and General Discussion . The recommendations basically apply to priests, too. Shoot for best spellpower bonuses.

I cannot fully agree with that. For many encounters Engineering is the most useful profession for holy priests because it's an extra cooldown for stressful situations (+10% haste for 12s every minute). People rarely die due to constant damage (if they do, the healers are bad).

In addition to that you get the Nitrobooster which saved my a** several times.