constantius

[top] Trinkets

The original write-up this is based on was written by Vurrin, of <Rebirth> on Khadgar-US.

Note that gems in the rough include [Bangle of Endless Blessings] and [Darkmoon Card: Blue Dragon], as well as [Eye of Gruul].

Profession-Based Options:

[Redeemer's Alchemist Stone]: the new Best-in-Slot trinket, only applicable if you are an alchemist. Conservative estimates place this item at 24 Mp5, with possible spikes as high as 50 Mp5. That, together with the 119 Healing S&E makes this an absolutely superb trinket, and a must-have for any serious min/max player.

[Alchemist's Stone]: as with the [Redeemer's Alchemist Stone] above, this trinket provides anywhere from 24 to 50 Mp5, along with 15 intellect and spirit. With typical T6-level gear, those stats are worth an additional 30 Mp5 OO5SR (or 10.2 Mp5 I5SR), making this one of the best regeneration trinkets in the game [best case: 60 Mp5 I5SR].

[Figurine - Seaspray Albatross]: 43 Mp5 overall, with the advantage, as below, of being able to define when you receive some of the mana.

[Figurine - Talasite Owl]: 29 Mp5 overall, with the advantage of being able to use the effect when you actually need mana, which should not be understated.

25-man Raid Drops:

[Glimmering Naaru Sliver]: the new end-game trinket 'drop' (from the token off Muuru), this is an exceptional trinket that provides amazing burst regen. We still don't know how the channel effect works, but are hoping it is not a true channel, but instead a Use effect that acts as a Mp5 'hot', similar to several other trinkets. The high healing on this trinket makes it the ideal complement to a [Memento of Tyrande], if you are so lucky as to be able to score both.

[Memento of Tyrande]: an excellent trinket, and still 2nd-best-in-slot (essentially tied with Glimmering Sliver). If you are not progressing rapidly in Sunwell, this is the trinket to aim for.

[Ashtongue Talisman of Acumen]: this trinket is not particularly strong for holy priests; don't use this unless you feel like playing around with it to try to make it effective. You would have to be using a large number of Renews to make this trinket viable, and CoH is a far superior method of healing in end-game raiding.

[Fel Reaver's Piston]: this trinket has a 45-second internal cooldown. It's basically the [Shard of the Scale] with a crappy proc.

[Earring of Soulful Meditation]: an amazing class trinket, and definitely something you should have in your bags for a variety of fights. If you think there might be a 15 second break where you can go without casting (phase change, or just a Clearcasting -> Inner Focus chain while tank healing), this is the trinket for you. The 66 Healing S&E is very solid, and the Use Effect is quite frankly obscene. In T6-level gear, and assuming 15 seconds OO5SR, this trinket is worth almost as much as a Super Mana Potion. If you cannot ensure time OO5SR, then you would be better served using one of the top 3 trinkets which come with significantly more Mp5.

[Eye of Gruul]: 44 Healing S&E and 450 mana reduction proc. Proc rate is exceedingly low, but if you can use a spell with a low enough mana cost (CoH, basically) you can drop OO5SR with the trinket. This trinket is basically designed for CoH spam; there is no other situation in which a priest will obtain enough benefit from it. A talented CoH is below 450 mana, and the chance-on-heal is per HIT, so each cast of CoH has 5 chances to proc it. Extremely valuable for a fight like Void Reaver (if you're healing melee) or Bloodboil (if you're healing a boil group). The proc rate can work out to something in excess of 50 Mp5 if you are spamming CoH non-stop.

10-man Raid Drops:

[Tome of Diabolic Remedy]: 18 mp/5 and an effective Healing S&E of ~ 65 is very well rounded, although relying heavily on the Use Effect being useful.

[Ribbon of Sacrifice]: 73 Healing S&E. The trinkets effect is of lesser single person benefit than many trinkets, but on fights where you either A) repeatedly heal the same people or B) have multiple people healing your target it can be effective. Some people consider this trinket to be of supreme use on CoH-based fights; your use may vary.

[Pendant of the Violet Eye]: this is not a priest trinket. It is exceptionally good for paladins and shamans, but the net benefit for regen for priests is pitiful, given that we have no real way to obtain the Mp5 except by using healing spells; shamans can use their totem drops, and paladins can use a low-rank of FoL. Even assuming the regen from the 40 intellect, this trinket is worth ~ 30 Mp5, with no other benefit.

Badge / World-drop / 5-man Drops:

[Essence of the Martyr]: 84 Healing S&E and an effective 49 Healing S&E if clicked every time. If healing throughput is what you're lacking this trinket is good.

[Battlemaster's Perseverance]: 88 Healing S&E and a HP Use Effect that is mostly useless in PvE. You may want to consider grabbing this if you are learning Naj'entus while wearing PMC, otherwise, [Essence of the Martyr] is functionally superior in every way.

[Vial of the Sunwell]: 15 Mp5 and a really badly designed Use Effect make this of minimal interest to anyone.

[Bangle of Endless Blessings]: the proc effect on this trinket is worth ~ 80-120 Mp5 depending on your spirit/intellect totals, and assuming 1 ppm, this is an overall 25 Mp5. The Use Effect is worth an additional 185 Mp5 OO5SR, so if you can get a 15 second break, each Use is worth 550 mana. Overall, assuming breaks, a solid 48 Mp5; an excellent option for those who do not have access to some of the above trinkets.

[Lower City Prayerbook]: 70 Healing S&E and assuming 7 casts per click 13 mp/5 or 18 mp/5 in the theoretical best case. 1 minute cool down. It should be noted that if you use all the clicks, this trinket is functionally identical to [Rejuvenating Gem]. If you have Rejuv, though, stick with it -- this one is a lot more hassle to manage, for no net benefit.

[Scarab of the Infinite Cycle]: 70 Healing S&E + haste proc. This trinket is actually fairly sexy for CoH spam in 2.4 given that haste now lowers our GCD, which would allow you to get off almost an additional CoH in the proc time.

[Warp-Scarab Brooch]: this trinket is a lesser version of Diabolic Remedy. It is underpowered compared to Prayerbook or Bangle which are of a similar level.

[Darkmoon Card: Blue Dragon]: due to low proc rate this trinket's effectiveness can vary dramatically, but overall (depending on gear) this trinket can be worth ~ 90 Mp5. Last tests done on this trinket showed it to be approximately 0.5 ppm, which given the current T6-gearing levels, makes it exceptionally strong as a regeneration trinket (Best-in-Slot, actually, if what you want is regen). As a note, at my current levels of gearing, an activation of this trinket is worth somewhere in the vicinity of 2900 mana, or as much as the top-end on a Super Mana potion.

If you raided before TBC, especially in Naxx:

[Eye of the Dead] the "next five casts" is a serious detriment to this trinkets overall usefulness.it is better than Martyr in situations where you're picking and choosing where you want your enlarged heals to land as opposed to trying to get off as many empowered casts as you can. As soon as you cast a sixth spell (during the first 20 seconds) [Essence of the Martyr] becomes a better trinket. Given that you would only cast more than 5 GHs in 20 seconds in an OMG TANK DYING situation, this is functionally the same as Essence of the Martyr, and last it was checked, the +heal Use Effects stacked.

[Warmth of Forgiveness]: overall effective at 24 Mp5 assuming you activate the Use Effect every 3 minutes. Not as powerful as the J/C trinkets available in TBC.

[Scarab Brooch] is mostly a toy. Fun to play with, though.

[Rejuvenating Gem]: this remains one of the finest healing trinkets in the game. Its balance between healing and regen is only matched at the level of [Memento of Tyrande], and remains a staple of many priests' gear right through to Illidan.

[Shard of the Scale] is basically useless now; if you want a pure Mp5 trinket, see the [Vial of the Sunwell] above.

Enchants

Head: [Glyph of Renewal] (HH - Revered) or [Prophetic Aura] (ZG Honored; requires a drop from Bloodlord Mandokir or Jin'Do the Hexxer)

Shoulders: [Greater Inscription of the Oracle] (Scryers - Exalted) or [Greater Inscription of Faith] (Aldor - Exalted) or [Zandalar Signet of Serenity] (Zandalar - Exalted)

Cloak: [Formula: Enchant Cloak - Subtlety]

Chest: [Enchant Chest - Major Spirit]; +6 Mp5; +6 all stats; +150 mana (poor choice)

Bracers: [Enchant Bracer - Superior Healing]; +6 Mp5; +4 all stats

Gloves: [Enchant Gloves - Major Healing]

Pants: [Golden Spellthread] (Aldor - Exalted : Recipe) or [Prophetic Aura] (Zandalar - Honored; requires drop)

Boots: [Enchant Boots - Boar's Speed]; Vitality (4 Mp5); 12 stamina

Rings: [Enchant Ring - Healing Power] (enchanter-only)

Weapon: [Enchant Weapon - Major Healing]; Spellsurge; +20 spirit (only for a spirit weapon)

[top] Interesting Math Stuff

Question 1: with respect to Inspiration, where is the 'sweet spot' for crit, beyond which stacking any additional crit sees reduced or nonexistant return.

Inspiration: 15 second duration, applied at the conclusion of a holy spell that crits

Assumption: all spells are independent, binary random variability on crit chance.

Using Greater Heal, what are the chances that Inspiration will be reapplied before the buff wears off? Now, of course, requiring continuous up-time is misleading. Really, what we care about is absolute uptime, as a percentage of total time-in-fight. We can find the expected number of casts until a crit fairly easily:

E[numCastsUntilCrit] is (1-p)/p where p = probability of a crit, so:

@ 8% : (1-0.08)/0.08 -> 11.5
@ 10% : (1-0.10)/0.10 -> 9.0
@ 12% : (1-0.12)/0.12 -> 7.3
@ 14% : (1-0.14)/0.14 -> 6.1

So if we have 14% spell crit, and are spamming on a single target, we expect a crit to occur within 15 seconds ~ 50% of the time.

To compute absolute up-time, the value 1-(1-p)^n, where n represents the number of casts required to cover the time span of the buff (15 seconds). Since we are using GHeal as our base spell, n=6.

What's the progression from 10 to 11 to 12 to 13 to 14?
@ 10% -> ~ 46% absolute up-time
@ 11% -> ~ 50% absolute up-time
@ 12% -> ~ 54% absolute up-time
@ 13% -> ~ 57% absolute up-time
@ 14% -> ~ 60% absolute up-time

Given that with the amount of intellect appearing on T6 gear, and the small amounts of spell crit we see (2 items in Sunwell), most priests will run ~ 10-14% raid-buffed intellect, it makes perfect sense to set yourself at a hard cap of 14%, and not bother stacking past that point.

Question 2: does stacking crit increase my CoH throughput (i.e. amount healed per cast of CoH)?

Let's assume you have a 10% Holy Crit rate. Let's also assume (as far as I know, this is correct) that each hit of CoH is computed separately with respect to crit chance. CoH scales as 22% HSE Coefficient. Assume Spiritual Healing for another 10%.

Some values to play with, assuming base of 0% crit and 2400 +heal, throughput of CoH per cast is:
@ 10% crit: (430+0.22*2400)*1.1*[0.90*5 + 0.10*5*1.5] is 5532 HpCast
@ 15% crit: (430+0.22*2400)*1.1*[0.85*5 + 0.15*5*1.5] is 5664 HpCast
@ 20% crit: (430+0.22*2400)*1.1*[0.80*5 + 0.20*5*1.5] is 5795 HpCast

Fixing crit @ 10%, and scaling +heal:
@ 2400: (430+0.22*2400)*1.1*[0.90*5 + 0.10*5*1.5] is 5532 HpCast
@ 2500: (430+0.22*2500)*1.1*[0.90*5 + 0.10*5*1.5] is 5660 HpCast
@ 2600: (430+0.22*2600)*1.1*[0.90*5 + 0.10*5*1.5] is 5787 HpCast

Increasing your +heal by 1 increases total healing done by a cast of CoH by 1.28.
Increasing your +crit by 1% increases healing done by a cast of CoH by ~ 26.

Note that this completely ignores Haste as a factor. I'll post more questions below with details on Haste.

Anyway, conclusion is ... to get the same benefit from increased healing / cast, you equate 20 HSE = 1.0% SpellCrit. So if you primarily spam CoH, gaining 1% spell crit is the same as gaining 20 HSE. Of course, 22.08 spell crit rating is worth a good sight more ilvl points than than 20 HSE, so very clearly stacking +crit on your gear is a stupid thing to do if you want to increase HpCast from your CoH. Of course, if you have the option of getting an item which has high HSE *and* high Spell crit, for the cost of regen, then you have a call to make.

I'm primarily thinking of items like [Ring of Calming Waves], which, for sheer CoH spam is worth (using above numbers) 24/22.08*20 + 27*1.1/80*20 + 64 = 93.1 HSE. This compares quite favourably to something like [Ring of Harmonic Beauty] which clocks in at 73 + 22*1.1/80*20 + 32*1.05*1.1*.35 = 92.0 HSE. Of course, I think most of us would rather have 32 spirit equivalence in regen than more crit ... maybe that's just me.

So, finally: for CoH throughput, Holy Specialization is worth (according to these numbers) 100 HSE. That's a reasonable gain from a Tier1 talent. Comparatively, Spiritual Healing is worth ~ 500 HSE at Sunwell gear levels, and is a Tier6 talent.

Question 3: How does Empowered Healing (the talent) work, and what is the net benefit?

Test: GH7 (no downranking penalty). 5/5 Emp Healing, 5/5 Spiritual Healing, 4-piece T6 bonus on. 2522 HSE.

84 casts, average heal of 6053.

(2609+ 2522*1.2*3/3.5)*1.1*1.05 = 6009
(2609 + 2522*3/3.5 + 2522*0.2)*1.1*1.05 = 6093

Test: GH6 (no downranking penalty). 5/5 Emp Healing, 5/5 Spiritual Healing, 4-piece T6 bonus on. 2522 HSE.

160 casts, average heal of 5714.

(2276 + 2522*1.2*3/3.5)*1.1*1.05 = 5625
(2276 + 2522*3/3.5 + 2522*0.2)*1.1*1.05 = 5708

[e] As a note: don't try to test this crap in Shattrah when the A'dal buff is up. It scales your healing without changing your spellbook or character screen, and really can make you doubt your sanity. Thanks to Kalman for helping me stay sane.

So formal calculation for a Greater Heal is:
which is actually something new to me. I haven't seriously looked at Empowered Healing for about a year; when I last did the numbers on it, I think I was running 1800 HSE, and in the testing I did, came to the conclusion that it was

Question 4: Why do my spellbook values not match up with Wowhead?
Wowhead and Thottbot report the spell values as they are when you learned them for the first time. There is a slight inflation in spell values as you level from when you learned them to your maximum level. As an example, Wowhead reports average GH:7 as 2590. My Spellbook reports it as 2609.