Balance Druid - DPS Trinket Discussion
 

Blizzard has really broken open trinket itemization giving caster DPS in particular a huge number of options to decide between. This thread will be available to discuss the comparative merit of these trinkets and a location to gather any hard analysis with first post updates to craft a tiered ranking and flashpoint analysis. In the future WWS posts from clears with static parameters outside of trinket usage can be used to seen some real-world results of the choices more at the mercy of the RNG. Conversations on the best way to maximize these procs/activated buffs would also be appreciated.


-


Initial First Tier Impressions:




[Dying Curse] (71 hit (2.71%), chance on spell cast to increase spell power by 765 for 10 seconds): Drops from Heigan in Naxxramas-25.

Starfire [124.676] /Wrath [144.442] : Calculated average spell power value of proc at (15% haste, 50% crit) assuming a 15% proc chance. These are chain rotations; dot applications will increase the value for Starfire and lower it for wrath. Starfall is tested as giving a chance to proc on every star that hits.

[139.2 Spell Power average]: Asheara's numbers using WWS to determine his live time between trigger events



[Sundial of the Exiled] (84 critical rating (1.83%), chance on spell cast to increase spell power by 590 for 10 seconds): Available for 40 Emblems of Heroism.

Starfire [96.155]/Wrath [111.399] : Calculated average spell power value of proc at (15% haste, 50% crit) assuming a 15% proc chance. No latency considered. These are chain rotations; dot applications will increase the value for Starfire and lower it for wrath. Starfall is tested as giving a chance to proc on every star that hits.

[107.4 Spell Power average] : Asheara's numbers using WWS to determine his live time between trigger events



[Illustration of the Dragon Soul] (18 Spell power for 10 seconds on spell cast and stacks to 10): Drops from Sartharion-25

Equivalent to a static 180 spell power in the trinket slot.

Just like the DMC: Crusade there will be encounters that it suffers in due to down-time phases. Can be stacked fairly quickly with initial dot applications and early Wrath usage. Pre-stacking with thorns before a boss fight is easy to do: Get it to 10 stacks then drink; when the buff gets to one second recast thorns. At that point your natural regen should bring you to max before you have to recast it. This method can also be used within the boss fight during phase changes. Herbalism Lifeblood heal will definitely maintain the stack at no mana cost; need testing on bandaging and other free abilities that are apparently able to proc Sundial/Dying Curse.

NOTE: Pre 3.0.8 this was worth 260 spell power.



[Embrace of the Spider] (98 Spell Power and chance to increase haste rating by 505 for 10 seconds): Drops from Maexxna in Naxxramas-10

Asheara's testing indicates a lower proc rate than the spell power procs. His results indicate a 10% proc rate for an average of 10 casts per proc with the 45 second ICD.

Starfire [73.857 haste rating or 2.25% haste average]/Wrath [89.578 haste rating or 2.73% haste average]. Calculated average haste value of proc at (15% haste, 50% crit) assuming a 10% proc chance. No latency considered. These are chain rotations; dot applications will increase the value for Starfire and lower it for wrath. Starfall is tested as giving a chance to proc on every star that hits.

[2.49% haste on average] : Asheara's numbers using WWS to determine his live time between trigger events


Consider also that during the proc you will only maximize its' value by using starfire. Wrath will be wasting the entire buff on every crit (. One to haste is starting a cast as the buff expires still gets you the benefit for the full cast making Starfire usage less likely to waste buff time. Hard to reccommend this trinket considering the other options.



[Mark of the War Prisoner] (73 hit rating (2.77%) and 346 spell power on use for 20 seconds)
Average 57.66 spell power for the activation.
Drops from Cyanigosa in Heroic Violet Hold
Deserves mention as a non-raid trinket with hit rating.




http://elitistjerks.com/f73/t37323-b...on/#post986849: Asheara's live data of proc rates with detailed breakdowns. Thanks for his work.




Discussion should center around the relative value of these procs and strategies for getting the best value out of them. My individual experience with Sundial is to maintain your current disciplined spell/dot refresh for the first 5 seconds; at that point if you're in a Starfire cycle a change to wrath is the best way to prevent wasted uptime. One of the enormous benefits of Wrath/Starfire being so close in DPS is the ability to adjust to procs in this manner. Applying glyphed dots under these effects IS beneficial; our glyphed dots only scale properly with spell power but they scale extremely well with it.


-

thedopefishlives stated here: http://elitistjerks.com/f73/t36776-d...p4/#post985366 that Rawr modeling has [Dying Curse] and [Sundial of the Exiled] as our top two trinkets at some gear loadouts.

-


Proc trinkets are listed in wowhead as having a 15% proc rate with a 45 second internal cooldown. Static proc rates greatly favor wrath usage for maximizing uptime. Once the ICD is up it should take an average of 6-7 casts to trigger and the faster you trigger it the sooner you start the next ICD.

Ashaera's live testing with wrath shows an average of 6-7 casts to trigger the proc for both Sundial and Dying Curse. This is consistent with the presumed 15% proc chance expected number of casts. Starfire still needs testing to determine potential for a PPM type system.


-

Opinions: I feel Illustration of the Dragon Soul is significantly better than the Sundial in most situations based on this initial calculation. It seems to be fairly equivalent to Dying Curse with similar usage issues (Procs are inherently vulnerable to random movement fights compared to activated effects.)

-

Testing/Analysis Needed:

Analysis of haste value related to these trinkets. Second run-through/vetting of theoretical calculations; perhaps at different gear levels.

An in-game test of actual proc rate of Dying Curse/Sundial of the Exiled to determine if it is PPM or flat percentage proc rate. (Wrath has been tested, Starfire test is needed)


Testing Starfall to see if each star builds IoDS stacks for quick ramp-up.


Adding second tier procs in the near future.

-


Note: I will be updating this frequently. Any requests and/or mistakes just PM me and I will adjust things properly. There is a definite plan to add the vast majority of the trinkets once discussion has expanded.

You can ramp up Illustration with any spell - we have a mage and a warlock with it (yes, 2 in 2 kills), and the mage uses Slowfall to stack it up while the warlock uses Detect Invis (or possibly underwater breathing, not sure which). This means they have effectively passive 260 SP, even through periods where bosses cannot be damaged/etc. I assume we can do a similar thing with Thorns (I can't think of anything else castable on friendlies in Moonkin) to stack up the buff pre-fight, although Thorns is not exactly a cheap spell (17% of base mana) so I'm not sure on the applicability of this.

On procs and recasting dots: The gain from the SP on the dots is generally not enough to be worth clipping the existing spells during procs, unless there are 2 ticks or less remaining and even then it is of marginal value.

One of the advantages of having a lot of instants means that even if we do get a proc just before/while moving, we can still (mostly) use it. In this case it's worth clipping dots, and using Typhoon and Starfall are also useful while moving. One thing I haven't properly tested yet is whether it's worth waiting for trinkets to proc in order to cast Treants (seeing as they tend to proc pretty quickly at the start of the fight anyway). Do Treants scale based on your SP at cast point, or on SP at the time they hit? It would certainly be worth waiting for procs (especially double proc if you have both Curse and Sundial) if they scale based on point of cast.

ninjaedit: Further trinkets to include:
[Embrace of the Spider] Naxx N
[Item not found!] Naxx H
[Tears of Bitter Anguish] (apparently a world BoE)

Those are all the epic trinkets I can find of worth for a damage caster.

I thought of pre-stacking but none of our spells we can cast out of combat is very cheap on the mana cost. Maybe if you were the raid leader and could do some pre-stack/drink/refresh then go shenanigans it might work well. With warlocks I wouldn't be surprised if lifetap builds it as well making it fairly trivial for them. If you have the mana available, thorns every nine seconds isn't a bad idea at all

Refreshing a dot with spell power buffs is a precision/awareness thing. With the Squawk and Awe tick timer if you can see a dot is going to tick right as your cast will finish you can squeeze some DPS out with it. It's not something that can be forced; you just have to be aware and let it happen. Certainly not a technique for most people.

Waiting for trinkets to treants is such a dynamic choice. Some boss fights you need to cast them immediately to get damage in before the aoe mechanics start; others the ideal time is in a specific phase. Look at the duration of the fight and if waiting too long will prevent a second Force of Nature: that means it's a bad time to wait for it. I've personally had trouble determining whether they scale at point of cast through raid experience just because of their damage range.


EDIT: I'll add those Soul of the Dead and Mark of the War Prisoner (Standout heroic trinket) as soon as I get home. Thanks for the input!

We had out first raid last night, and cleared 3 wings...with this and looking at loot the bosses drop, I am finding a scary lack of hit, considering how much we now need. Granted I am still in my full t6/sunwell gear, so far all of my upgrades have had 0 hit and almost no haste. They do however have a ton of stam, crit and spellpower.

I think with this trend the trinket with 71 hit on it is pretty much a guaranteed trinket spot.

The +hit trinkets are very likely going to be a requirement since none of the leather spell armor seems to have any +hit on it. There is a lot of cloth armor with +hit on it, but there's also a lot more competition for those pieces. I'm currently running the JC +hit trinket and another +hit trinket just to hit the 262 number (they're both on-click +dmg, which is a problem, but I don't really have another choice for right now).

Some initial testing of [Embrace of the Spider], [Dying Curse], [Sundial of the Exiled] shows me the following :
- Insect Swarm counts as a harmfull spell cast, i.e. proccing Sundial of the Exiled aswell.
- Starfall has a chance to proc atleast Embrace & Dying Curse on each individual star. Over atleast 10casts I havnt seen Sundial of the Exiled proc once.
- Dot ticks will not proc any of the trinkets (obvious I guess).
- Hurricane's initial cast can proc them, but they don't have a proc chance on each "tick".
- Force of Nature: I would assume the cast to proc Embrace & Dying Curse - But 0 procs on any of the 3 trinkets in well over 10 attempts.
- Practically anything will proc the "Your spells has a chance too" trinkets, including silly stuff like making bandages, buffing etc.


All tests done spamming wrath to reach max amount of casts / minute.

[Embrace of the Spider]
Time to proc:
Test 1: 0.09 - 1.10 - 2.00 - 2.50 - 3.48 - 4.42 - 6.00
Test 2: 0.03 - 0.49 - 1.56 - 2.43 - 3.30 - 4.58 - 6.01 - 6.53 - 7.58
Minimum time between proc appears to be : 45seconds
Average 54.9 seconds between procs.

Casts to proc:
8, 1, 15, 4, 19, 15, 3, 11, 4, 16, 16 - An averaged 10,2 casts to proc.


[Dying Curse]
Time to proc:
Test 1: 0.14 - 1.02 - 1.54 - 2.42 - 3.38 - 4.30 - 5.20
Test 2: 0.09 - 0.57 - 1.46 - 2.31 - 3.23 - 4.26
Minimum time between proc appears to be : 45seconds, with the red marked period being a mix of luck & a result of me using ingame stopwatch+paper instead of a proctracker.
Average 50.3 seconds between procs.

Casts to proc:
1, 3, 11, 2, 18, 4, 2, 2, 5, 20, 9, 2 - An averaged 6.6 casts to proc.


[Sundial of the Exiled]
Time to proc:
Test 1: 0.13 - 1.01 - 1.49 - 3.10 - 4.16 - 5.02 - 5.58 - 6.47 - 7.38 - 8.38 - 9.27 - 10.28
Minimum time between proc appears to be : 45seconds
Average 54.8 seconds between procs.

Casts to proc:
7, 1, 1, 9, 5, 3, 2, 18, 7, 7, 6, 6 -- An averaged 6casts to proc.


Allthough this is a rather small sample size I think it gives some strong indications.
- The trinkets seem to share a 45second cooldown between procs, dying curse might be 40seconds only.
- [Embrace of the Spider] appears to have a lower proc chance than [Dying Curse] & [Sundial of the Exiled]
- [Embrace of the Spider] & [Dying Curse] procs off Innervate & individual Starfall stars hitting.


Weak attempt to calculate procs into a static effect
Numbers below taken from my last Patchwerk wws. (Rounding up abit here, using proper numbers in calculations)
I had 1.475 seconds between each event capeable of proccing [Embrace of the Spider] & [Dying Curse] & 1.66 seconds between each event capeable of proccing [Sundial of the Exiled].
Most of the difference comes from Starfall ticks, which comes in 2 close together periods, so i'll use the number for [Sundial of the Exiled] in further calculations.

Assuming
10 events needed to proc [Embrace of the Spider].
6 events needed to proc [Dying Curse] & [Sundial of the Exiled].
45seconds as minimum time between possible procs.


Time to proc:
[Embrace of the Spider] : 1.66*10 = 16.6seconds to proc.
[Dying Curse] : 1.66*6 = 9.96 seconds to proc.
[Sundial of the Exiled] : 1.66*6 = 9.96seconds to proc.

Uptime:
[Embrace of the Spider] : Time between procs (45+16.6) 61.6seconds so (10/61.6)*100 = 16.2% uptime.
[Dying Curse] : Time between procs (45+9.96) 54.96seconds. (10/54.96)*100 = 18.2%
[Sundial of the Exiled] : Equal to Dying Curse = 18.2%

Result:
[Embrace of the Spider] : 98spellpower & 2.49% haste.
[Dying Curse] : 2,71% hit & 139.2 spellpower.
[Sundial of the Exiled] : 1.83% crit & 107.4 spellpower

Removing 98spellpower that gives:
[Embrace of the Spider] : 2.49% haste.
[Dying Curse] : 2,71% hit & 41.2spellpower.
[Sundial of the Exiled] : 1.83% crit & 9.4 spellpower

Making Dying Curse a clear winner if u can utilize the hit & gives you a hard 2nd choice.

Thanks for the awesome live testing. It certainly seems to confirm the 15% proc chance numbers that was datamined and used with the theoretical numbers I ran (Your spell power values are decidedly similar to mine which is some additional confirmation). I'm assuming this was done with just normal personal moonkin form buffs? Likely your haste was a bit lower than the 15% I used as a sample number and that alone could account for the slight differences in wrath numbers. Next step would be to include these trinkets in raid DPS modeling and use that to find out how they affect the stat weight of haste when using one and/or both trinkets.




I actually acquired an [Illustration of the Dragon Soul] last night so I will test with Starfall to see if each star builds the stack and if it builds two stacks if both stars can be used. I'm going to update the main post with some of your data and observations as well as some thoughts on Illustration.

For my estimation on events to proc I used a Patchwerk wws - I swapped a few items since then, but I was around 10-11% unbuffed haste, 13-14% with 3/3 Celestrial Focus & rougly 16% if you count [Embrace of the Spider].
The procs scale (allthough not by much) with your haste %.

[Illustration of the Dragon Soul] seems very strong & in most current raid content it would be fairly easy to keep the buff stacked just by doing what you would normally do.


At some point I'll try doing a longer test with Dying Curse to see if I can push the proctime below the 45second mark - For consistancy I was expecting a 10second uptime to produce a 30second downtime, as most TBC trinkets seemed to use the uptime*3 = hidden downtime formula.
With procs being as powerful as they are, then maybe this has changed though.

2 other trinkets to take in consideration are [Item not found!] since moonfire should be up 100% and [Soul of the Dead] just for the huge amount of crit.

[Soul of the Dead] -> 2.07% crit from a trinket slot is not massive, compared to the approximate number for some of the other epic options.
There are better blue options: [Forge Ember], [Mark of the War Prisoner] or even [Pendulum of Telluric Currents] - Or you could use the jewelcrafting or alchemy trinkets for greater benifit aswell.

[Item not found!] is harder to put a value on - I cant get Rawr to funktion on my system, but you would have to put an approximate value on how much dps u gain from various stats (which is very gear dependant) & then try to figure out how much dps the proc is worth with your current rotation.
Does it have an internal cooldown? Is the 10% proc chance listed on wowhead correct?

Having played mage for 3years I tend to just ignore trinkets like that one, but maybe my gut feeling is wrong here. Either way an affliction warlock/shadow priest will probably get more benifit from it.

[Item not found!] is something that would need tested to get the definitive proc rate and cooldown numbers on. My initial thought is rather negative with the poor itemization value of crit.

Considering the wowhead datamining indicates a 45 second cooldown you're looking at around 22 DPS maximum assuming it procs immediately off of cooldown. That doesn't hold up to the value of Mark of the War Prisoner's on use value. Not even worth putting in as a viable raid trinket when several heroic trinkets surpass it; Forge Ember on the other hand is one that is another excellent pre-raid option I'm adding.


I haven't included [Soul of the Dead] yet because the main utility would be mana return. It's just not a very good pure DPS option unless you're mana starved. Trying to keep this focused on ideal options first; especially considering the fairly big gap between trinket values at this point. Likely to slowly separate the list into pre-raid and raid acquired trinkets.

Anyone been able to test a [The Lightning Capacitor] + [Thunder Capacitor] combo? I have gotten good results from the thunder trinket in AoE situations, especially since it can proc OoC during Hurricane (which has its own proc chance for OoC reduced). I am trying to decide if it is worth going back for the Illhoof trinket. From what I hear, if you have both trinkets on they each gain charges at the same time, and a crit on the bolt from one trinket adds a charge to the other trinket.

I want to toss in one point about Dying Curse + Mark fo the War Prisoner. As most (all?) 'kins dont pick up IFF having Mark in the bags after getting Dying Curse is a real help for when a shadow priest isn't in the raid. The 73 hit almost makes up for the missing 78.69 hit rating. I plan to be about 6 over hit max just to hedge the bets.

I really, really, want to use Dying Curse, but I can't seem to itemize out enough hit for it to be good in the long run. Here's the hit gear I'm looking at having going into Ulduar.

Valorous Dreamwalker Cover
Valorous Dreamwalker Mantle
Leggings of the Wanton Spellcaster
Wyrmrest Necklace of Power
Surplus Limb
Cincture of Polarity

It's already 30 rating over my 9% hit cap, and I can't find any good haste/crit items to replace these slots with to itemize out enough hit to make Dying Curse worth while. At best, I can trade the hit on some of the pieces for spirit, at which point it would be better to just use another trinket.

Does anyone see any ways to drop some of that hit for enough crit/haste/spell power to make Dying Curse better than Embrace of the Spider?

I can't see anything in terms of gear that would make it better directly, but can you make it so you're 2% over the cap and drop a point in Balance of Power in order to pick up a point in something else?

fakeedit: You could go with [Cosmic Lights] for the neck, which I'm pretty sure is better than the Malygos quest one you've linked if all the hit is going to waste (unless you can put in a dragons eye). Combine that with switching a point in Balance of Power for something else while using Dying Curse, and you gain overall I believe.

You could also do with fixing your profile - you're a draenei hunter with no alts looking at end game moonkin raiding gear? :P