Specialization Tier List - Hekili/hekili GitHub Wiki

Hello, I am the tier list. Yes I am up to date for TWW.

Note: Thanks to @PandaTVoce / Twitch for supporting the upkeep of this article.

Which Specializations Work Best with Hekili?

The short answer...

DPS

Tier Specializations
S Feral, Retribution, Enhancement, Outlaw, Fury
A Frost DK, Havoc, Beast Mastery*, Windwalker, Assassination, Subtlety, Affliction*, Arms, Devastation
B Arcane*, Frost Mage*, Unholy, Balance, Augmentation, Marksmanship*, Survival, Shadow*, Elemental, Demonology*, Destruction
C
D Fire*,**

* - Moved up since the June 2020 review.
** - Specialization mechanics require more CPU usage.

Tanks

Tier Specializations
S Brewmaster, Protection Paladin
A Blood, Vengeance
B Guardian, Protection Warrior

How They're Ranked

There are several factors that impact these rankings.

Melee vs. Ranged

Melee specializations have an advantage over most ranged, because melee specializations can rely on nameplates for target detection. Ranged specializations have historically relied upon damage- (or DOT-) based damage detection, which means you have to hit things before the addon knows which targets you care about. The reason is that the WoW API doesn't provide positioning data in most settings. The addon can tell roughly how far a target is from you, but not where that target is in respect to other targets.

One improvement in this area, since Battle for Azeroth, is ranged classes that use a melee pet. Marksmanship/Beast Mastery Hunters and Affliction/Demonology Warlocks can enable pet-based target detection in their specialization settings (i.e., /hekili > Marksmanship > Targeting) and with a little bit of setup, when you are targeting an enemy near your pet, the addon can count targets within a certain radius of your pet. That led to these ranged specializations moving up a tier, when configured accordingly.

Predictability vs. Reactivity

Some specializations have a great deal of unpredictable variability. For example, Havoc Demon Hunters can talent for Demon Blades, which makes a significant portion of your Fury generation unpredictable. It would not be safe for the addon to make recommendations assuming that you're going to have Fury that can't be predicted, so the model is forced to be reactive: if you auto-attack and that 60% chance procs, the addon updates its recommendations accordingly. Similarly, Fire Mages have to watch their critical hits carefully to convert Heating Up into Hot Streak. The addon can assume a spell is going to crit during Combustion, but outside of Combustion, the addon has to wait and see -- leading to reactive updates that might not afford you a great deal of time to react to the changes.

Single vs. Multi-Target

Some specializations have dramatically different AOE priorities compared to their single-target priorities. While this addon will happily model most of those differences, some priorities are written such that they take advantage of raid_events in SimulationCraft, knowing that add waves will arrive in a specific amount of time. In game, the addon does not model encounters this way; you either have multiple enemies up now, or you don't. The addon provides confabulated data regarding raid_event.adds.up and similar expressions, returning true if there are multiple targets now and false if there aren't.

I'll note that target detection itself is a key part of single vs. multi-target recommendations. If your ranged specialization needs to rely on damage-based target detection, you won't get those multi-target recommendations until you've hit the targets with something OR you use a display mode that uses the AOE display to show AOE recommendations even before those targets are truly detected.

Class Review

All of the following class review information is presented with the caveat: You will always get more out of the addon, in more challenging content, the more you know about your class and specialization mechanics. If you're a Blood DK and you know you're about to get clobbered and you need to Death Strike after, holding onto RP for Death Strike is probably a wise choice. If you're a Feral Druid surrounded by enemies but there's a priority target that needs to die immediately, you should probably skip Primal Wrath and hit Ferocious Bite instead. (And remember, you can toggle the addon to make single-target recommendations during that sort of phase, even with enemies nearby!)

Demon Hunter

Havoc: A
Vengeance: A

Both specializations continue to play well with the addon, though Havoc comes pre-loaded with a warning about RNG-based resource generation. I'd encourage Havoc players to also keep track of Soul Fragments on their own, as the addon handles those reactively -- you won't get a recommendation to move to pick them up. Move to get those when your Fury is lower. The addon assumes you'll pick up fragments when encounter mechanics allow it.

Vengeance works well, though the active mitigation mechanics of Vengeance are a little underwhelming. Remember that priorities can always change over the course of an expansion.

Death Knight

Blood: A
Frost: A
Unholy: B

All DK specializations continue to work well with the addon. Unholy was held back slightly in Battle for Azeroth as trying to pool Festering Wounds to pop during Death and Decay for Festermight had varying results throughout the expansion. As Festermight fades away, that mechanic becomes easier to ignore and less intrusive. There remains some pooling logic for wounds with respect to Apocalypse and so forth, so timing your next ability recommendation can be a bit challenging for the addon. If you ever see missing recommendations, this is likely the cause -- and this will be sorted out as priorities are iterated upon during Shadowlands.

Blood works rather well. The Ossuary and Bone Shield changes emphasize the need to keep Bone Shield stacks high. The default priority does a pretty solid job of trying to keep Blood Shield up (i.e., refreshing before it falls off) while reserving RP so you can also Death Strike when you need to recover health or build a bigger shield.

Druid

Balance: B
Feral: S
Guardian: B

Feral is what I've spent the most time playing and that leads to more time confirming that the addon's model is accurate and consistent. The Shadowlands priorities include some functionality around calculating how many ticks of Rip you'll get from Primal Wrath vs. Rip based on remaining duration and number of enemies. The addon approximates this well.

Balance's revamp is simpler and easier to model, generally. Stellar Drift seems popular for M+, and there's some logic tweaked to try to keep that up so you can cast and move. The first revision of Balance for Shadowlands wasn't quite modeled in line with SimulationCraft, which led to some inconsistencies that were recently resolved.

Guardian works well, but Guardian is so boring that it's hard to rate it highly. You don't really have fun buttons to push, though that is a subjective judgment.

Hunter

Beast Mastery: A
Marksmanship: B
Survival: B

Survival's big challenge remains that Wildfire Bombs are a ranged cone. The addon expects that you're shifting your position to make sure those bombs hit as many targets as possible, but you might not actually do that.

Beast Mastery and Marksmanship's rankings improved a bit because they have the option of using pet-based target detection now (though many Marksmanship players will opt for Lone Wolf). Using pet-based target detection with a Blood Beast's Blood Bolt ability (note: you don't have to use a Blood Beast, you just need to have tamed one to put Blood Bolt on your bar) is pretty phenomenal for Beast Cleave. Barbed Shot is still a little risky in terms of daisy-chaining your bleeds... If you know a fight's mechanics is going to force you to move out of range, you might want to refresh early. Or not. Play thoughtfully.

Mage

Arcane: B
Fire: D
Frost: B

I've moved all Mage specs up a bit, though Fire still sits at a D simply because there is so much reactivity to the specialization. As many of the extremely overwrought aspects of the Battle for Azeroth arsenal fall away (the bracers, Memory of Lucid Dreams, etc.), the modeling does get a bit easier which has resulted in less CPU usage and somewhat stabler recommendations. But it's still far and away the most volatile specialization.

Arcane's model has been successfully updated for Shadowlands, though there have been multiple iterations to enable or undermine the "Arcane Missiles" spam that some folks have enjoyed. The Shadowlands version of Hekili has several key improvements for modeling channeled spells and aura ticks, with improvements to accuracy and processing time. Arcane should work pretty nicely, though you'll still have to remember to "prime the pump" for multi-target.

Frost is a B but almost an A. It can't use pet-based target detection since the Water Elemental is a ranged caster, too. But I've made improvements even in the past 24 hours that help to time Ice Lance impacts effectively and give proper recommendations based on whether you can double-dip on a Winter's Chill or Frozen debuff.

Monk

Brewmaster: S
Windwalker: A

Brewmaster has changed a fair bit with Celestial Brew replacing the maintenance buff Ironskin Brew. Tanks are generally updated more slowly in SimulationCraft, so I fully expect to see Celestial Brew usage evolve over time. Based on user feedback, I've emphasized using Purifying Brew such that you aren't sitting at max charges when you could be building up Purified Chi for your next Celestial Brew, and then using Celestial Brew more proactively. With all your cooldown reduction effects, you get a fairly good uptime on Celestial Brew anyway and you recover Purifying Brew charges more rapidly than you might expect.

Windwalker is one of my favorite specializations and always works well. Over the course of beta, the priority had to be refined a few times to consistently provide recommendations without getting blocked by mastery or Hit Combo, and seems to be in a healthy place.

Paladin

Protection: A
Retribution: S

These pink power rangers continue to be easy to model, so their rankings remain unchanged. The Retribution priority will likely see some continued iteration in early Shadowlands -- aligning Seraphim, Final Reckoning, and Execution Sentence in a Patchwerk simulation has proven a bit easier than in-game, so I'll be keeping my eye on it.

Priest

Shadow: B

Shadow was previously ranked a bit lower, but with the improvements to resource modeling (it's easier to keep track of incoming Insanity and Voidform-based decay isn't the real way to play) it gets bumped up. It also benefits from new modeling around channeled spells, breaking channels, and tick effects. The main issue that results in more reactivity than I prefer is that your opener will likely see a Void Eruption recommendation pop in while Mind Flaying or Void Torrenting that should be more predictable.

Note: I do plan on adding Discipline DPS at some point, along with other healing specializations.

Rogue

Assassination: A
Outlaw: S
Subtlety: A

Since rerolling is less of a "thing" in Shadowlands, Outlaw continues to be strong and functional and remains in the S-tier. If it had been A-tier previously, I'd have bumped it up to S. So it's really an S+ at the moment. There was a period in Shadowlands beta where Roll the Bones required more energy than your other combo spenders, which was problematic in terms of keeping Roll the Bones up, but that's been resolved.

Subtlety continues to require some pooling behavior, which can appear a bit misleading. People get confused when they see a cooldown spiral on Envenom despite Envenom not being on cooldown. Remember -- if you see a timer on something that's not on cooldown, it probably means (1) you're pooling resources or (2) you're waiting for a de/buff to fall into refreshable range.

Assassination continues to work well, with snapshots and Exsanguinate modeled as expected.

Shaman

Elemental: C
Enhancement: S

Elemental could probably even be a B now, as the class theorycrafters have done a great job of cleaning up the overly complicated priority that we saw in Battle for Azeroth. I'm keeping it at a C for now.

Enhancement is where the addon started, so it will always be S-tier. Most feedback has been positive for Shadowlands, and the spec is expected to be stronger in this expansion which is also good news. Given the volume of changes to this specialization, I'd just like to emphasize that (1) I'm rather confident about the addon's recommendations and (2) you may need to reset your expectations about priority because it's just so different now.

Warlock

Affliction: A
Demonology: B
Destruction: B

Affliction and Demonology both move up a tier thanks to pet-based target detection (assuming you're using a Voidwalker, Felhunter, or Felguard). Destruction isn't exceptionally different and remains where it was, though I could almost promote it to an A due to another recent addon improvement. If you use a mouseover macro to apply your DOTs to other targets, the addon can tell in advance that its not going on your current target and adjust accordingly. For example, if you're casting Unstable Affliction on Dummy A via mouseover macro while targeting Dummy B, the addon will know that Dummy B is not going to have Unstable Affliction on it at the end of the cast. Previous versions of the addon would assume you're applying effects to your current target, so this is an improvement to multidotting for sure.

Multidotting will be further improved in a future update that treats secondary targets as fully-formed units. That will enable the addon to do a better job of warning you that Agony is going to fall off a secondary target ahead of time, rather than reacting once the debuff has fallen off.

Warrior

Arms: A
Fury: S
Protection: B

All melee, all good. Arms sits at an A (rather than S) because of the long swing times of their two-handers, which makes Rage generation slow and spiky. It's modeled accurately, but having to sit and wait can sometimes feel incorrect.

Protection continues to work by being cautious about dumping Rage that you might need for mitigation, which keeps it a tier below most other tanks.