Yeah, I know, long time no watercooler. I’ve been busy.
I wanted to take a moment to provide context for some of the Patch 5.4 class changes that we’ve been making. If you haven’t already, you’ll want to get acquainted with the 5.4 PTR Patch Notes
, as this watercooler is going into detail on the changes listed. I’ll discuss the changes to each class, but here are a few things to keep in mind before we get started:
- The team made a lot of changes to glyphs that weren’t working out well or were just to provide some new choices. There were a lot of changes so to keep this blog manageable, I’m not going to go into them in detail.
- I’m not going to get into specific damage or healing tuning unless it requires extra explanation for some reason. In general, you can assume that your damage was too low or too high if you see generic buffs or nerfs to abilities.
- As I write this, Patch 5.4 has not shipped yet, so if there are changes yet to come, I’m obviously not going to be able to cover them!
One final caveat: please don’t consider the number of changes or the size of the paragraphs as an indication of the degree to which we love or hate your class. When players start weighing the number of words we use on each class, the end result is just to make us less likely to want to engage in discussions like this. Paladins and Hunters have a lot of words below because their changes are fairly complex and I felt they required more explanations than other classes.
With all that in mind, here we go!
Most of our core changes were to Blood DKs. We added the
passive to DKs (and Warriors) to allow them to increase their damage using the avoidance stats that naturally occur on their tanking gear. We also removed the Runic Power cost of Dancing Rune Weapon
to allow it to be used more as a cooldown when needed. Finally, we changed
to also be proc’d by avoidance, to make sure avoidance stats can help provide more active mitigation.
In terms of DK talents, as with all the classes, our primary goal was to open up some more talent choices and address some underpowered talents. While Anti-Magic Zone
was capable of mitigating a lot of damage when used at precisely the right time, it was also easy to not gain the full benefit of the cooldown. We changed it to just be damage reduction similar to Power Word: Barrier
. We also buffed Death Siphon
and Plague Leech
, which were underutilized.
We buffed Guardian spec Stamina outright because we felt the spec needed it. Druids have less reliable damage reduction, which isn’t a problem itself, so long as they have the health pool to compensate for it. For Feral Druids, we reverted an earlier nerf to Cyclone
’s cooldown but removed Cyclone
from Predatory Swiftness
. We felt like this ability had become a PvP problem that was too unpredictable to counter. We also lowered the PvP duration of
to make it less of a hard counter to stealthed players. We changed Innervate
to scale with Spirit so it would scale with gear better. We also made a few changes to the Restoration healing toolkit. Our change to Efflorescence
is an attempt to get Swiftmend
back to its original design of being a potent single-target heal. We’re making that change slowly by having the 5.4 Efflorescence glyph move the ground heal from Swiftmend
s. We further tweaked Efflorescence
s for Restoration by limiting it to a single mushroom.
We had several specs whose passive damage reduction we felt was starting to cause balance problems, and Balance’s Moonkin Form
was among these. Originally we converted the physical damage reduction to spell damage reduction at a time when the spec needed a PvP buff, but we felt like the fact that Balance required less healing in PvE was causing raids to favor them over other casters.
Most of the changes to Druid talents were because of the situation where certain specs leaned heavily on certain talents in a row and tended to ignore the others—
, Nature's Vigil
, and Soul of the Forest
fall into this category. We changed
to be a core Restoration ability because we felt the healers were dependent upon it. We replaced it with a new passive talent, Ysera's Gift
We had three main issues we wanted to resolve for Hunters this patch:
, Readiness, and Stampede
The problem with
is it’s just too powerful as a talent. No self-respecting Hunter is going to choose any talent but
, especially in PvP, but in PvE as well. At the same time, we think Hunters, especially Beast Mastery, have proven really powerful in PvP this expansion—and frankly the game has too many blanket silences already, so we didn’t want to just give
to all Hunters. Our decision was to give
to Marksman, who needed a PvP buff anyway. We did give all Hunters a base interrupt with the new Counter Shot
, which can interrupt a spell being cast, but can’t pre-silence a caster.
Readiness has been another problematic ability for us to solve. It started as a talent, but was too good compared to the other choices. We ended up giving Readiness to all Hunters, but we were never happy with this implementation. For PvP, it made Hunters really difficult to balance with the ability to reset so many offensive and defensive abilities. For PvE, we felt like Readiness wasn’t doing anything interesting besides making the opening attack sequence twice as complicated as it needed to be. If Readiness was an ability that a clever Hunter would bust out at a clutch moment, then you could argue it would be an interesting ability, but of course it was never actually used that way. Given that Hunters already have a lot of cooldowns, situational abilities, and just buttons in general, we didn’t think Readiness was worth saving. We did make small buffs to Deterrence
and we will make sure DPS is where it needs to be (which, if I had to guess, will be relatively higher than it was in Patch 5.3).
We felt like Stampede
had been nerfed too much in PvE for PvP reasons. It’s the intent that Stampede
is a potent cooldown for Hunters, so we increased its damage back to an exciting level and just had it do less damage in PvP. Finally, we made a change that some (though to be fair not all) Survival Hunters had asked for a long time, to remove Explosive Trap
(and therefore from the shared cooldown with
) so that they could use Explosive Trap
(typically with the knockback glyph) without interfering with
Hunter talents were in a pretty good spot overall after a bit of iteration throughout MoP. We did nerf Iron Hawk
, because as with Shadow Priests and Balance Druids, we felt like the damage reduction was just too potent. We changed Narrow Escape
to be dispellable as a PvP nerf. We buffed
, which has felt particularly weak, and we buffed A Murder of Crows
We didn’t think Mages needed many changes overall. One goal of ours was to fix Frost Mastery, which we felt was increasingly making them difficult to tune in PvP. When you take a spec with high burst and good control and then give them Mastery that links the two together, you’re bound to have trouble. The new Mastery, called Mastery: Icicles
, spreads the burst out over a few seconds but in a novel way that doesn’t feel too similar to, say, Mastery: Ignite
. As a consequence, we removed the mechanic where Frostbolt
buffs other spells and just let those spells do higher damage to begin with. As of this writing, we haven’t made any mechanical changes to Fire, though if you saw any earlier patch notes, we experimented with a few. Long term, we still have concerns about the interplay between
, and we still have concerns about Fire scaling with Critical Strike, but we’re not convinced either of these problems is so egregious that it needs to be fixed in Patch 5.4.
In terms of talents, we increased the damage of the bomb tier, and buffed Ice Floes
, which were underutilized. Long term, we still think the level-90 tier needs some work, but we were concerned that attempting to replace two or three of these talents for Patch 5.4 would lead to a lot more iteration before they felt good again.
But we did ease the restrictions on Rune of Power
. Since it didn’t feel like a crisis, we think more dramatic change is better saved for an expansion, when we have more time to consider new visuals, test, and get feedback.
Earlier in the PTR cycle, we attempted to solve Mistweaver mana problems, by which I mean Mistweavers didn’t have to stack Spirit to the extent other healers did. We became concerned that it would take quite a number of changes to solve this problem, resulting in a lot of relearning for players, so we agreed to accept the current status quo and balance Mistweavers around the fact that they have less Spirit (and therefore more of other secondary stats) compared to other healers. For Windwalkers, we redesigned their Mastery (yet again) and buffed the Storm, Earth, and Fire
ability so that the cleave aspect of the cooldown would feel really potent.
Brewmasters were in pretty good shape, but we did feel like Keg Smash
made their damage too high compared to other tanks.
Several Monk talents were underpowered or difficult to use. We buffed Black Ox Brew
, Zen Pulse
, Power Strikes
, and Healing Elixir
. We made Xuen easier to control by providing him with a true pet bar. We changed Chi Burst
to require no target. Ring of Peace
has a better visual that’s more representative of the affected area, and we tweaked its numbers. We didn’t think we could solve Rushing Jade Wind
’s problems, namely that it was too similar to Chi Burst
, so we redesigned it as a replacement for Spinning Crane Kick
Most of our core changes were to Holy Paladins. Specifically, we felt like the healing style of blanketing a group with
to proc absorbs from
had become too widespread, and didn’t fit the healing style we wanted for Paladins. When some Paladins choose to use a heal-over-time spell, that’s the kind of diversity we want the talent tree to deliver. However, when every Paladin uses heal over time spells, then it’s just the way Paladins work.
In this case, we didn’t want to make
a core ability. Unlike Restoration Druids, Mistweaver Monks, and to a lesser extent Holy Priests, we don’t think every Paladin should be extensively using heal-over-time spells. The only change we made was to have
’s periodic heal not proc the
absorb. It can still be used the way any HoT is used, to provide a buffer that can be layered with other heals, but it will no longer be an efficient way to blanket many targets with long-lasting absorbs. We made significant changes to
and Selfless Healer
as well to help offset this nerf, which I’ll discuss below.
We also felt like
had a design problem. When Holy Paladins could melee targets, they had no mana problems at all, but when they had to heal at range (which we wanted to at least be an option, if not the default healing style), they suffered for mana. We thought it best to unshackle Holy Paladins from the pressure of having to stand in melee, so we removed the mana return from
. We had to provide more mana to Protection Paladins as well, since they typically tank using
. We also buffed
for Holy, because we felt like it was too focused on single-target healing at a time when most healers are required to do a lot of area healing. Finally, we allowed the
passive to also affect Holy Shock
, with the intent of making Haste slightly better for Holy.
Aside from the mana buff mentioned above, we changed Crusader Strike
to also provide Weakened Blows as a quality of life improvement. We also changed Grand Crusader
to proc from avoidance instead of attacks. We aren’t trying to make Dodge or Parry a Paladin’s best tanking stats, but those stats are going to appear on tanking gear, so we want to make sure they can tie into active mitigation. We’re happy with how active mitigation is working out overall and we want to put even more emphasis on it, for all tanks, in the future.
Retribution didn’t need much attention overall in our minds, but we did make Inquisition
require less maintenance overall, and we reduced the cooldown and strength of Guardian of Ancient Kings, so that Retribution could use the cooldown more frequently. Retribution players had concerns that their damage was too tied to long cooldowns.
For Paladin talents, we tried to improve Selfless Healer
for Holy by allowing Judgment to provide Holy Power, and we tried to improve
by removing the target cap and having a cooldown and charges instead. We’ve come to the conclusion that the same talent choices for healers and other specs just won’t always work, which is why you increasingly see talents work slightly different for healers in the case of Paladins, Priests, and Druids in particular. We replaced Burden of Guilt as a talent (and made it a slightly weaker glyph) with the ability to use Turn Evil on players. We hope that this will open up some PvP comp choices for Paladins, since fears are not always easy to get. We buffed
so that it will still have some benefit on boss abilities that can’t be prevented from immunities. We buffed Sanctified Wrath
for Holy and Protection. We also simplified Unbreakable Spirit
by just having it apply the cooldown all the time rather than having to be “driven down.”
We made a few changes to each Priest specialization. For Discipline, while we don’t think the shields, Atonement
, or potent cooldowns are overpowered individually, we felt that having all three in one spec gave them too much versatility compared to the other healers. We know Atonement
is fun, and while it’s useful, it doesn’t truly compete with a dedicated healer or DPS, so we think it’s in a good spot. We like Discipline to have a strong shield kit, and we’re happy to see more Power Word: Shield
use in 5.2 than in 5.0. That left the cooldowns as targets, and we felt like Spirit Shell
was the right one to nerf. We also felt like Rapture
scaling so well with Spirit turned what was originally supposed to be a mechanic offering a discount on Power Word: Shield
to reward smart use had instead become a bona fide mana regeneration mechanic. For Holy, we buffed
and Holy Words
to help keep Holy in mana a little longer. We also finally gave in and reversed Lightwell
and Lightspring, so that the more passive (and popular) totem-like use of the spell became the default. We reduced the damage reduction of Shadowform as we did for Hunters and Balance Druids, because we felt like it offered too much of a passive benefit. We also completely rebuilt Shadowy Apparitions
so that it keeps the kit of a delayed source of damage, but won’t get stuck on terrain or objects or struggle with bosses that move or fly.
For Priest talents, we are trying one more time to make Angelic Feather
compete with the other movement talents, and let Mindbender
and Surge of Light
compete better with Solace and Insanity. We also buffed
for Discipline and Twist of Fate
for all three specs.
We buffed Evasion
for Rogues outright to increase their survivability against melee. We also buffed
because we agree that it went from being a potent heal in Cataclysm to a more middling one in MoP. The other core changes were really focused on Combat. We increased the cost and damage of Sinister Strike
in order to help address Rogues having too much energy with high Haste values, which led to a spammy, RSI-inducing style. (It’s fine for Combat to feel fast-paced, but it had gotten out of control.) To offset the loss of combo points from less frequent Sinister Strike
s, we added (or returned, depending on your point of view) the Ruthlessness
passive. The other Combat change was to redesign the targeting system of Killing Spree
. Killing Spree
now hits a single target if used without Blade Flurry
. During Blade Flurry
, it continues to work as it does on live.
For talents, we were worried in previous patches that we had overbuffed Burst of Speed
, but clearly that is not yet the case, so we lowered its cost. We buffed the numbers on Cheat Death
, and Shadowstep
, since they were seeing less use than other talents.
We’re happy with Enhancement and Elemental performance overall, but we are keeping an eye on their burst in PvP. We did feel that Restoration wasn’t really delivering on healing when grouped and stationary, which should be a Shaman’s strength. To remedy that, we changed Chain Heal
to no longer decrease effectiveness per jump and buffed Healing Rain
’s radius and healing. We also gave Healing Tide Totem
to all Shaman, because we felt like Restoration was never going to choose another talent so long as they had the option to choose a group healing cooldown.
With Shaman talents, we still had the problem that some talents were only attractive to some specializations. We buffed Astral Shift
, Stone Bulwark Totem
for this reason. We had to replace Healing Tide Totem, so we introduced Rushing Streams
, which allows Healing Stream Totem to heal two targets at once. We redesigned Conductivity to make it more useful for all Shaman by having it increase the duration of Healing Rain
, saving GCDs (global cooldowns). We were convinced that Totemic Restoration was just a problematic talent. It saw little use in PvE, but was really annoying to deal with in PvP, since a Shaman could just drop totems for a split second and benefit from their effects. The replacement talent, Totemic Persistence
, allows Shaman to summon a second Water, Earth, or Air totem without destroying the first one. (Including Fire Totems would have been a non-trivial DPS increase, making this the only talent attractive to Elemental or Enhancement.)
Because Warlocks had so many changes coming into Mists of Pandaria, we had to iterate on a lot of these new mechanics throughout the expansion and tried to not change things too much for Patch 5.4. We did nerf Fel Armor
for the same reasons we lowered the passive damage reduction of Balance Druids, Shadow Priests, and Hunters. We increased
’s damage but removed its DoT extension, which had become a liability for
in PvE rather than letting it provide damage on the move as it was intended. For Affliction, we shifted more damage to DoTs and away from
is a cool spell, and we like channeling it to increase DoT damage, but we felt like too much damage had been shifted away from DoTs, which hurt Affliction in PvP and made all Affliction locks have difficulty moving. We also simplified the Soul Swap
mechanic slightly by removing the glyph and the initial damage component. For Destruction, we changed Rain of Fire
to not be so essential to the single-target rotation by reducing its Ember generation. We made Howl of Terror
baseline for all Warlocks because we thought they had lost too much of their fear-based control package.
Our biggest challenge with Warlock talents was the level-90 tier, where most Warlocks chose
(often to the frustration of raid leaders forced to deal with slow-moving Warlocks). We removed the snare from
but reduced the number of spells it affected. The talent should help with movement but not guarantee characters never need to stop moving. We buffed
by allowing it to increase the damage as well as radius of area-effect spells, but on a cooldown so the Warlock has more control over the effect and so it’ll feel more potent when active. We decided we could not safely balance Archimonde’s Vengeance in its “damage reflection” kit, so we redesigned and renamed the talent to emphasize the Dark Soul cooldowns (with the intent that one talent needed to be attractive in fights with no movement or area damage).
In addition to these substantial changes to the level-90 row, we wanted to buff several unattractive talents.
never had a strong niche, so we made it improve
rather than act as an area-effect
. We replaced Howl of Terror
(now baseline) with a talent inspired by the Cataclysm version of Shadowflame that many players asked for.
is a cone-based snare. Soul Link
was also returned to an older design which is easier to use and more powerful.
Most of our Warrior changes were to increase Protection Warrior damage, try to improve the area-effect and cleave damage of Arms, or to address PvP quality-of-life issues. Buffing
helped both Arms and Protection. Changing
for Protection Warriors provided both more damage and active mitigation by making Critical Strike a more valuable stat. We don’t expect Prot to heavily stack crit, but they’re going to have some anyway just from core stats and group buffs, and it does open up the possibility of using crit gear.
We also gave Protection the same
ability we gave to Death Knights, to make sure the avoidance stats that do show up on tank gear provide some active mitigation.
For Arms, we reduced the cost of Thunder Clap
, and gave Slam a cleave ability. We didn’t think Fury needed much attention overall, though we did change the animation system to allow them to dual-wield polearms. For the PvP changes, we reduced the cost of
, gave the Safeguard root break to
baseline, took Hamstring off the global cooldown, and reduced the cooldown of swapping stances. We also made a change to allow Shield Wall
and Spell Reflection
to not require a shield to be equipped, but to still provide the visual cue that a Warrior was using these abilities by having a shield visually appear even if one was not technically equipped.
For Warrior talent changes, we buffed Bladestorm
, Storm Bolt
, and Warbringer
in addition to the
change mentioned above to open some more choices for Warriors, especially in PvP. We understand that Shockwave
is hard to live without in PvP, but we didn’t think that was easily fixable without nerfing Shockwave
—and probably speaks more for to the relative strength of stuns compared to other forms of crowd control, which is something we’d like to address long term. (Perhaps stuns need to be shorter, dispellable, or break on damage so they aren’t quite so game-changing.)
So there you have it. Hopefully you now have a little more insight into why we’re making the changes we’re making. I know it can feel like the developers are out to punish players sometimes, but really our primary goal is to keep everyone having fun. That includes redesigning awkward mechanics and buffing weak spells, but it also means nerfing overpowered mechanics so that other classes or specs don’t feel as if they can’t compete. We made a lot of talent and glyph changes because we want to deliver on the promise that players have a lot of choices in both of those systems. I won’t pretend that we’ll get everything perfect this time around either, but we should be closer. Please keep providing us with specific, constructive feedback. We still have time to respond to feedback, even after the patch launches.
As far as watercoolers go, I will personally try to get back to releasing them more frequently than I have over the past few months. No promises, but the one I am thinking of now is my own personal preemptive post-mortem of Mists of Pandaria. We’re very proud of the expansion overall, but we’re also our own harshest critics.Greg “Ghostcrawler” Street is one of the lead designers on World of Warcraft. He has been bitten by, but not limited to, the following animals: Atlantic brief squid, Roseate spoonbill, Texas ribbon snake, coatimundi. Follow him on Twitter at @Ghostcrawler.