What the final Forever notes mean for druid PvP

Blizzard posted the final Forever development notes on Monday night US time, two days before launch, and if you play a druid you’ve probably read them looking for your class and come away unsure how to feel. That’s a fair reaction, because the notes do two things at once, reworking all three druid specialisations and raising the damage of most of the classes you’ll be fighting. This post reads them the way a druid PvP player needs to, spec by spec with equal weight, then works through what the other eight classes got that you’ll be on the receiving end of.

Whilst I’ve played druids for years across both factions, nobody has fought at level 60 on Forever yet, me included, so treat this as a starting position that I’ll correct in print once the fights start. My short version is that Feral wins, Restoration pays, and Balance gets a bigger Starfire that I still don’t believe belongs in a PvP talent build, which gets three sections of its own because I expect it to be the most argued. The changes come from Kaivax’s post on the official forums, with Icy Veins’ copy of the full notes and Steven Mills’ write-up at outputlag for the class sections, and anything that comes from the game client rather than Blizzard’s notes is labelled where it appears.

Shorter fights for everyone

The change that matters most to a druid isn’t in the druid section at all. Blizzard’s developer note explains that holding each class to its Classic power budget meant every new ability had to be paid for by weakening something else, so over the last two weeks of the beta their answer was to raise damage on classes across the game and give higher level creatures more health to make up for it. That’s creatures, not players (the only player stat change I can find is a correction to each race’s base stats), so the increase that tougher mobs absorb in the open world lands on you in full when the attacker is another player.

Essentially, time to kill (how long a player lasts from full health under pressure) goes down across the board, and that changes what wins. The traditional druid PvP game is attrition, with HoTs (heals over time) ticking, Roots and forms buying distance, and the fight going to whoever still has mana in the third minute, and shorter fights cut that third minute off. They reward damage that arrives quickly and punish anything that needs a long, uninterrupted cast, and that one shift explains most of what follows. Notably, Blizzard doesn’t plan big tuning changes before the raids open on 9 December, outside of bugs and serious outliers, so this is the balance you’ll be fighting in for at least five weeks.

Feral’s gains

Feral got the largest set of buffs in the druid section, and nearly all of it is damage. Shred goes from 155% to 200% weapon damage, Claw from 110% to 140% and Ravage from 350% to 435%, whilst Rip, Rake, Ferocious Bite and Pounce gain about 25% on both their base damage and their attack power scaling. In addition, a new passive at level 32 called Poised takes 10 Energy off your next Energy ability whenever you’re out of combat at full Energy, and Rend and Tear’s bonus rises to 15% at 5/5 (all five ranks, up from 10%).

The way I read it, and this is my opinion, the line that matters most for PvP is Claw, not Shred. Shred needs you behind your target, which against a player mostly happens whilst they’re stunned, rooted or busy with someone else, so the strike you’ll throw most often is Claw, and it’s now 27% stronger. The ratio between the two barely moved (Shred was 1.41 times Claw and is now 1.43 times), so getting behind someone is worth what it always was, and you simply hit harder from the front. Likewise, the bleed increase is bigger than it looks against plate, because bleeds ignore armour as they did in Classic, so a Warrior who shrugs off part of every Claw takes the full extra from Rip and Rake.

Poised is an opener talent in all but name, the opener being the first thing you hit someone with from stealth. For example, in Prowl (Cat Form’s stealth) you’re out of combat and usually at full Energy, so your Pounce or Ravage costs 10 less. Energy comes back at 20 every two seconds, which is Classic’s rate and one I’m assuming carries over until I’ve seen the launch client, so Poised is a free second of Energy at the moment you most want one. Add a Ravage that hits 24% harder and a Pounce about 25% stronger, and the stealth opener is the largest single gain anywhere in the druid section.

Feral’s losses

That isn’t to say Feral only gained. Berserk is gone, with Primal Bite taking its place on row 7 behind Leader of the Pack, which leaves you no burst cooldown to save for a kill window (the few seconds where your target can’t escape or be healed). My earlier talent post repeated, I believe wrongly, a report that Berserk went to every specialisation, and since the notes only list it as removed I’ve added a correction there. Shifting Power, the talent that turns mana into 40 Energy, has also moved to row 7, behind a new row 4 talent called Predatory Discipline, which makes it a deep Feral talent rather than something a hybrid (a build split across trees) can pick up on the way. The client also shows its cooldown at 8 seconds, down from 16, which is datamined rather than in the notes.

Predatory Discipline is a quiet buff of its own, because Ferocious Bite keeps the extra damage from your spare Energy without spending that Energy. Lacerate is now a Feral talent instead of something every druid learns, and Bear Form’s threat bonus going from 50% to 115% is a raid change that does nothing for you in a battleground. All of this to say, Feral is the specialisation shorter fights suit best, mainly because its damage arrives in seconds rather than over the length of a DoT (damage over time spell). So were you weighing Feral against a caster build for world PvP? If you were, the notes have pushed you towards Feral.

A bigger Starfire

Balance has fewer lines, and one of them is large. Starfire’s base damage goes up by about 40%, and Genesis now adds 2/4/6/8/10% to your periodic damage and healing (from 1 to 5%). Moonkin Form’s aura goes back to raising every kind of critical strike chance for your group, but it no longer stacks with Leader of the Pack, so a premade (a group that queues together) running a Moonkin and a Feral now gets one 3% aura, not two.

There’s also a change for every class that a druid should notice, which is that various talents no longer lose their chance to proc (trigger) when you cast a low rank of a spell. The one exception the notes name is the Mage spell Blizzard, whose low ranks keep a reduced chance to trigger Arcane Concentration, Frostbite and Fingers of Frost. The notes don’t list which talents are covered, and Forever still penalises the spell power on low ranks, so my advice is not to rebuild your bars around rank 1 spells until someone has tested which druid procs it reaches.

Genesis is the quieter change, and for PvP the more awkward one. A 10% harder tick on Moonfire and Insect Swarm is welcome, but a DoT pays out across its whole duration, and the shorter fights get, the more of it is still ticking after the fight has been decided. As I see it, that makes it a good talent the rest of the notes have made slightly less valuable, not a reason to re-plan.

Starfire’s cast problem

In my last talent post I argued that Starfire needed a better reason to exist in PvP, and on the surface a 40% increase is exactly that. So does a bigger Starfire earn its place now? I don’t believe it does, and since I expect a lot of Balance players to spend points here that they don’t need to, I want to argue it properly. The short version is that the buff changes how hard Starfire hits, and Starfire’s PvP problem was never how hard it hits, it was whether the cast finishes.

Starfire is still a 3.5 second cast. Improved Starfire takes 0.1 seconds off per rank, so all five ranks get you to 3.0 seconds, plus a 3% chance per rank to stun the target for 3 seconds (15% at 5/5). Wrath with Improved Wrath at 5/5 is a 1.5 second cast at half the mana, which is why it’s the spell a PvP Balance druid casts most. So for the same three seconds of standing still, would you rather have one Starfire or two Wraths? The answer depends on what each of them does when the other player gets a say.

Every cast is an offer to be interrupted, and a longer cast is a longer offer. A Kick, a Pummel or an Earth Shock can land anywhere in those three seconds, and one interrupt ends the Starfire along with every tenth of a second you put into it, whereas against two Wraths the same interrupt only takes one, and the other player then waits out their cooldown (10 seconds for Kick and Pummel, 6 for Earth Shock) before they can do it again. The notes make this worse, too, because Rogues now have Shank, a 5 Energy strike that hands back 15 Energy every 10 seconds, so the Energy for a Kick will be there far more often.

Pushback (the delay added to your cast each time you take a hit) makes the same point with numbers you can check in your head. Nature’s Focus at 5/5 gives you a 70% chance to ignore the pushback from each hit, and in Forever it covers Arcane spells as well as Nature, so it does protect Starfire, which sounds like a lot until you count the hits. For example, a Warrior who lands one hit during a Wrath will usually land two during a Starfire, so the Wrath goes through untouched 70% of the time and the Starfire 49% of the time (0.7 times 0.7). Against a dual-wielding Rogue landing four hits it’s 24%, and every pushback that lands makes the cast longer, which invites another hit, so twice the cast time doesn’t mean twice the risk, it means risk that stacks on itself, the way a debuff does.

The 40% in practice

The 40% also applies only to Starfire’s base damage, not to the spell damage you add from gear. The top rank on the pre-launch client was 350 to 412 (about 381 on average), and Starfire has historically gained 100% of your spell damage, as it did in Classic as a 3.5 second cast. Wear 300 spell damage, a placeholder until I’ve seen real PvP sets at 60, and an average hit goes from 681 to about 833, which is 22% rather than 40%. The more spell damage you wear, the smaller the buff gets, and spell damage is exactly what a Balance PvP set is built around.

So what about the moments when the other player can’t interrupt? The ones a druid can make on its own are short, and most of them end themselves. War Stomp (the Tauren racial, usable in forms) stuns for 2 seconds, so even if Starfire starts the instant the stun lands, the target gets the last second of your cast to interrupt it. Bash and Pounce both need Bear or Cat Form, and each spends a global cooldown (the short lockout after most abilities) of the stun before you can shift back and start casting, and Bash also needs Rage, which a build without Feral points won’t have on arrival now that Furor lives in Feral. Roots and Hibernate last far longer, but damage can break both, so the Starfire that lands is also what frees your target.

Improved Starfire’s stun doesn’t rescue it either, because it can only proc off a Starfire that has already landed, and at 15% it’s a bonus you can’t plan around. Eclipse, the new talent where your Wraths shorten your next Starfires, is the closest the tree comes to an answer, and it has the same flaw in a different place. The charges come from Wraths that finished, so Eclipse speeds up the Starfire you cast after the other player has already let you cast freely, which is the one situation where you didn’t need a faster Starfire.

Where Starfire still works

That brings me to points, and what they return. A 31/0/20 build has 31 points in Balance and every one of them is contested. Improved Wrath makes the spell you cast every fight faster and cheaper, whilst Nature’s Reach, Improved Entangling Roots, Overgrowth and Moonglow all change fights you’ll actually be in. Five points of Improved Starfire buy half a second on a spell you’ll cast in a handful of situations, and in every one of them the half second doesn’t matter, chiefly the following:

• The opener from range, against someone who hasn’t seen you yet. A melee player running at 7 yards a second needs about 3.6 seconds to close from 30 yards to melee range, and that’s after they’ve noticed you, so even an untalented 3.5 second Starfire lands first (with room to spare if Nature’s Reach has added to your range), and the 40% makes that first hit noticeably bigger. Nobody is trying to interrupt you, so the shorter cast buys nothing.
• A Hibernated target, since Hibernate works on enemy druids in animal forms and on Hunter pets (both count as Beasts), and it lasts far longer than any cast. Your Starfire wakes them, so it’s a one-cast opener on a target that can’t respond, and again it doesn’t need the shorter cast.
• The bait, because Starfire is Arcane and nearly everything else you cast is Nature, and an interrupt only locks the school (the magic type) of the spell it hits. A Kick on your Starfire locks Starfire and Moonfire but leaves Wrath, Roots and every heal available, so starting a Starfire to draw the interrupt, and cancelling if they don’t bite, is a legitimate play at 3.5 seconds or at 3.0.

So my position is to take the 40% as a free improvement to your opener and the Hibernate trick, keep Starfire on your bars for those, and not spend five points of a 31 point tree making it faster. If the launch client shows Starfire’s cast time coming down, or Eclipse turns out to be something you can build without casting into an interrupt, I’ll revisit this here.

The cost to Restoration

Restoration is where the notes cost druids the most, and most of it was already in the hotfix cache last weekend, so if you read my post on the datamine it’s now official rather than new. Wild Growth has a 3 second cast (2.5 with Naturalist at 5/5), Bloom replaces Gift of the Earthmother, and Swiftmend needs a Rejuvenation or Regrowth that you put on the target yourself. Naturalist adds 2 to 10% to all of your damage, up from 1 to 5%, and Heart of the Wild and Furor have swapped trees.

The clearest way I’ve found to see the cost is to time the sequence you would use when a teammate is about to die. In the beta that was Rejuvenation, Swiftmend and an instant Wild Growth, and with Gift of the Earthmother taking half a second off each of their global cooldowns, all three were out in 3 seconds (those two beta values come from the client’s own tooltips, via ForeverChanges, because the notes only say what each one changed to). At launch the same three take 5.5 seconds, 1.5 for each of the first two and a 2.5 second cast for Wild Growth, which can be interrupted or pushed back like any other cast. That’s nearly twice as long to deliver the same healing whilst everyone hitting your teammate hits harder, and it’s the clearest case in the notes of a specialisation paying for everyone else’s buffs.

Notably, Wild Growth now has Starfire’s problem. Everything above about interrupts and pushback applies to a 2.5 second heal as much as to a 3 second nuke, except that a lost Wild Growth usually means a lost teammate rather than a lost hit, and with that in mind Nature’s Focus becomes close to mandatory for a dedicated healer. How often does your team actually give you two and a half seconds to stand still? When they don’t, what are you casting instead?

Bloom and Swiftmend

Bloom cuts against shorter fights as well. It pays when a Rejuvenation or Regrowth runs its full duration (half when an enemy dispels it), so it rewards HoTs left alone for twelve seconds or more, and the faster fights are decided, the fewer of those finish. How often does one of yours get twelve seconds alone in a battleground? It does make offensive dispels slightly worse against you, since a Purge or a Dispel Magic on your HoT hands your target half a Bloom on the way out. Swiftmend needing your own HoT mostly matters in premades, where several druids have been able to keep Rejuvenation on one target since the 15 October notes, and a second druid can no longer spend yours.

The clear gain is Naturalist doubling, and it pays your damage rather than your healing, which tells you where Blizzard sees a Restoration hybrid’s value. All of this to say, a dedicated Restoration druid in battlegrounds should plan for less healing per second under pressure with more damage per second coming in, and in my view, which is only a view, that’s the reason to think hard before you commit a PvP character to deep Restoration this week, not after you’ve paid for the respec (resetting your talents, which costs gold).

Harder matchups

The other half of reading these notes as a druid is the damage coming the other way. A number of classes gained damage in the same pass, but three of them gained the most against you in particular. Which class did you lose to most in the beta, and is it one of them? In order of how much they worry me, they’re the following:

• Shamans, with Lightning Bolt’s base damage up on every rank (by 80% at the top) and Lightning Overload duplicating a spell 10/20/30% of the time where it was 3/7/10%. Call of Thunder also replaces Lava Burst, adding 5% critical strike chance to Lightning Bolt and Chain Lightning. The shocks are back at their Classic Era values, and Stormstrike now hits for 150% weapon damage on a 6 second cooldown (from 100% on 8). An Elemental Shaman was already the hardest caster for a Balance druid to out-trade, and with Earth Shock both an interrupt and a bigger hit, every long cast against them is a worse idea.
• Hunters, with Aimed Shot’s bonus damage up about 110%, Arcane Shot up about 30% with a higher attack power coefficient (0.15, from 0.1), and Serpent Sting about 20% stronger with more scaling per tick. An opening Aimed Shot on a casting druid is now a large share of your health, so expect to spend more time in forms against Hunters, not less. Serpent Sting is a poison as it was in Classic, so Abolish Poison still clears it, and all of their traps now share one 30 second cooldown, so a Freezing Trap on you puts the rest of their traps on cooldown with it.
• Rogues, with Shank, a new level 24 strike that costs 5 Energy, can’t miss and gives back 15 Energy on a 10 second cooldown (25 at rank 2), which leaves more Energy for Kick and Gouge. Deadly Poison and Instant Poison doubled their attack power scaling, and Rogues get Poised too, so their openers got cheaper exactly as yours did. Deadly Poison is the one you can answer, since Abolish Poison removes it, whilst Instant Poison’s damage is done the moment it lands.

The rest matter less to a druid, starting with Mages, who lost Frost Blast (the 40% slow they gained during the beta). Ice Lance now deals 300% more damage to frozen targets on an 8 second cooldown it shares with Fire Blast, so staying in a Frost Nova costs more than it did, and since forms remove roots, the answer is to shift out of it. Warriors mostly changed in ways that matter to tanking, and the Paladin rework is broad enough, with new talents in all three trees and Salvation gone, that it deserves a proper read of its own rather than a line here.

Easier matchups

It isn’t all bad news, and a few changes in the same notes work in your favour. Two of them reward the shapeshifting and HoT play a druid is built around, and the third makes one matchup a little less one-sided:

• Chained roots, because procced roots now have diminishing returns (DR, where each repeat of a control effect lasts less time), covering Frostbite, Entrapment, Improved Wing Clip, Counterattack and Improved Hamstring. You can shift out of all of them, but every shift costs mana and a global cooldown, so fewer chained roots is a direct saving if you want to stay in caster or Moonkin Form. Notably, Nature’s Grasp isn’t on Blizzard’s list, so don’t assume your own proc root changed until it’s been tested.
• Warlock pets, which now gain 2 health per point of their master’s Stamina, down from 7.5. That makes a Felhunter a realistic target, and it matters to a druid more than most, because a Felhunter’s Devour Magic is what strips your HoTs. A Feral druid can now take the thing that undoes your healing out of the fight quickly.
• Shadow Priests, whose Shadow Word: Death now costs them 30% of their maximum health when it fails to kill (from 10%). That turns their execute (a finishing spell meant for a target on low health) into a real risk to themselves against a druid with HoTs running, so keeping a Rejuvenation up as you drop low is worth more than it was.

Before you re-plan

So before you move a single point, ask yourself a few things. When you lost fights in the beta, was it because a cast didn’t finish, or because you ran out of health before your HoTs caught up? Do you play mostly in premades, where a Moonkin and a Feral now share one aura, or in random battlegrounds and world PvP, where you’re usually the only druid? If you heal, how often does your team give you three uninterrupted seconds? If your answers point at burst and short fights, the notes have made Feral the better home for you, and if they point at long fights, the hybrid still works, with less room for error than it had a week ago.

If you’ve weighed all this and you still want to play the druid that wins the long fight, Balance and Restoration both still do things no other class can, and I wish you every success with them. I’ll be testing all of it from the moment Restokins hits 60, starting with the Starfire windows and the three matchups above, and from then you can find me live on the stream weekdays from midday AEST, where I would love to hear how your version holds up.

For the mechanics behind interrupts, school lockouts and shapeshifting out of roots, read the PvP mechanics guide.