Repost from EU forums on behalf of Malon: (https://eu.forums.blizzard.com/en/wow/t/feedback-mage-class-changes/142106/3)
A little background: I’ve played a Mage since Classic, been writing guides about the class on Wowhead since MoP and am a moderator over at Altered Time, the Mage Discord server.
Here I’ll begin by going over some of the class-wide changes and my thoughts on it, then move on to my notes and a wishlist for each spec. This post is extremely long, but that’s because I am passionate about both this class and the game. I’ve added a TLDR summary to each section for easier reading.
- Alter Time is not an adequate replacement for Displacement for Arcane
- Mirror Image should be a utility cooldown, not a DPS one
- The loss of Shimmer’s second charge should be compensated with a lower cooldown
- Focus Magic overlaps with the role of Incanter’s Flow and is only suitable for Frost.
Great to have it back; I am, perhaps, in the minority as I always preferred the WoD version (a simple mobility/survivability spell) over the MoP one (a damage cooldown). However, taking Displacement away from Arcane to give Alter Time back to all specs, at the same time as Shimmer is reduced to one charge, is a significant hit to Mage mobility, and Arcane’s in particular. The spell should really be tweaked for that spec, at least, to refund Blink/Shimmer’s cooldown - they are the masters of time and space, after all.
As glad as I am to see the spell return to the base kit (it was never a competitive choice against the other options), MI on the alpha is not an interesting button to press; it’s simply ‘Do a bit more passive damage for 40 seconds’. By the end of Mists - the last time the spell was a baseline ability, rather than a talent - Mages were using this as a threat-drop or tool to help with soloing content, but it wasn’t a regular part of a damage rotation; if the spell remains as-is (uninteractive, simply casting Fireball/Frostbolt/Arcane Blast), this may be the route Blizzard should pursue for it.
Simply a case of the unpruning going too far, these are not iconic abilities and I fail to see why they’re making a return. A tuned (for example raid-wide) version of Dampen and/or Amplify Magic would be much more welcome.
It certainly needed a nerf; the spell has been a must-pick for all specs since day one of Legion. However, dropping it to one charge while retaining the five-second increased cooldown on Blink is just…bad, to the point where picking the talent on the alpha genuinely feels like you’re choosing to downgrade Blink. Considering the insane mobility of other specs, especially melee classes, as well as Blink’s ongoing bugginess (failing at terrain changes like slopes, stairs and doorways), we will feel this in both PvE and PvP.
This spell fills too similar a niche to Incanter’s Flow, favouring a flat damage curve. That immediately makes it near-useless for Arcane and Fire, both of which are cooldown-based and simply won’t sacrifice Rune of Power to take a minor crit buff (which Fire doesn’t care about, anyway). Frost, of course, takes Incanter’s Flow and it’s hard to see Focus Magic out-performing that unless the numbers are dramatically tweaked. This is the same problem Mirror Image had, except it encroached on Rune’s territory instead of Incanter’s.
To differentiate it, the Mirror Image/Focus Magic talent spot should be significantly different to ‘short cooldown burst spell’ or ‘passive damage increase’, such as a burst spell with a long cooldown.
- Only the Kyrian Covenant works well in all content for all specs
- Radiant Spark has issues with multiple projectile spells and travel times
- Mirrors of Torment is unusable in PvE, but Venthyr have the best Signature ability (a teleport)
- Deathborne is powerful for Arcane but loses out for Fire and Frost
- Shifting Power is immobile, provides little cooldown reduction and needs to be cast too frequently, taking away from the class rotation.
Note that all of the below said with the realisation that we haven’t seen the legendary items, Soulbinds, etc., but weaknesses need picking up early on and I don’t believe that using one system to fix the flaws of another is healthy.
This may be grass-is-always-greener syndrome, but Mages really seem to have been given the short end of the stick when it comes to Covenants, with low damage, awkward or flat-out pointless abilities. Preach said it best, in his video about the Night Fae: “Poor Mages - every time we get to them, I feel bad.” Let me explain:
Radiant Spark (Kyrian)
This is a flat damage increase, and is probably the one most players will end up taking as it just works in all situations (which makes it unfortunate that the Kyrian Signature Ability is by far the weakest compared to the shields, movement speed increases and teleports of the other Covenants). The low cooldown is great and the damage increase is of course useful. The issue comes with spells that are treated as separate hits: things like Flurry and Arcane Missiles consume multiple stacks, and spells with different travel times (Ice Lance travels faster than Glacial Spike; Fire Blast is instant and hits before Pyroblast) also interfere with the debuff. That makes it awkward to play ‘correctly’ and line up your biggest damage spell with the fourth stack.
Mirrors of Torment (Venthyr)
This spell will almost never get its full value in PvE, as NPCs rarely cast frequently enough to consume all three stacks, plus you generally want to be interrupting their spells anyway. The cast speed slow effect, while useful, makes proccing the root + silence at the end even less likely. In addition it appears to just simply not work on most PvE bosses, and in PvP begs to be dispelled.
Another contender for the default option, this ability has the advantage of having a use in cleave situations, although the cooldown is much longer than Radiant Spark. In addition, Fireball and Frostbolt don’t deal a huge amount of damage anyway, so the Necrolord Covenant will be less attractive for these specs than it is for Arcane - which, again, pushes all Mages towards Kyrian.
Shifting Power (Night Fae)
Aesthetically this is by far the best-looking Covenant ability, but that’s where the attraction ends. Being forced into melee range is traditionally unhealthy for casters (we’re treated as valid targets for ranged mechanics no matter where we stand), and standing still for six seconds while the spell channels is even worse. Coupled with the fact that it only ticks every 1.5 seconds (so four instances of damage/cooldown reduction), and the cost of interrupting the spell becomes even worse. On top of all that, it’s only three seconds of CDR with each tick, so just 12 seconds per full channel. With a 45-second base cooldown, that is a LOT of channelling to make Shifting Power worthwhile.
- Arcane has between nine and 15 seconds of set-up time before dealing damage; cooldowns should be removed/taken off the global cooldown/combined to address this
- The spec lacks cleave, but Touch of the Magi could fill this niche with appropriate tuning
- Mobility is a big issue and Arcane needs an adequate replacement for Displacement, which Alter Time is not.
Arcane’s burn/conserve rotation has seen some changes going into Shadowlands, specifically with the doubling of Arcane Power’s cooldown and the addition of Touch of the Magi as a new base spell. This means one big burn every three minutes, and smaller burn phases every 45 seconds. It also reduces the power of the Overpowered talent, which is welcome. However, the spec still has the glaring weaknesses that have prevented it from being mainstream throughout BfA.
The biggest issue with Arcane on the alpha is its set-up time. The spec has always been slow to start, with the need to build four Arcane Charges before it can enter the burn phase - the Charged Up talent addresses this and is a must-pick in most raiding environments. However, even with Charged Up there are now fully six global cooldowns - nine seconds - of doing nothing except casting buffs before starting to deal damage, or around 15 seconds without Charged Up. That is: Charged Up/Arcane Blast x 4; Mirror Image; Covenant ability (if Kyrian or Necrolord); Touch of the Magi; Rune of Power; and Arcane Power. The last four of those are also only active for a short time, so some - maybe most - of the duration will be wasted while casting additional buffs.
Moving Mirror Image to be a utility cooldown instead of a DPS one, taking some spells off the global cooldown (Touch of the Magi and/or Arcane Power would be the top picks) or combining some (like Charged Up and Touch of the Magi, which have similar cooldowns of 40/45 seconds) would go some way towards alleviating this problem - but is at best a sticking plaster.
I can’t stress enough how bad it feels to play with all of these buffs and cooldowns on the alpha. The spec drags behind others, and most mobs are half dead by the time you can actually start doing damage.
Of the three Mage specs, Arcane has always held the crown for the highest theoretical single target damage (movement is a limiter, but that’s a later discussion). However, where it really falls down is in its cleave damage, meaning passive damage executed as part of the single target rotation. Because very few fights these days are pure single target, cleave is an important feature for every damage spec. Frost has Splitting Ice and Frozen Orb, while Fire spreads Ignite (passively for now, every time it hits Phoenix Flames in Shadowlands). The only cleave damage Arcane has is Arcane Barrage, and while it hits hard the need to cast it with four Arcane Charges massively limits the number it is possible to cast, even using Arcane Orb. Consuming your Charges in this way is also hugely disruptive to your single target rotation.
Touch of the Magi could be a fantastic way to give Arcane cleave potential using the spec’s strong single target burst, but the tuning is currently well off.
The issue is that the splash damage from Touch is far, far lower than the damage the main target takes, and falls even further with more targets (the formula is ‘100% damage divided by [square root of mob count]’, so 57% splash damage to three (total) targets, 50% to four targets, 44% to five targets, etc. This was confirmed with testing). It’s far from a reliable or high-damage cleave tool, especially considering its limitations in needing to be cast on a mob that will live for the full eight seconds to get the maximum benefit; although that might also be a hindrance if the pack needs to die quickly. Still, we’ll accept that as a limitation of the spell as it’s currently designed for now.
But…what if we don’t accept it? Let’s look at a possible redesign of Touch: still a debuff cast manually on a mob, but instead it acts more like Prophecy of Fear (https://www.wowhead.com/item=124230/prophecy-of-fear&bonus=1801) from WoD, exploding for damage each time the Mage hits the target. This design gives Arcane some on-demand ranged AoE and cleave, without needing to worry about the target or other mobs dying before the effect activates.
Of all the Mage specs, Arcane loses the most DPS from movement as all of its damage (barring the non-spammable Arcane Barrage) comes from hard-casts. Live Shimmer, with Displacement, partly counteracts this weakness over long distances, although micro-movement (moving out of a void zone, for example) is still painful. There’s a reason all the top Mage logs on Shad’har - the only pure single target fight in Ny’alotha - are Fire, not Arcane.
The nerfing of Shimmer to one charge and the (long overdue) reversion of Clearcasting to stack increases the viability of the Slipstream talent, which is welcome. However, this leaves Arcane with the opposite issue: long-range movement. Alter Time is not a valid replacement for Displacement; in fact, it feels like the other specs are gaining something while Arcane is losing power. Lowering Alter Time’s cooldown for Arcane, having it refund Blink’s cooldown, or both, would be very welcome - again, masters of time and space.
- Arcane needs a reliable ranged AoE ability to both remain safe in raids and avoid dropping mechanics on melee players.
There’s little to be said here; Arcane has always been forced into melee range to AoE; another element that has limited the spec in raids. While Blizzard, Frozen Orb and Flamestrike can be cast safely from range, Arcane needs to position to avoid melee mechanics while also being a valid target for ranged mechanics, which often means they are simply not allowed near the melee pile.
The addition of Touch of the Magi again goes some way towards addressing this, and gaining mana from Arcane Barrage (a Shadowlands levelling perk) is a great bonus; but for years Arcane has been punished for trying to AoE, and that doesn’t appear to be set to change in Shadowlands. If nothing else, the return of the larger range on Arcane Explosion that was present in Legion would be welcome.
- The new Enlightened talent emphasises remaining at high mana levels, which impacts talent choice in the rest of the tree.
- Dead talents like Mana Shield, Supernova and Nether Tempest remain.
Amplification’s change to fire two additional missiles, and Clearcasting stacking three times instead of once, will really increase the power of Arcane Missiles. Whether it will be enough to take away from Rule of Threes - a major conserve tool - remains to be seen. The new Enlightened talent makes conserving at high mana even more important, so it’s unlikely at current tuning.
Mana Shield’s buff is nice but the talent will still rarely be used in competitive PvE. Movement is just too important for the talent to compete with Slipstream or Shimmer.
On our 45 tier, Arcane is locked into Rune of Power. As mentioned above, Focus Magic and Incanter’s Flow do not fit Arcane’s burn-conserve playstyle; especially now that Touch of the Magi encourages a burn every 45 seconds - the exact same as Rune of Power’s cooldown.
Resonance remains a great AoE talent while Charged Up is the go-to for a single target, and Supernova is still underpowered. The latter has some utility in Mythic+ but is rarely, if ever, worth the opportunity cost of missing one of the other options on this row.
The 75 tier has little impact on PvE, but all three talents are usable in different situations.
The 90 tier has a new problem. In BfA, in common with the 60 tier, it forced a(nother) choice between single-target (Touch of the Magi) and AoE (Reverberate), with Nether Tempest as the trailing third option. Thankfully Enlightened, replacing Touch of the Magi, increases all damage done, so while Reverberate might be stronger in AoE, Arcane Mages won’t feel like they have a dead talent if called on to attack multiple targets. The issue is that, because Enlightened is strong and fits all situations, other more niche talents will just not be chosen. It also encourages playing at high (above 70%) mana unless burning, which plays poorly with ⅓ of Temporal Anomaly’s effects and any level 15 talent except Rule of Threes.
At the 100 tier, the increase to Arcane Power’s cooldown has nerfed Overpowered, which was really necessary; it was always the best choice. With that we may see more choice in this tier, which is very welcome. However, as mentioned above, with Enlightened (currently looking like the go-to choice) we want to remain above 70% mana. This means that any Evocation procs from Temporal Anomaly are going to be mostly wasted. Lowering the mana threshold on Enlightened would go some way towards addressing this, and also make Rule of Threes less important. That is a minor issue, though, and I welcome having some actual choice on these talents again.
- Significant reduction in set-up time/global cooldowns
- Touch of the Magi to more clearly fulfil the cleave role that damage specs need for viability in the current PvE end-game. As well as a damage buff, acting like the Prophecy of Fear trinket from Warlords of Draenor - which Arcane Missiles had a unique interaction with - would be one way of addressing this.
- Alter Time to have an Arcane-specific component.
- Some form of reliable ranged AoE. Touch of the Magi could fulfil this role with the above redesign.
Of course Fire in Shadowlands is going to feel very different to how it does at the end of BfA, where rental power has increased output to silly levels. The spec currently flows well but, much like Arcane, is designed around high peaks and low troughs. A single fluffed Combustion can be the difference between success and failure. Having lost much of that rental power (Memory of Lucid Dreams, Hyperthread Wristwraps, etc), this would be a good time to increase Fire’s damage at a base level.
- Fire has no effective burst AoE against low health enemies.
- Manual Ignite spread was tried and failed in Warlords of Draenor; limiting it to eight targets is enough.
AoE has, ironically, been a problem for Fire for several expansions. There are no easily spammable spells like Blizzard + Frozen Orb, or Arcane Explosion + Barrage. Rather, Fire’s AoE ability comes from spreading a big Ignite, but that requires having a single large target to build and spread the Ignite from; it only looks good now because we have the tools to create burst windows to build that massive Ignite. Without that, Fire doesn’t have the tools to cope in its base kit, and this is very obvious in low-keystone-level dungeons or when facing many small mobs.
Another issue, which Blizzard addressed while I was writing this post (thanks!), was in using Fireblast to spread Ignite. That functionality is now on Phoenix Flames, which is an excellent change. However, the spell should be changed so it does not interact with Hot Streak/Heating Up , otherwise we’ll face the same issue (using it during Combustion, rather than for spreading Ignite) that we did with Fireblast. An adjustment to its travel speed/spreading Ignite on cast rather than impact/being off the global cooldown would also be welcome. As it is, we lose a minimum of two Ignite ticks before Phoenix Flames hits and spreads the DoT.
In addition, manual Ignite spread enabled what came to be called ‘tick compression’ in WoD (see here ). How will this be avoided in Shadowlands?
- Fireblast is too important to the rotation to give up Flame On.
- Flame Patch still encourages hard-casting Flamestrike; perhaps it should only apply to those cast using Hot Streak.
- Conflagration is almost irrelevant.
- Kindling desynchs Combustion from other abilities, which plays against Fire’s cooldown-stacking playstyle.
The importance of Fireblast exacerbates the disparity in our level 60 talent tier; as it stands, no-one would consider taking anything except Flame On, although I’ll admit that we haven’t seen Phoenix Flames’ replacement yet (please god not Cinderstorm). Likewise, Rune of Power is the only competitive option in high-end PvE with the existing peak-and-trough style of play.
Further down the tree, the level 90 tier is built around AoE and, with tuning, could work; but at current levels the choice has almost no impact on a single-target encounter. Currently the damage increase from Conflagration is absolutely minimal on a single target: taking another option, even if you never cast Flamestrike/Living Bomb in a fight, is less than a 2% DPS loss. The talent needs some love.
Living Bomb could be a tool to handle Fire’s weak AoE when it can’t spread Ignite, but at current levels of tuning (without the high Mastery we have at the end of BfA) it’s not worth casting on an average-sized (4-5) pack of mobs, and until late 8.3 everyone has taken Flame Patch for dungeons. At the moment, though, Living Bomb has actually been nerfed going into Shadowlands (it deals reduced damage to secondary targets), while Flame Patch has not.
Flame Patch is a problematic talent, in that it encourages what I would consider degenerate gameplay; namely, ignoring your entire rotation to hard-cast Flamestrike against anything more than two mobs in a dungeon setting. Ideally the talent should only encourage you to spend your Hot Streaks on Flamestrike, so a simple fix would be to make it only apply to instant-cast Flamestrikes - perhaps with a small buff to compensate.
At the very end of the tree, the level 100 tier currently looks pretty balanced. Both Pyroclasm and Kindling have been buffed to compete with Meteor and that is appreciated (whether the hassle of playing Pyroclasm will be worth it in a real-world setting is a different story). That said, Kindling still has two issues: first, that it lowers the cooldown of Combustion but nothing else, which makes stacking cooldowns like Covenant abilities difficult (again, the current peak-and-trough playstyle means desynching cooldowns is not desirable); and second, that it works with all of our rotational abilities - even Phoenix Flames - except Scorch, which makes it play against Searing Touch.
- New/different ability to spread Ignite, removing some of the rotational weight from Fireblast - FIXED WHILE WRITING THIS POST, PRAISE THE LORD
- Flamepatch to not encourage hard-casting Flamestrike - this is not simply a BfA issue, but something seen in Legion as well.
- Smoothing rotation by moving damage out of the burst window
- Flurry and Winter’s Chill have issues with travel time and do not allow the Shatter combo to work effectively at either maximum or minimum ranges.
- Fingers of Frost is still ‘munched’ (useless) with a Brain Freeze proc.
- Frost is still stuck spamming Frostbolt without procs, and may once again drop Ice Lance from the rotation at high Mastery levels; a proposed redesign addresses both of these issues and is linked below.
- Frozen Orb remains buggy when used in melee range.
Frost has had few changes going into Shadowlands; mechanically, the spec plays well and has a tool for most situations. Its major issue in BfA has been the degenerate gameplay of the No Ice Lance build (which will not be a thing in Shadowlands) and the range issue of the Winter’s Chill debuff.
In BfA, Brain Freeze makes Flurry apply Winter’s Chill: a very short buff allowing the next spell to act as if the target were frozen. The most common way to consume Brain Freeze is to cast a Frostbolt/Glacial Spike/Ebonbolt (Spell #1) and queue an instant Flurry immediately afterwards, followed by an Ice Lance; because of Flurry’s travel time, and the fact that it hits three times, this enables Spell #1, as well as the Ice Lance, to benefit from Winter’s Chill (‘the Shatter combo’). However, because of spell travel times standing in melee range means that - in the situation above - Spell #1 will not benefit from Winter’s Chill, as it hits before Flurry.
Winter’s Chill has had a slight redesign in Shadowlands: it now lasts six seconds and affects the next two spells to hit, enabling the double Ice Lance gameplay that first appeared in Legion. However, the melee range issue is still present, and the combo now has the same issue at maximum range - especially on large bosses like Stormwall Blockade, Immerseus and Lord Marrowgar. In these situations the third projectile from Flurry may hit before Spell #1, causing that spell to consume a stack of Winter’s Chill.
The Shatter combo is a massive part of Frost’s damage, and the inability to stand at either maximum or minimum range will have a serious impact on the spec’s DPS on fights where standing elsewhere is impossible. There have been many suggestions to change how the combo operates, most recently and -comprehensively from Norrinir and other Mages of Altered Time: https://vituscze.github.io/frostmage/ . While these are significant changes, they address some core weaknesses of Frost by providing bad luck protection, future-proofing against No Ice Lance returning at high Mastery levels and emphasising the Shatter combo.
A related issue with Frost’s procs is that Brain Freeze renders Fingers of Frost - both of which interact with Ice Lance - useless and results in munching the FoF proc (because the target is already considered Frozen due to Winter’s Chill). The above proposed redesign handles this by changing Flurry to an instant-cast spell with two charges that always applies Winter’s Chill, and making Brain Freeze refund part of a Flurry; this gives players the choice of when and how to use Flurry.
Aside from the Winter’s Chill issues, Frost - although a little boring and suffering from severe repetitiveness when it fails to get procs - is otherwise well prepared for most of the challenges it faces and puts on a respectable showing in all content. An honourable mention should go to Icy Veins, though, which is both a fairly weak cooldown and doesn’t interact with the spec at all (Frost gets no additional procs or power from adding haste than ‘more casts in this time span’). The changes suggested by Norrinir et al., above, adds Flurry interaction to the spell.
Finally, a quality of life issue: when enemies are in melee range, Frozen Orb tends to either slow down too late (after having passed through most of the enemies) or not slow at all. Combined with the unreliability of the Flurry-powered Shatter combo in melee range, this makes Frost very frustrating to play whenever the player is forced into melee range. One potential solution is to use the targeting reticule to place Orb, as with the Concentrated Coldness PvP talent.
Little to say here. The Frozen Orb/Blizzard interaction works and Frost performs well in sustained AoE. Adding Frozen Orb to its Mastery is a way to keep that stat relevant when attacking groups of enemies, and is very welcome.
However, if Frost is meant to excel at two-target cleave - a stated design intent during the BfA beta - it needs some sort of two-target capability in its base kit. Currently, this comes through the Splitting Ice talent.
- No glaring issues on live, many talents saw play in BfA thanks to the support of different Azerite traits. Without support from Soulbinds/legendaries etc, talents like Rune of Power, Thermal Void and Ray of Frost will not be used.
As with other Nova talents (Supernova/Blastwave), Ice Nova is underwhelming compared to other talents in its row, which provide much higher damage over the course of a fight. It is useful for bursting or trapping an enemy in PvP, but has little-to-no use in end-game PvE. It needs a significant buff to compete with Bone Chilling or Lonely Winter.
Nerfing Shimmer makes Ice Floes much more attractive on the 30 tier (although Frost already handles movement reasonably well thanks to its procs), but Glacial Insulation suffers from the same problem as Ice Nova: it is near-useless in PvE. If it is the ‘I’m not moving’ talent, it should do more than increase armour, as the vast majority of incoming PvE damage is not physical.
Rune of Power doesn’t line up with any of Frost’s other cooldowns and the spec has a flatter damage profile than Fire or Arcane, so it is forced into Incanter’s Flow. Focus Magic may actually have some viability here, but I’m not a fan of creating a class-wide talent for a single spec.
There are no outstanding issues with the 60 tier; all of these talents have their place. Frozen Touch could be rewritten to make it more obvious that the increase is multiplicative rather than additive. Likewise with the 75 tier and 90 tiers; each of these talents has a place, although at high Mastery levels Comet Storm falls off compared to Splitting Ice, even on a single target.
At 100, Thermal Void has fallen far from its glory days of 100% uptime in Nighthold, but found some relevance again with Orb Build in BfA. Hopefully the talent will continue to see support, perhaps through a legendary item increasing Fingers of Frost generation, in Shadowlands. Ray of Frost is simply not that interesting, simply a button you press on cooldown, while Glacial Spike actually interacts with Mastery and changes your rotation; it is difficult to beat its power for a mainstream build.
- Fix the Shatter combo and its interaction with Flurry travel time/Winter’s Chill to work at all ranges.
- Icy Veins to interact with spec mechanics, rather than simply being ‘Do what you always do, but faster’.
- Perhaps a long-shot, but take lessons from the efforts of high-level theorycrafters to redesign the way Frost works. This would protect against dry streaks of no procs (bad luck protection), future-proof the spec against dropping Ice Lance and make Icy Veins work with the spec: https://vituscze.github.io/frostmage/ .