Jump to content


Frost Mage Discussion [Cataclysm]

  • This topic is locked This topic is locked
136 replies to this topic

#1 Zeldyrr


    Piston Honda

  • Members
  • 233 posts

Posted 07 September 2010 - 05:49 PM

Welcome to the consolidated thread on the Frost Mage spec for Cata. This initial post will try to summarize the current state and common concerns. I am not going to make a list of definitive links to wowhead or other sources for talents, spells, and so on--one can assume that an EJ reader is sophisticated enough to find those things on their own. Instead I will concentrate on what is changed and what outstanding issues remain. The first post will list changes, the first reply will focus on potential concerns. From there it is open to discussion to all would-be frost mages.

Signature Abilities

Mages get two signature abilities for specializing in the frost tree.

Summon Water Elemental: Summon a Water Elemental to fight for the caster.

Frost Specialization : Increases the damage of your Frost spells by 25%.

One of the big changes in Cata is that pet scaling is suppose to be fixed, meaning that pets scale as equally as possible with buffs and don’t double dip. For the most part this is implemented. Current scaling on the water elemental in the beta has been reported as follows.

  • Health and Mana: 50%
  • Armor: 0%
  • Spell Power: 40% but this may be academic. Waterbolt may get 1/3 of the mage’s spell power and the spell power of the elemental itself is ignored.
  • Hit, Crit, Haste: 100%
  • Mastery: 0% but Ghostcrawler is indicated that mastery will “scale” as pure damage for trees where sharing the actual mastery would be complicated / illogical.
  • Cit damage: 200%.
Other than scaling the elemental has changed significantly in other ways. It is now a permanent pet without any glyphs and retains the ability to cast Freeze. Furthermore, casting Freeze in a raiding environment is now part of the basic rotation because of this talent:

Improved Freeze : Gives your Water Elemental's Freeze spell a 100% chance to grant 2 charges of Fingers of Frost.


The frost mastery is termed Frostburn.

Frostburn : All your spells deal 20% increased damage against Frozen targets. Each point of Mastery increases damage by an additional 2.5%.

It should be noted that Frostburn has been changed from previous beta builds. It is no longer a debuff applied by frostbolt. It is a straight-up damage multiplier to frozen targets. As expected In this case, targets are considered frozen for the standard list of affected spells when Fingers of Frost is active.

New Spells

Mages get three new spells at levels 81, 83, and 85. Frost is in the enviable position that it is able to take full advantage of all three of the new spells (when the appropriate talent points are chosen.)

Flame Orb [Level 81] : Launches a Flame Orb forward from the Mage's position, dealing 228 Fire damage every second to the closest enemy target for 15 secs.

A Tier 6 talent turns Flame Orb into Frostfire Orb. This had a number of repercussions. One, Frostfire Orb (FFO) benefits from the Frost Specialization, so damage increases by 25%. Two, the orb attack chills and the snare is very strong. Three, this chill can proc Fingers of Frost and Brain Freeze. Four, the cooldown on FFO is reset by the Cold Snap talent. All in all, FFO is a useful spell with good synergy with other frost talents. It becomes a "regular" part of the raiding rotation (see below).

Ring of Frost [Level 83] : Summons a Ring of Frost, taking 3 sec to coalesce. Enemies entering the fully-formed ring will become frozen for 10 sec. Lasts 12 sec. 10 yd radius.

This spell has changed repeatedly during the course of the beta but has ended up in as a pretty unique and effective utility spell. The targets frozen by the ring are indeed considered frozen for talents such as Shatter and Deep Freeze and the freeze effect breaks on damage much like all such effects. The ring will reapply the freeze if the target crosses the ring a second time, but the pulse between freezes is slow enough that this shouldn't be considered overpowered. The spell has obvious uses in PvP and while soloing. PvE uses are probably less frequent but the spell has the potential to be very effective in the right situations in 5-man and raiding environments as well.

Time Warp [Level 85] : Warp the flow of time, increasing melee, ranged, and spell casting speed by 30% for all party and raid members. Lasts 40 sec. Allies receiving this effect will become unstuck in time, and be unable to benefit from Bloodlust, Heroism, or Time Warp again for 10 min.

Yes, mages get heroism. There has been a lot of whining on the forums on both sides about Time Warp. Other classes thing isn't fair that mages get it. Mages complain that the spell is just a repeat of what shamans can do (and special hunter pets) and that we were robbed of a new, unique level 85 ability. Both of these arguments are rather weak.

The fact remains that 10-man raiding, if it doesn't eclipse 25-man raiding in popularity, will still be much more common and much more important in Cataclysm. It is not reasonable to build 10-man raids and always have a shaman present. With 10-man content tuned to be as challenging as 25-man content, it isn't acceptable to have groups without access to heroism. Mages won the lottery of which other class gets to bring heroism to the party.

It should be noted that having heroism also will affect arena teams (RMP gets heroism), 5-man dungeons (good for those achievements that require burst), and soloing. That last area should not be discounted--having heroism available should allow a mage to solo some pulls while questing that would otherwise require the help of a group.


Frost talents (as with all trees in Cata) have changed dramatically. The talent calculator can be found at wowhead or mmo-champion. Each talent will not be linked and discussed in detail, but notable changes and new talents are as follows.

Fingers of Frost : FoF now procs with a single charge but the proc rate has been increased to 20%. This is almost universally seen as an improvement. It makes the rotation feel more like real shatter combos, removes many instances of double instants in a rotation which felt awkward and procs happen quite frequently. The last point is important in that one of the complaints about frost in LK was that you could go forever without a proc.

Frostfire Orb : With one point, this talent changes the new Flame Orb spell and turns it into a Frostfire Orb. This allows the Frost Specialization of +25% damage to take effect. This is a big enough boost to make it a required talent and to make Frostfire Orb part of the “rotation”, being cast whenever the 1 min cooldown is up. The second point adds a chill effect to FFO and this chill effect can proc both Fingers of Frost and Brain Freeze. This can represent a significant (4%+) increase in damage assuming one can react to all the FoF/BF charges. Both points in this talent are now therefore required of any serious raiding build.

Brain Freeze : The notable change here is that instant cast FFBs under Brain Freeze now benefit from Fingers of Frost.

Piercing Chill : This is a new talent that has an interesting cleave-like mechanic associated with it. The basic effect is that frostbolt crits spread the chill effect to two nearby targets. But what seems a soloing / pvp utility talent has raiding implications because these chill effects can proc Brain Freeze and/or Fingers of Frost. So in “cleave” situations where there a few adds near the primary target, damage on the primary target is increased due to the increased number of procs. This is a unique and interesting form of cleave damage; damage goes up but only on the primary target--the adds are there to boost damage but are not themselves damaged.

Enduring Winter : This talent now combines the effects of two live talents. It provides a 10% reduction in mana costs to all spells and it adds a replenishment effect to frostbolt. This is a frost mages raid utility talent combined with a selfish benefit. It should be noted that the first point provides the full replenishment effect--the other two points are the selfish bonus only. If the additional mana reduction is not needed or the overall build picks up efficiency talents from the other trees, then Enduring Winter is probably a 1-point talent. There was discussion of frost mages getting 5% crit debuff raid utility but Ghostcrawler has recently indicated that such possible changes are currently not planned.

Raiding Talent Builds

An issue (or an advantage, depending on your point of view) frost currently has with raiding talent builds is that there seem to be very few logical choices for the discretionary points that all specs are suppose to have in Cata. The starting point for almost all builds is represented by the following partial spec.

23-point Base Spec

We are two points shy of getting to the next tier. One could make some arguments for other talents, but two points in Piercing Chill seems like the most obvious choice. This gives frost mages extra damage on boss fights with adds close to the boss. Once the next tier is unlocked, two points into Frostfire Orb. There are now three points to choose and these are about the only free points in a frost build. Logical choices would be Permafrost (for pet healing and the healing debuff), Enduring Winter (for more mana efficiency and Replenishment uptime), and perhaps Reactive Barrier (for survivability and to help out healer mana). After those three points, the remaining talents fall into place:

Complete Frost Raiding Spec

Ignite is simply too much of a damage improvement to pass up so that locks in most of the remaining points. The build as listed above probably runs with mage armor. If molten armor is to be used, then some of those three points in frost must be put in Enduring Winter and perhaps a point or two moved from Netherwind Presence to Clearcasting. Simulations are showing mana to be really tight without mage armor. So yes, 6-8 points in mana regen / efficiency talents may be required with molten armor.


The choices of glyphs is affected by the choice to run with molten armor or mage armor. When running with molten armor, the associated glyph is quite powerful until the 33% soft crit cap is reached. Obviously when running mage armor, a different glyph should be chosen. The following two glyphs are far more effective than any of the other choices and should be used in every possible case.

Glyph of Deep Freeze : Your Deep Freeze deals 20% additional damage.

Glyph of Frostfire : Increases the damage done by your Frostfire Bolt by 15% and your Frostfire Bolt now deals 3% additional damage over 12 sec, stacking up to 3 times, but no longer reduces the victim's movement speed..

If one is running molten armor, then the appropriate last glyph choice is:

Glyph of Molten Armor : Your Molten Armor grants an additional 2% spell critical strike chance.

Otherwise the correct last choice is:

Glyph of Frostbolt : Increases the critical strike chance of your Frostbolt spell by 5%.

The only other glyph under consideration is the Glyph of Ice Lance:

Glyph of Ice Lance : Increases the damage of your Ice Lance spell by 5%.

but current simulation results show Glyph of Frostbolt to be a larger DPS increase.

Tier 11 Set Bonuses

The initial set bonuses for Tier 11 have been released. From past experience, Blizzard rarely changes these during the beta process unless there is an obvious and egregious problem with them. The set bonuses for all classes seem to fit the "effective but not inspired" mold, which is what one might expect for the initial tier.

  • 2P Bonus : Increases the critical strike chance of Arcane Missiles, Ice Lance, and Pyroblast by 5%. / Instant
  • 4P Bonus : Reduces the cast time of Arcane Blast, Fireball, Frostfire Bolt, and Frostbolt by 10%. / Instant
There is the very minor concern that a +5% bonus on crit could push Ice Lance crits above 33.33%, which would be wasted (since Ice Lance is always cast when Shatter would be active and has x3 crit chance). Given crit levels initially at level 85, this is unlikely to be a problem.

Similarly haste levels are unlikely to be high enough that the 10% reduction in cast time pushes Frostbolt down to GCD levels--though this could be more problematic with Early Frost.


Perhaps the biggest news for frost in Cata is that there is an actual rotation to use in raiding and that there are enough levers to adjust that it is to be expected that frost damage should be raid-worthy without being overpowered in pvp. This is a huge change for frost. The current theorycrafted rotation can be expressed as a priority list. You are monitoring various procs and cooldowns and make your decisions based on the following rules.

  • Frostfire Orb, if cooldown is up.
  • Deep Freeze, if cooldown is up and Fingers of Frost is active.
  • Frostfire Bolt, if Brain Freeze is active and Fingers of Frost is active.
  • Ice Lance if Finger of Frost is active.
  • Freeze, if cooldown is up, Deep Freeze cooldown is up, and Finger of Frost is not active.
  • Frostbolt.
This may seem more complicated than it actually is. The basic rule is that you spam frostbolt and interrupt that spam for FFB on BF procs and for DF/IL when FoF procs. It is best to save BF-FFBs for a FoF proc. One can force FoF to proc twice with Freeze and should do so whenever you have DF active to take maximum advantage.

External Links of Interest

EJ Cataclysm Mage Resource Thread

#2 Zeldyrr


    Piston Honda

  • Members
  • 233 posts

Posted 07 September 2010 - 05:51 PM

There are a number of outstanding issues and concerns for frost mages. What is perhaps most remarkable is that most of these could be classified as “fine tuning” or “minor concerns”. This is a welcome change from previous expansions where the concerns were more along the lines of “frost is going to be 25% behind in raids”. I will add to this list as questions come up.

No Choice in Off-tree Talents. There are two damage boosting talents available to frost mages in the other two trees--Netherwind Presence in tier 1 arcane and Ignite in tier 2 fire (for FFB crits). Unfortunately one can (almost) get both of these. This locks frost mages in the a 2/8/31 or 3/7/31 build where the only real choice in off-tree talents is whether to max NP or Ignite. This totally goes against Ghostcrawler’s stated goal that raid specs could be described as “take these talents and then you have your choice of these 7-10 utility talents”.

(As an additional note, there aren't enough free points in the frost tree itself, when working towards a raiding build, to give us these "free utility points" either.)

Frostburn too bursty? The current implementation of the Frostburn mastery works quite well to set up an interesting raiding rotation. There has been some concern that it makes frost too bursty, which could complicate pvp balancing. The counter argument is that health pools will be much larger in Cata relative to damage so burst damage that doesn’t result an unhealable gib is not a concern.

Water elemental mana issues. With Freeze being part of the main rotation and used on every cooldown, there exists the possibility that a permanent water elemental could run out of mana. Initial indications that this will not be an issue but it something to watch.

Piercing Chill cleave mechanic intended? The increased number of FoF and BF procs from PC spreading is interesting. There hasn’t been confirmation from a blue poster that the effect is intended.

AOE Rotation boring. While this is true for many specs, the frost AOE rotation is nothing but spamming blizzard over and over again. Additionally, the flavor of blizzard has been greatly reduced in Cata. Blizzard no longer benefits from fingers of frost, there is no more frostbite, and the (talented) snare of blizzard is down to 40%. An uninteresting AOE "rotation" is probably not important. However, compared to fire, which has spreading DOTs and many more AOE spell choices, frost does appear to be lacking in this department.

Lack of Water Elemental abilities. Frost mages gain the water elemental at level 10 but other than Waterbolt and Freeze, there isn't much interaction between the player and the pet. Some have suggested the elemental have some sort of defensive or healing cooldown such as turning into an ice block or turtling like a hunter pet. Having some sort of AOE attack could also help with the overall AOE rotation--frost mages are certainly free to send a pet command while chain-channeling blizzard.

Only ranged DPS with < 40 yard range. Ghostcrawler has said he likes "X is the only class to..." but being the only class to have less than a 40 yard range seems like a pretty big "only". Time will tell if frost spells are adjusted as they have been for all other specs.

#3 Vykromond


    the staleness of Max's dumps

  •  Benefactors
  • 7777 posts

Posted 06 December 2010 - 08:41 PM

New thread! No posting about level 80 stuff here, this is the fully-Cataclysm thread.

#4 Zeldyrr


    Piston Honda

  • Members
  • 233 posts

Posted 06 December 2010 - 09:09 PM

Updated the section on glyphs in the OP to include the correct recommendations.

#5 Forell


    Glass Joe

  • Members
  • 4 posts

Posted 07 December 2010 - 05:44 AM

Curiously, what changed to make Early Frost part of the base spec? I was under the impression that its ability to take you under the GCD was undesirable, and that it was mostly seen as a replacement for Rank 1 Frostbolt in PVP.

#6 Thaelyn


    Glass Joe

  • Members
  • 8 posts

Posted 07 December 2010 - 06:43 AM

Curiously, what changed to make Early Frost part of the base spec? I was under the impression that its ability to take you under the GCD was undesirable, and that it was mostly seen as a replacement for Rank 1 Frostbolt in PVP.

I guess it couldn't hurt, though I'd imagine the dps gain would be so negligible, it still wouldn't matter. As you said, with haste/frostbolt base cast time/Tier 11 bonus, the cast time is going to be significantly low as it is. However, those two points really have little elsewhere to go.. I'd maybe do something like this Frost Raiding Build

Also, I noticed that Time Warp is labeled as "Time Stop" above. ;P

#7 Zeldyrr


    Piston Honda

  • Members
  • 233 posts

Posted 07 December 2010 - 06:45 AM

At last report from the beta, Early Frost not only lowered the cast time of Frostbolt, but it also lowered the GCD when used. This was the change that made it more than a marginal talent. If this has changed, points in EF might be reconsidered.

#8 Lhivera


    Bald Bull

  • Members
  • 1519 posts

Posted 07 December 2010 - 12:59 PM

Currently, disabling Early Frost in the SimC profiles results in only about a 0.3% DPS decrease. That doesn't agree at all with what I was estimating at the end of beta after the GCD change, but of course it's possible that it simply declines in value quickly as you gear up. I'll need to investigate this.
At Veridian Dynamics, we can even make radishes so spicy, people can't eat them. But we're not, because people can't eat them.

#9 Fröstler


    Glass Joe

  • Members
  • 13 posts

Posted 08 December 2010 - 08:33 AM

Early Frost only decreases the cast time to 1s on live server, not below.
But every additional haste buff decreases the cast time under 1s and also the GCD.

[Edit] Hmm, it seems that i always do not get the full 0.6s shorter cast time.
Is the tooltip wrong?

#10 Incoherent


    Von Kaiser

  • Members
  • 53 posts

Posted 08 December 2010 - 01:00 PM

Hmm, it seems that i always do not get the full 0.6s shorter cast time.
Is the tooltip wrong?

Are you testing hasted? The 0.6 seconds reduction is calculated before haste effects, which means the cast time reduction (of EF or any other haste effect) is itself reduced proportionally to haste.

Currently, disabling Early Frost in the SimC profiles results in only about a 0.3% DPS decrease

My models are not up to date due to busy RL. The 0.3% dps increase sounds low to me, unless Haste has also devalued to where an amount which shaves off 0.6 seconds from total casting time every 15-20 seconds or so gives about the same dps increase. (allowing for extra WBs etc)

I must admit to being suspicious of anything that says Early Frost gets weaker with improved gear, unless you are bottoming out on the GCD floor when casting it. On a purely conceptual level: If you are casting higher power stuff, casting even more higher power stuff must be a DPS increase. The only remaining question is whether it is better than something else.

#11 MalaclypseTY


    Glass Joe

  • Members
  • 20 posts

Posted 09 December 2010 - 05:11 PM

So it has been shown to definitely be better now to hold freeze until DF is off cooldown? Or is there something more complicated like if Freeze comes off cooldown and there is X or less amount of time left on DF cooldown to save it.

#12 Lhivera


    Bald Bull

  • Members
  • 1519 posts

Posted 11 December 2010 - 07:22 AM

I must admit to being suspicious of anything that says Early Frost gets weaker with improved gear, unless you are bottoming out on the GCD floor when casting it. On a purely conceptual level: If you are casting higher power stuff, casting even more higher power stuff must be a DPS increase. The only remaining question is whether it is better than something else.

Yeah, I'm not at all confident in this. Let's throw some output from the latest build of SimC out here with both specs on it (plus, for comparison, Frost-Frostfire and Frost-Frostfire with Fire Power). Maybe somebody can spot something that doesn't look right about the effects of Early Frost. You can also get a preview of the new SimulationCraft output styling I've been working on.

Simulationcraft Results
At Veridian Dynamics, we can even make radishes so spicy, people can't eat them. But we're not, because people can't eat them.

#13 Light4


    Piston Honda

  • Members
  • 164 posts

Posted 11 December 2010 - 02:51 PM

Reactive Barrier seems to be broken. I leveled with frost and also ran some instances. Reactive Barrier only activates if
a) you have no CD on Ice Barrier (as expected) and
B) you have no spell queued.

Since we should always be casting this involves constantly queuing spells, effectively giving this talent the chance to proc only during movement or by stopping casting to save the mana for an Ice Barrier (but that means you have to drop below 50%).
I think there are better talents until it is fixed.

#14 Lhivera


    Bald Bull

  • Members
  • 1519 posts

Posted 12 December 2010 - 01:54 AM

There really doesn't seem to be a huge difference between a maxed spec and a spec that drops 2-3 points in DPS (even from Ignite) in favor of utility. A few variants on Frost, all within a 4.5% range:

Simulationcraft Results
At Veridian Dynamics, we can even make radishes so spicy, people can't eat them. But we're not, because people can't eat them.

#15 Requizen


    Glass Joe

  • Members
  • 7 posts

Posted 13 December 2010 - 05:17 AM


After looking at your Simulationcraft results, I see you made two different Frost-based Frostfire builds, but both of which take Early Frost.

Could there be a possible Frost FFB build that ignores Early Frost and reallocates those two points (maybe to Enduring Winter for mana efficiency or Permafrost for more utility/pet healing) and uses nothing but FFB, ignoring Frostbolt? Or is the dps gain from EF FB really that high?

#16 Naqaj


    Glass Joe

  • Members
  • 23 posts

Posted 13 December 2010 - 08:44 AM

FFB can't proc Brainfreeze, that's why you don't completely replace Frostbolt with it. Early Frost is just an afterthought.

#17 smoug



  • Banned
  • 3 posts

Posted 13 December 2010 - 09:07 AM

If you get a haste bugg that will make your FB go below GCD.
Will it then be better to cast FFB instead?

#18 Lhivera


    Bald Bull

  • Members
  • 1519 posts

Posted 13 December 2010 - 12:13 PM

The builds I'm including in the output are maximized, as best I can -- I haven't even had the time yet to really tune the sub-372 sets. I can't try all possible variations myself, but I encourage people to download the program and experiment with different options if they're interested in seeing how close they can get to the optimal DPS while dropping some of the less-powerful talents. Early Frost and Netherwind Presence are both fairly small contributors, so they'd be good options for experimentation. If you discover something interesting, let us know!

I can say that I tried out a build that maxed Netherwind Presence and dropped Ignite, and was surprised to discover that it's less than 4.5% behind the maxed build. So a more casual, utility-minded build can be used without completely tanking your DPS, which is a good thing. Messing around with NP and EF and Ice Floes, as examples, would likely give results between the maxed builds and the Igniteless build, so they'd all fall within a pretty reasonable range.
At Veridian Dynamics, we can even make radishes so spicy, people can't eat them. But we're not, because people can't eat them.

#19 Nathyiel


    Von Kaiser

  • Members
  • 96 posts

Posted 13 December 2010 - 01:48 PM

I have made a quick test with Magegraf for testing Crit cap or not. The result is a just a little for crit capping (~1%) and only if the cap is obtain by reforging hast/mastery to crit (no enchant/gem change). I haven't had time to test it on SimC for more accurate data.

I also made a few calcul for mana use with the beautiful 379 trinket: . For a 300s fight, you can drop from 390k mana spell cost to 310k mana. In the same length, maga armor give ~92k mana. Maybe, with one more point in Enduring Winter, we can change back to Molten Armor.

#20 NateB


    Glass Joe

  • Members
  • 9 posts

Posted 14 December 2010 - 04:55 AM

Aside from using trinkets designed for a healer to backdoor a 5% crit increase, I'm curious about break points.

Testing using magegraf, at high levels of crit Fire starts to race ahead of the other specs in terms of theoretical dps.
This isn't surprising, because with greater crits there will be a) more free Pyroblast! and B) more mana returned from master of elements, which will, in turn, permit us to use more Fireball and less scorch, a significant dps increase.

I'm curious about frost acting the same way, at least insofar as using molten armor full time, no matter the length of the fight. Consider that you get 30% mana back from crits. As you hit 33%, every Ice Lance and Deep freeze will return 30% of their mana, but they cost half the mana frostbolt does.

Frostbolt = base 2264 mana per cast; 2037 with fully talented enduring winter

Assuming Molten armor 4000 crit rating and 3250 int, you have a ~37% crit chance when using molten armor glyph, or if you swap out the molten armor glyph, a ~40% chance with frostbolt.

at 37% crit, you save an additional 251.3 mana per frostbolt
at 40% crit, you save an additional 271.68 mana per frostbolt

My question is, at this level of crit, will molten armor be viable full time, or do we still need the mana from mage armor? Is there a crit point at which we no longer need it?

0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users