Lhivera’s Library

Adventures in World of Warcraft, Dragon Age, the real world, and beyond

How to Solve Frost's Soft Cap Problem?

Sunday, February 24 2013 at 1:50pm CST

Many specs have a soft cap on one of their secondary stats: an amount they try to reach and not go over, because past that point, the value of the stat declines significantly.

Frost deals with not one, but two soft caps: haste and crit.


About 50% of Frost's DPS stops benefitting from haste at 50% haste, when the global cooldown hits the minimum of one second: Frozen Orb, the selected bomb talent (the bomb's damage benefits from haste past 50%, its GCD can't go any lower), Ice Lance, and Brain Freeze Frostfire Bolts.

50% may sound like a lot, but with the raid buff and Frost Armor, the additional haste required drops to about 33.5%. A switch to Molten Armor is problematic due to the crit soft cap. A switch to Mage Armor is generally unwelcome because Mastery is a weak stat for Frost.

Even so, few Mages are likely to sit at 50% static haste or more. The problem arises when we mix in active haste buffs. Because Frost likes Haste so much, a Frost Mage will carry as much passive haste as possible, which means an active haste buff will put us way over the hit cap, sharply devaluing the buff. It's not a simple matter of the player carrying less haste: it's our most valuable stat, so we're usually better off stacking static haste and avoiding haste procs than we are reducing our static haste to accommodate the haste procs.


About 40-42% of Frost's DPS stops benefitting from Crit at 28% crit: Ice Lance and BF-FFB. Because these spells are cast only when they will benefit from Shatter, 28% crit rate gives them a 100% chance to crit against a boss. Against adds and spawns that can be frozen, the percentage of lost damage is higher, since Shatter affects more spells.

Frost does its best to get to 28% buffed passive crit, and thus receives sharply reduced benefit from dynamic crit buffs.


  • Volatile Talisman of the Shado-Pan Assault: ~20.7% haste proc
  • Sinister Primal Diamond: 30% haste proc
  • Unerring Vision of Lei-Shen: 100% crit chance proc
  • Tier 15 2-piece bonus: ~4.24% haste, crit and mastery proc

All of these items will be undesirable for Frost. Additionally:

  • Breath of the Hydra: Locks us out of Frost Bomb, which is undesirable to some Frost Mages because Frost Bomb provides predictable BF procs.

This leaves only two trinkets as strong options for all Frost Mages:

  • Cha-Ye's Essence of Brilliance (admittedly a very good option for Frost)
  • Wushoolay's Final Choice

There is no alternative to the legendary meta gem; Frost Mages may be better off using the non-legendary mega gem instead. There is no way to avoid the two-piece set bonus (the four piece is strong).


This problem is only going to get worse in later tiers. Glyph of Icy Veins helps us avoid the soft cap with passive haste, but since we want to be as close to the soft cap as possible, it doesn't solve the problem of active haste being undesirable.

Frost needs a long-term solution to dealing with the soft caps. Until it has one, it will suffer from significant limitations in its selection of gear with dynamic buffs.

The question is: how can it be fixed? Convert haste in excess of 50% to something else? (Not crit, obviously!) Change the way Shatter works? Improve the value of Mastery so that Frost prefers it over the soft-capped stats? Post your ideas in the comments!


Submitted by Rincewind on

The most straightforward I agree is to make Mastery stronger, but it would need to be stronger enough I think that it would be at least equal if not stronger than Crit. Otherwise I would see a lot of dancing to try to hit 28% crit exactly (which some people do like), which would make Frost the only spec that would try to juggle 3 caps. Instead I think the best solution would be to eliminate at least one cap. Since the GCD cap is pretty universal, that means eliminating the crit cap.

To go about eliminating the crit cap means that Shatter needs to not increase crit chance anymore. Instead maybe Shatter should increase the critical strike damage of spells against frozen targets. This is increadibly similar to Frost's mastery, but this isn't unheard of (Shadow Priest Mastery is basically Crit for DoTs).

This in itself may pose balance problems (primarily in PvP) for Fire & Arcane, but if so maybe it would be best for Shatter to be Frost only. This would still mean that Frost would have a higher burst potential (again a PvP issue) so this would still be a potentially dagerous idea. It would also replace the 25% damage increase on Ice Lance from Fingers of Frost.

To go with that Fingers of Frost and Brain Freeze would also need to grant Ice Lance and Frostfire Bolt a 100% critical strike chance. This would basically bring things back to the current status quo.

So overall, this would eliminate the Crit cap for Frost, but its not a perfect solution. There are potential balance problems in PvP doing this, but they could be overcome (maybe this version of Shatter doesn't work on players?). This may also require other nerfs (primarily because Frostbolt & Frostfire Bolt would be buffed) but given they've had to buff Frostbolt already, perhaps this is a good thing (also means Frostbolt now scales better with Crit). Its also possible that this makes Crit too good, where for thematic purposes I would rather Haste remain our best stat (at least up to the soft cap).

And maybe after all that making Mastery better just makes more sense :).

Submitted by Nathyiel on

Difficult to say ...

For band-aid fix, your solution feel a little too-big (pvp problem ?). My idea would to increase Mastery's coefficient so it's value would be superior of the haste one.

In long term, I think it's the whole spec that need to be redone.
Frost have a quick and nervous game-play. Recently, it's was toned down by the debuff for PvP reason to nerf burst from Ice Lance. It's the same problem that haste: all lot of instant.

How to deal with it ?
Maybe by reducing the number of instant spell by changing one of the 2 instants for a "short-cd" rotational spell (< 10s).
Maybe by reintroducing a big damage on a mid-long CD, like Deep Freeze in Cataclysm.

edit at Rincewind :
changing Shatter is very difficult, maybe the formula by reducing the "+50%" to "+25%", by example.
Making BF-FFB and IL  guaranteed crit will devalues crit scaling enormously. THe result won't differ a lot from the actual situation.

Submitted by Sombrelune on

Hi, To begin with, I apologize for my poor english. I ' am a french man living in Reunion Island. I can't post on us forums.

Frost is my loving spec and I want to give my ideas about improving it.

1 : Remove the frostbolt debuff or make it more easy to handle.

Frost was really strong in Cata, in fast switch. Now switching means to lose dps.

If we remove the debuff and give the 15% to the respective spells, we must nerf icelance damage in pvp by 20% ( like frostbomb).

Compared to live it will a small nerf to frost pvp ( who need it) and a small up to frost pve ( aoe + cleave + switch), but also a huge improvement in QOL.

If removing this debuff is too radical, my other idea is to make this debuff easier to apply and to maintain.

Give all the " bolts " the capacity to apply the debuff : FrostBOLT, frostfireBOLT, Mirror image's frostBOLT, Blizzard's frostBOLT, Frozen orb's frostBOLT, elemental's waterBOLT

If if too much, remove 1 or 2 types of " bolts" .

It will be a small up to frost pve ( aoe + cleave + switch) and a huge improvement in QOL but we still have manage the debuff.

Maybe it is too easy.

So at least we could make that ' brain freeze FFB, and waterbolt  only refresh the stacks ( but don't apply it).

With that we could keep the pet on our main target, maintaining the debuff, while we switch to new targets. We could save a FFB for that new target.

Our bond with the elemental will be stronger ( thanks to him we don"t lose dps when switching back).

2 : Make mastery a better stats in pve without a big up in pvp.

We could raise the scaling of WE with mastery. In pvp it is easy killable and it is part of our sustain dps and not burst dps.

Furthermore we could give back damage to Deep Freeze ( again immune stun targets). Deep Freeze was a great spell in pve, with a huge mastery an sp scaling.

Obsviously if we give his damage back, we must lower frost damage somewhere else.

Or we coud change Icy Veins. In addition to his effect, the caster's target is treated as frozen for all singledps bolts ( frosbolt FFB, waterbolt).

Mastery will scale better in pve ( bigger burst phase) and in pvp, icy veins is dispellable ( but maybe a 5% icelance nerf againt players will not be

unneedfull )

3: Make FFB a desirable spell filler when there is a big hast buff ( non glyphed IV or Blood Lust or Legendary meta gem proc).

We have problems with hast mainly because our filler is a short cast spell. We do have and another cast with longer cast time ; FFB.

I love FrostBolt, but i will be ok if we use FFB when FrosBolt is under 1 sec cast ( and only at that time).

thoses ideas are not mean to be all implemented. But there are differents ideas that could save Frost spec in pve in my mind.

Thanks for Reading, and maybe if some us people likes thoses ideas they can, express them on blizzard us forums ( damage dealing or ptr )

PS:  Give us Freeze back XD ( or at least water jet )

Submitted by Rincewind on

Well, my changes to BF-FFB and FoF-IL were in the context of changing Shatter  - if you don't do the Shatter change, then making these spells guarantee a crit doesn't make sense either. The Shatter change was to ensure that Crit still has value to Frost overall, but one without a cap.

I would be weary of just inreasing the Crit cap by reducing the Shatter bonus, primarily because that would mean having to shift more damage into baselines to avoid it being a straight nerf. It may be a reasonable band-aid, but it would still interact poorly with crit procs. If anything it I think it would make more sense to reduce the multipler instead.

Overall its a hard problem and I think a spec with 3 caps (Hit-Haste-Crit) just ends up in a bad place. Much the same problem happend to many casters during WotLK, except there wasn't an uncappable 4rd stat to feed excess stat points to.

Submitted by Lhivera on

What if BF-FFB were not instant -- what if it were a 2-2.5 second cast that hit harder and treated the target as frozen? The root of the problem here is how much damage comes from instants -- maybe less damage should come from instants.

Submitted by Citizenpete on

A hardcasting BF-FFB would be another loss of mobility and another loss of frosts identity. Pet Freeze in most of the fights, Deep Freeze and BF-FFB benefiting from FoF, all gone and were great and fun tools in Frost`s repertoire.

With the current 10nh/hc and 25nh T15 tests, I can say Frost has a real problem regarding mobility, on demand burst, target switching and AoE (multi dotting). Because I´m a mage, I will sit on bench for some encounters during the progress and even more, if I stick to frost. 

Tortos (multi dotting, high mobility and target switching required)  Durumu (mobility - high if u get a ray and target switching required) Primordius (multi dotting, on demand burst, heavy target switiching and high encounter buffs to secondary stats, make this impossible for frost) Horridon (multi dotting, target switching and mobility required). As far as I have seen, there are at least 5 - 6 encounters I will see in farm raids at the earliest.

According to my tests on PTR, the loss of pet freeze is not really compensated (less FoF proccs even with T15 4p) but the buff to Frostbolt offset this and a little part from the huge Invocation nerf.

With the pretty bad secondary scaling, legendary meta, troll racial (a must have for progression raiders in Throne of Thunder), bloodlust, Volatile Talisman and Tier 15 2p in mind, there is a prospect of an end for frost (already have 12k for an arcane haste/invocaton build).

A band-aid fix would not really help in my opinion, as Nathyiel already mentioned, the spec has to be redone! One suggestion would be, give Frost a few dots: lower damage from Ice Lance and FFB by X% and the rest as a (dispellable) dot.

Submitted by Nathyiel on

An interesting change could be a Starsurge like spell with a DoT component, ie a casting spell with a short CD that proc make it instant.
I mainly think of FFB/BF-FFB. Cast-it as opener, instant cast when it's proc and cast-it to refresh the proc. (like FFB in cata).

For a big damage, Df like, spell, I easily see it cast by the Water Elem. Good interaction in PVE.

at Citizenpete: more DoT = more haste for more tick. Or Dot that don't scale with haste. For me, Frost fell bad with DoT.

edit: They can also change the way BF proc from bomb : more haste = more proc.
edit @Lhivera : just check your green post. Why not.

Submitted by Citizenpete on

I proposed a dot for PvP reasons. Maybe a base tick like living bomb + benefit from mastery could solve haste dependence. Blizzard just should think about -20% vs players.

I´m on the PTR right now with full haste reforge and gem as frost. Starting a fight with 46% haste (scaled to lvl 522) is not fun! I guess 2 haste > 1 Int? 

Another thing I have recognized is, that I do significant more dps with T14 4 piece than with T15 4 piece. I hope that´s on Blizzards radar, otherwise a frost mage would keep 4p till the end of T15 tier.

*Edit* After 2 25 man heroic tests, I have to say, mage is pretty low. Jin'rokh middle of the pack and Tortos at the end of the damage meter:  ~100k behind classes like locks, hunters, boomkins etc. My guild strongly recommend to reroll (Warlock or Boomkin), otherwise I will sit on bench for most of the fights : /

Submitted by Citizenpete on

I have played around with the recent svn Simcraft version and the result was as expected for frost mage (10% behind the other two) with T15 heroic gear. Frost around 180k and the other two around 200k dps. Did anyone else some simulations?

Submitted by Nathyiel on

Mine is still ongoing but they show that:

Wushoolay + Cha-Ye is better than Breath + Cha-Ye by near 2k dps.

Wushoo + Cha-Ye is scaling good with both haste and crit, no bad with Mastery.
For Int = 5.51, Crit=2.44, Haste=2.41 and Mastery=2.13.
With Human/NT/RoP

Test that I feel do :

  • Invoc Vs RoP
  • NT Vs FB
  • homogeneised stat (crit=hast=mastery)
  • haste>crit>mastery
  • haste(IV)>crit(cap)>mastery
  • mastery>crit>haste

edit> and I have 180k as human.

Submitted by Nathyiel on

My better result is 183.1k dps with a troll, Wushoolay+Cha-ye, in Haste>crit>mastery
To resume my test:

Glyph of Icy veins
Not having the glyph will drop the DPS by 1.3k or more, even if haste is under or equal to 4786.

Frost bomb versus Nether Tempest
Choosing Frost bomb reduce the DPS by ~2k in average

Stat preference

  • ha>cr>ma = 180.5
  • cr(cap)>ha>ma = 178.6
  • cr(cap)>ha=ma = 173.6
  • cr(cap)>ha(cap)>ma = 166.8

Human versus Troll
just add 2k dps with Troll.

Trinkets choice
Wusholay (26.7%) with Cha-Ye (28.2%) give 2k more DPS than with Breath (24.6%).

Stat scaling for Troll, Wu+Cha-Ye, ha>cr>ma

  • Int     / Hit   / Crit    / Haste / mastery
  • 5.48 / 6.47 / 2.35 / 2.42    / 2.17
  • 1       / 1.18 / 0.43 / 0.44    / 0.40

all stat are near each other but Haste is over the cap and Crit is near cap (-3k) for me, par of the problem came from the Mage Bomb interaction with Brain Freeze. The more haste, the more bomb dps and the more proc. Making BF act like Early frost (auto activation on cooldown) could help here.

With the number of fight with cleave, it can be interesting to choose Crit(cap) > Haste > Mastery has it's not so far behind.

Submitted by Nathyiel on

I just have change trinket like this :


I haven't change the stuff, and all. I just play around with "gear_hit=" and all. It's more easy to test.

I haven't tested if change all +320 haste to +80 int/ +160 haste could be better as Int > haste*2.
The same can be done with all the +160 hit/+160 haste to +80 int/+160 hit. like this, haste will be around soft cap.


Submitted by Nyteshade on

One way you could get around the crit cap is to have any crit in excess of 100% increase the damage of spell.

ex.  If you have enough crit rating that shatter would give you 110% chance to crit, then the damage is increased by 10%.

Submitted by Tiga on

The Icy Veins glyph could be removed if Icy Veins also reduced the shortest possible GCD to 0.85 seconds. Given the haste meta gem (and the haste proc valor trinket), it's not enough though. This should be a relatively simple change to implement.

Another way to soften the haste cap is to limit Frost haste to a maximum of 50%. Any haste above 50% could be automatically converted to mastery using some sort of multiplier.

If haste needs to have even better scaling for frost, you can do things like make the Fingers of Frost chance increase slightly with haste. Or, as has been suggested on the forums: lower the Frozen Orb cooldown based on haste (I'm not a big fan of this as it becomes much harder to plan your cooldowns). Brain Freeze frequency already increases as a function of haste, so it would be similar to that.