If frost spellpower scaling is improved, it also improves haste, crit and hit values (in terms of DPS - not in terms of relative DPS increases).

There's not much they can do with haste or hit, so if scaling should be changed, spell power, crit and spirit are the best candidates. I think the best they can do with crit and spirit are to make them "less bad". I think the thing to remember about crit is that making our talented crit chance higher will just make crit rating on gear less effective as a scaling stat and making our crit bonus higher will make it better, so if DPS is to be increased by adjusting crit, it should not be done by increasing the chance to crit.

I'm sure frost PvE could be fixed (at least for the existing raid tiers) without touching crit in any way.

I'd agree that crit scaling is the reason that we're falling behind Fire and Frostfire builds, but there are plenty of ways to fix this without adjusting crit rates at all. It would probably be better to get better scaling on any stat other than crit just to set us apart in gearing choices.

A few more points on Empowered Frostbolt or anything that gives us a better mutliplier on spell power could come a long way to solving the scaling issue.

But I think it would make more sense to have them tweak haste though. That's a pretty direct way to improve DPS and unlike all the the other scaling items haste doesn't improve DPM. We have more than enough mana in PvE. There are multiple ways to tweak haste without just trying to add a talent that increased Haste by some set % (although that would work too). Blizz could allow hast to increase the duration of icey veins or reduce its cooldown.

Any frost changes would really have to happen deep in the tree as significant changes to IV would also affect Arcane and FFB builds. Dependant on the nature of the change, we could potentially see the SWP era double IV FB builds reappearing. That's the tricky bit really (beyond the PvP element) as any frost changes could have a knock-on effect for other builds.

The difference is that in order to pick up double Icy Veins, you would need 21 point Cold Snap, and realistically speaking, Blizzard will never buff Icy Veins to a point where two Icy Veins outweighs Living Bomb. Glyph of Living Bomb alone is practically more DPS than Icy Veins.

Making Brain Freeze Fireballs scale better would go a long way towards improving frost dps. I particularly like the idea of swapping Brain Freeze from Fireball to Frostfire Bolt

There are other better changes that could surely be made, but they would have a bigger impact on PvP.

I for one am constantly annoyed with the PvP balance issues that prevent needed PvE changes. I fully understand the need for PvP balance, but there are ways around the issue. It would be nice to see spells simply put do different damage to players than to NPCs.

However, I digress; how to buff Frost in PvE, a challenge that has been around ever since we left Ony/MC/BWL and the world of Fire immune bosses.

OK, I need to revise my assessment. Remember I said earlier that they had balanced Warlock specs within about an 11% spread (actually closer to 8%, now that I run the latest sim myself) despite the scaling factors having a spread ranging from 30% to 38%? I just ran simcraft using the T8 Mage data file, using the following command:

Simple run, if anyone wants to suggest changes I'm happy to rerun. But the results I got were these:

Spec	DPS	% behind top
53/18/00	6643	10.6%
57/03/11	7029	4.6%
00/53/18	6926	6.1%
20/51/00	7349	0.0%
18/00/53	5739	28.0%
Total Spread	28%

Spec	Spell Power	Hit	Crit	Haste	Factor Ranks
53/18/00	1.42	2.91	0.81	1.16	0 1st, 1 2nd, 0 3rd, 3 4th, 0 5th
57/03/11	1.51	3.00	0.85	1.21	1 1st, 0 2nd, 3 3rd, 0 4th, 0 5th
00/53/18	1.63	2.50	1.26	1.62	2 1st, 1 2nd, 0 3rd, 1 4th, 0 5th
20/51/00	1.69	2.79	1.11	1.58	1 1st, 2 2nd, 1 3rd, 0 4th, 0 5th
18/00/53	1.36	2.30	0.48	1.13	0 1st, 0 2nd, 0 3rd, 0 4th, 4 5th
53/18/00 Spread	19%	3%	56%	40%
18/00/53 Spread	24%	30%	163%	43%

So you'll note that the spread between 20/51/0 and 53/18/0 is very similar to the spread between the top (0/13/58) and bottom (3/52/16) Warlock specs (11.9%). So it makes sense to compare the scaling factor ranges between these two pairs of specs as well. As you can see above, the spread between 20/51/0 and the best scaling factors ranges from 3% (Hit) and 56% (Crit), with Spell Power sitting at a reasonably low 19%.

The spreads between 3/52/16 and the best scaling factors for Warlocks range from 10% (Hit) and 36% (Crit), with Spell Power sitting at 34%. The total range is smaller, but Arcane's lower spread on Hit and Spell Power help to balance the higher spread on Crit and Haste.

So this is a good example of how stat scaling doesn't need to be homogenized to result in a decent total performance spread.

But then we turn our attention to Frost, and we see that Crit is, indeed, an astronomical outlier. 163% spread between Frost and FFB. 131% between Frost and Fire.

The other problem, of course, is that Frost is dead last in benefit from all four of the stats. Even 53/18/0 comes in second on one (fourth on the rest). It's probably that scaling with all factors needs to be improved for Frost, which likely means an increase in the flat % damage multipliers is necessary.

But contrary to my earlier statements, crit probably will need to be addressed directly. Compensating for a wild outlier would require creating another wild outlier (or two), and that wouldn't be very elegant.

In short: diversity good, wild outliers bad, and Crit is a wild outlier for Frost, more so than I realized.


ETA: Either I'm doing simcraft wrong, or the Warlock thread is outdated. The actual spread in their specs is a mere 8.8%. High spec is Affliction+Doomguard, low is Meta+Ruin. Max factor spread is about 30%, and two factors actually do more for the Demo build than the Affliction build.

Here is a direct factor-for-factor comparison between the top specs, rather than just frost's scaling factors vs. the top spec for that factor. Afflic+DG vs Demo first for comparison:

Spell Power
53/0/18: 1.59
3/52/16: 1.19
Spread: 33.6%

Hit Rating
53/0/18: 2.24
3/52/16: 2.43
Spread: 8.5% (in favor of Demo build)

Crit Rating
53/0/18: 1.59
3/52/16: 1.19
Spread: 12.2%

Haste Rating
53/0/18: 1.59
3/52/16: 1.19
Spread: 10.6% (in favor of Demo build)

So the maximum spread is 33.6%, and two of the factors actually go in Demo's favor. Now let's look at Fire vs. Frost:

Spell Power
20/0/51: 1.69
18/0/53: 1.36
Spread: 24.3%

Hit Rating
20/0/51: 2.79
18/0/53: 2.30
Spread: 21.3%

Crit Rating
20/0/51: 1.11
18/0/53: 0.48
Spread: 131.0%

Haste Rating
20/0/51: 4.85
18/0/53: 4.39
Spread: 10.4%


It's clear that, when addressing performance, Frost needs two things:

1) Moderate general improvement to all scaling factors.
2) Significant specific improvement to scaling with Crit Rating.

Sure you can -- with enough haste, you can fit one shatter and a few ice lances into the 5 second Deep Freeze stun window. I think the haste number I've seen floating around that PvPers aim for is 13%, which is 426 haste (though I've seen people say 500 is the mark they aim for).

So your global cooldown after Deep Freeze would go from 1.5 to 1.32 seconds. Then you'd have a 2.2 second Frostbolt cast, plus an ice lance with another 1.32 second GCD. With Icy Veins up, this increases to 35.6%, so 1.5 GCD is decreased to 1.1. Frostbolt is decreased to 1.84. That's 1.1 + 1.84 + 1.1, which is 4.04 seconds. Throw in [Platinum Disks of Swiftness] and we get even further. 426+375 = 801 = 24.4% haste, with Icy Veins is 49.28% haste. Global cooldown is decreased to 1 second, Frostbolt cast time is 1.67. You can now do 1 (GCD) + 1.67 (FB) + 1 (GCD after Ice Lance) + 1 (GCD after second Ice Lance) + 1 (GCD after third Ice Lance). The third ice lance leaves your hand after 4.67 seconds, when the target is still frozen.

So you see, Deep Freeze indirectly revolves around haste. Similarly, haste can indirectly influence things like Pyroblast. The 10 second window to cast Hot Streak Pyroblast never increases when T8 4P bonus procs. It's possible then that if you were extremely lucky with procs, the more haste you had, the more pyroblasts you could fit into that 10 second window. You have your normal 1.5 second GCD after the first spell, so 8.5 / 1.5 = 5.6. After 5 pyroblasts, you have 1 second left on Hot Streak buff, but you have a 1.5 second global cooldown. But... with 13% haste, you increase that number to 6 possible Pyroblasts.

Another example of haste based talents is boomkins. DPS is their single strongest stat (although it is a step-function so single points could be devalued).

I have no idea what any part of this post means--which of our talents scales strongly with haste?

I believe he's trying to say that Boomkins are an example of a model that scales strongly with Haste. I know about 0 about DoomChickenry, but I do know that because of Frostbolt being our spam before procs, Haste scales much more strongly than additional crit for Frost Mages, since we are effectively soft-capped in crit.

Earth and Moon has a significant haste component to it from what my boomkin friends tell me.

The DPS value of crit will go up if frost DPS goes up, so buffing frost in almost any way (except adding more base crit) will also improve the value of crit rating. That's why the simulationcraft numbers can be a bit deceiving.

I think it would be possible to take the empowered frostbolt and winter's chill crit chance increases and convert them into crit scaling improvements. That would improve crit rating scaling by lowering base crit chance and by increasing crit scaling. I think it would be a relatively PvP-neutral change.

Here's a new crit-related idea: every spell of yours that crits will increase the lifetime of your currently existing water elemental by one second. (Replacement for the fixed 15 seconds from enduring winter.) I just came up with this idea, so I'm not 100% sure yet that I would want a talent like this.

I guess a mage (selfish only) version of "Demonic Pact" might be more fun.

I was thinking along lines more similar to the old Night of the Dead talent for unholy DKs: each critical strike from Frostbolt or Ice Lance would lower the remaining cooldown on Water Elemental by x seconds (1/2/3 seconds seems like a good place to start). If crit still scales poorly, they could make the cooldown reduction apply to other talents as well (Cold Snap? Icy Veins?) or increase the rate at which crits lower the cooldown (bearing in mind that if it scales too fast, you can also cap the crit benefit from this talent).

Frost scales well with haste and suffers from pushback. Why not make the Icy Veins talent a crit-proc? Something along: Your critical strikes have a 50% chance to increase your spellhaste by 5% for 10s. Stacks up to 5 times. While under the effect of Icy Veins, your pushback is reduced by 20% x number of stacks.

In principle it should then be active all the time and close the 28% dps gap nicely.

A raid buffed Frost mage will be approaching 90% crit rate with Shatter up from Fingers of Frost. Since Fingers of Frost has no bearing on crit, haste becomes a dominate stat. More haste results in more casts, resulting in more Fingers of Frost procs. Although it doesn't increase the amount of procs you have, it does increase the number of Frostbolts you can fit into the fight duration, which increases the average percentage of them being affected by Shatter (since more events brings you closer to the average).

Even if a talent was added that affected the value of crit, it would still be somewhat of a lackluster stat. I mean, I'm using [Embrace of the Spider] over [Eye of the Broodmother] in my frost build; I'm even using [Illustration of the Dragon Soul] over Broodmother.

It's almost a double edge sword. On one hand, Shatter gives you a very high crit rate, so much so that every time Shatter is up, the value of crit from your gear goes down. And then on the other, Fingers of Frost affecting your next two spells greatly increases the value of its 15% proc rate. And the only way to "influence" that proc is through haste.

In other words, I'm not sure you could ever improve the value of crit without changing how both Shatter and Fingers of Frost work. I particularly wouldn't like to see something done with crit. When I think of frost, I think of a spec that is very mana effecienct and very fast at casting.

There's no interaction between haste and Fingers of Frost. Personal DPS is proportional to (1+haste), and that's that. Yes, I'm specifically excluding the Water Elemental to simplify that statement.

Frost is in the very strange position of having a high effective crit rate (remember, FoF adds a ton of crit, upwards of 13.75% in Frostbolt spam) with comparatively weak crits. Arcane Potency is only 3% on average, so it has a much less distorting effect, and since Frost DPS is still linear with crit, this highly suppresses the value of additional crit and puts more emphasis on haste and spellpower instead.

Aye, but that doesn't matter, because you also have more casts that do not happen while FoF is active.

Say you have 0% haste, and you cast 100 times. You get 70 casts without FoF and 30 with.

Now you add 20% haste. You cast 120 times. You get 84 casts (+20%) without FoF and 36 (+20%) with.

Your DPS goes up 20% -- exactly what we expect from the 20% haste.

"More casts = more chances for X to proc" only matters when the effect has a cooldown (as it can mean a slightly faster proc after the cooldown expires).

What I think Enthorn is saying is that the proc rate for FoF is more consistent at higher cast amounts. It is more likely to get a proc rate close to 15% when there are a 120 casts involved instead of just 100. That example is a bit extreme, but it is like flipping a coin. The more you flip the closer it gets to the theoretical outcome.

That being said, frost mages are likely to hit a soft cap of 50% crit outside of FoF procs meaning that beyond 50% crit the value of crit drops significantly, while haste has no such cap until you are getting 1 second Frostbolt casts.

An interesting idea that I had that could improve frost dps would be a rolling buff that would increase frost damage by a certain % for X seconds or even make it increase critical strike bonus damage. Make it work similar to Incanter's Absorption and have each event start their own modifier and have it continuously roll. With more haste one could work with a greater amount of this buff rolling. The only problem I see is that it makes Frost mages stack even more haste or try to get their haste to distinct amounts similar to how Shadow Priests only need a certain amount of haste before the value of the stat diminishes significantly.

This is exactly what he's saying, but the way it was worded, it looked like there was an implicit assumption that when the proc rate is inconsistent, it's more likely to be lower than expected than it is to be higher than expected (we hashed this out in PM's). This is of course not the case. Additional haste will narrow your deviation from the expected DPS by effectively increasing the sample size, but there's no reason to expect it to be an improvement to your average (beyond the improvement granted directly by the haste itself). Predictability is not a bad thing, of course.

Let's see where the soft cap comes and hits us:

4% from empowered frostbolt
8% from winter's chill
5% from moonkin aura/elemental oath
x% from molten armor (let's say 5% and the rest depends on gear)
----
22% total (17% leaving out molten armor)

In order to go over 50% crit before FoF, you need 28% crit on your paper doll (or 33%, if you are raid-buffed and have molten armor on you). It's reachable, but I still have way over 10% to go. Frostfire mages speccing frost might hit it if they don't change their gear priorities, but current best in slot frost gear is quite far off the soft cap.

Past the soft cap, the benefit of crit rating is diminished by about 28% (1-0.85*0.85 is the chance of FoF to be up for frostbolt spam).

I think from a game design point of view, the soft cap may some day add depth to how you gear up for frost. Less harsh than the hit cap, but similar, I suppose.

I think if people expect frost pve to be competitive in this expansion they are just going to be dissapointed by the QA, because mages have viable dps specs, and virtually all buffs to frost mages would take a well balanced pvp spec and potentially make it OP.

I saw some posting (not sure where, but last few pages), and people seem to be confused about the usage of FoF procs. Someone was saying that your timing needs to be really good to make both of them count, but really thats not the case. Whether or not something crits is calculated when it leaves your hands, so its just a question of whether you ice lance fast enough to determine if that one counts as frozen (obv the frostbolt that precedes it always will). If you are spamming ice lance during your bolt cast this should work pretty much 100% of the time.

iirc its actually more likely that you will not have the lance be considered frozen if you have really low or high latency. Really high causes obvious problems because you dont cast quickly enough after the bolt, but really low makes it so the server and client know more quickly that you have used FoF and it doesnt queue your lance enough in advance to cause it to calculate as a crit. With slightly lower latency you should always get it off (i play with 60ms and i dont think i have ever had a lance on the "3rd" FoF charge not work).

Lastly I think if they have a chance at buffing frost pve its through spirit. They cannot give higher crit modifiers to frost because shatters in pvp would be too high (think of warlocks who currently crit chaos bolt / conflag instantly, which should see some nerfs sometime soon i think) that consitently double crit (not as consitently as it used to, would probably be <40% chance to double crit on resil target now, but those times the nuke would be very big).

So you cant really touch crit modifiers i wouldnt think, cant give more spell or haste because that would probably be too strong for frost pvp also, what you can give is a new spell that somehow doesnt work well in pvp (ilke a very slow cast, pyro/starfire type) or make spirit give some benefits to pve dps, because its generally an avoided pvp stat.

My personal guess is that frost pve will get slightly buffed, but its unlikely to get within 15% dps of ToTW Fireball IMO

I should add that I think there is an idea for pve dps that blizzard hasnt used yet that I think would be fun and could help frost pve. Make a few main nukes that get increased dmg when used in some sort of alternation, and decaying dmg as you spam it.

For instance lets say they added Ice Boulder which is essentially a frost pyro blast and each time you cast it your next Ice boulder's dmg is decreased by 10% for some time period, and frostbolt somehow works the same way so that you have to cast a few of them, then alternate back to the other. This could be fun I think and work even better with 3 in the rotation. This would have to be available via some talent otherwise it would mess up other elements of the game. Its not a terribly polished idea at this point, but I have always wondered why they always just give things a cooldown, seems like this would be a new interesting way to ensure that a rotation is used (unlike current frost dps bolt spam)