Growing Gap: Frost Concerns before 5.1

his is really sad. Might as well rename it wet-noodle-bolt. Without a frozen target, I swear it hits for barely more than a DOTs tick. ><


How much does Fireball hit for anyway? My Frostbolts consistently hits for 35k per hit, non-crit.
It's mainly the ramp up that's the problem. It's doing only 80% of it's full damage on a non-debuffed target.

Frost only became the PvP spec durring WoTLK btw, for anyone saying frost was the pvp spec from the start.
Reply Quote
90 Undead Mage
5285

Frost only became the PvP spec durring WoTLK btw, for anyone saying frost was the pvp spec from the start.


Frost has always been the dominant general-purpose pvp spec, with fire and arcane being viable/good at various times or for people that enjoyed being a 3-minute mage in bgs. Frost has never been a bad or subpar pvp spec.
Reply Quote
90 Blood Elf Mage
4590
11/15/2012 11:12 PMPosted by Icyheart
Frost only became the PvP spec durring WoTLK btw, for anyone saying frost was the pvp spec from the start.

Definitely not true. 17/0/34 was the dominant spec at 60, then 17/0/44 at 70 (I popularized 17/3/41 as well). AP/Fire was viable at 60 and 70 but not by any means more widely used than frost.
Reply Quote
100 Gnome Mage
15615
I ran the simulations I spoke of earlier. Here are the results:

Relative scale factors using 5.0.5 T14H profiles:

Spec Int Hit Haste Crit Mastery
Frost 3.69 4.92 1.60 1.59 1.50
Fire 4.03 3.45 2.14 2.67 1.82
---------------------------------------------
Rel. Diff. 0.092 -.30 0.34 0.68 0.21


(Frost and Fire values are x 1e-5).

I think this is very troubling for the future of Frost. For int and all the other secondary stats (beside hit), Fire is scaling more quickly, by at least 20% for each of haste, crit, and mastery. That portends a dark future for the spec without significant adjustments.
Edited by Muphrid on 11/16/2012 12:56 PM PST
Reply Quote
100 Troll Mage
11590
A lot people saying fire is higher because there are more good mages playing fire, which is true, but still not enough to hide the fact that frost is much weaker than fire. I'm saying this because I've been able to pull world rank NO.1 on both fire and frost. I've been playing frost since the first day Mogu'shan vaults open and I spent much more time practising frost than fire, but now my fire record is way higher than my frost record.

With bloodlust, frost already reach the soft haste cap 50% and will reach crit cap in the future, while fire doesn't even have any cap. So the difference between frost and fire will be larger and larger.

That's just my opinion.
Edited by Modoshi on 11/29/2012 10:13 AM PST
Reply Quote
100 Human Mage
14910
I ran the simulations I spoke of earlier. Here are the results:

Relative scale factors using 5.0.5 T14H profiles:

Spec Int Hit Haste Crit Mastery
Frost 3.69 4.92 1.60 1.59 1.50
Fire 4.03 3.45 2.14 2.67 1.82
---------------------------------------------
Rel. Diff. 0.092 -.30 0.34 0.68 0.21


(Frost and Fire values are x 1e-5).

I think this is very troubling for the future of Frost. For int and all the other secondary stats (beside hit), Fire is scaling more quickly, by at least 20% for each of haste, crit, and mastery. That portends a dark future for the spec without significant adjustments.


I kinda figured that was the way it would go given how far behind frost still was in T13 after the patch dropped.

Oh well, maybe arcane might be worthwhile with the patch.
Reply Quote
90 Troll Mage
13965
11/29/2012 10:01 AMPosted by Modoshi
With bloodlust, frost already reach the soft haste cap 50% and will reach crit cap in the future, while fire doesn't even have any cap. So the difference between frost and fire will be larger and larger.


Were you using Glyph of Icy Veins? It'll take a lot more gear to reach the cap if you use that glyph. I've been able to push very high WoL ranks and parses with Frost in MV and HoF(for normals at least, guild hasn't reached heroics for the group I raid with), and I don't see a significant gap that isn't explained by Fire RNG until you reach the 3rd boss onward in HoF. Though with the Combustion changes, I don't know how much this would affect the gap.
Reply Quote
100 Troll Mage
11590
[quote="71989854714"]
Were you using Glyph of Icy Veins? It'll take a lot more gear to reach the cap if you use that glyph. I've been able to push very high WoL ranks and parses with Frost in MV and HoF(for normals at least, guild hasn't reached heroics for the group I raid with), and I don't see a significant gap that isn't explained by Fire RNG until you reach the 3rd boss onward in HoF. Though with the Combustion changes, I don't know how much this would affect the gap.


Of course the Glyph is used. Your gear and raid buffs provide you more than 20% haste, right? and lust is 30%, plus Troll's racial ability, it's more than 70%. The GCD cap is 50%
Reply Quote
MVP
90 Human Mage
10015
Muphrid: one of our problems here is that, as of the end of beta, we knew for certain that the Mage simulations were running very high — anywhere from 10-20% high. Last I knew, we still had not been able to figure out why; every number we went over by hand — damage, cast times, crit intervals, expected proc rates, everything — checked out. And yet the differences remained.

I feel pretty confident that if you're looking at Frost, and it tells you that Crit is worth 20% more than Mastery, that's pretty accurate. I'm not at all confident that you can make such comparisons across specs at this time. That doesn't mean there isn't a problem, but I don't think SimC output is sufficient evidence to say that there is.

Real life is eating up all the time I would normally spend trying to figure this stuff out right now, and that's not likely to change soon, but if you had the time and inclination to dive into the SimC dev community and try to find problems, I'm sure it would be much appreciated.
Reply Quote
100 Gnome Mage
15615
Yeah, I know this is an area that is still somewhat nebulous. Still, I've tried to work with what information and models we do have and to insulate them against systematic errors in the simulations as much as possible. That's the reason I don't use raw scale factors, only DPS-normalized ones (the numbers in the table are DPS gained per point / total DPS of the spec). The table says that frost gains .0016% DPS from haste, while Fire gains .00182% DPS from mastery, its worst stat?

Yeah, the models may be wrong--I don't strongly dispute the belief that simc is overestimating Frost DPS (though I do wonder how much the apparent discrepancy comes down to selection effects--a spec believed to be higher DPS will attract more talent and lead to skewing, almost making the state of things self-fulfilling). But it was my hope that normalizing the scale factors in this way would allow us to draw some conclusions. If simc can be reasonably trusted with scale factors (which I admit is a big question), then the reality I see coming is dire: Frost's scale factors are the poorest of the three mage specs, full stop. So the question is this: is simc underestimating Frost scale factors, are they that bad, or is it a combination of the two?

Beyond that, I look at the damage distributions from having played both Fire and Frost, and I don't see any obvious discrepancies. Again, though, in my mind, as long as these distributions are right, even if Frost's damage is systematically overestimated, the conclusions about stat scaling ought to be somewhat reliable.
Reply Quote
100 Gnome Mage
15615
A lot people saying fire is higher because there are more good mages playing fire, which is true, but still not enough to hide the fact that frost is much weaker than fire. I'm saying this because I've been able to pull world rank NO.1 on both fire and frost. I've been playing frost since the first day Mogu'shan vaults open and I spent much more time practising frost than fire, but now my fire record is way higher than my frost record.

With bloodlust, frost already reach the soft haste cap 50% and will reach crit cap in the future, while fire doesn't even have any cap. So the difference between frost and fire will be larger and larger.

That's just my opinion.


I feel this can be misleading. By nature, good RNG with Fire can lead to disproportionately good DPS. Look at simc's charts--as much as we wonder about simc with frost--and examine the distribution of Fire sims. There's a much longer tail toward higher DPS than Frost has. Even if they performed identically on average, Fire's best performance would be much, much better than Frost's, just by the nature of the spec.
Reply Quote
MVP
90 Human Mage
10015
Another important thing to remember here is that the devs don't plan for scaling to carry specs along a perfect tier-to-tier trajectory. They focus on the current tier. As long as the specs are reasonably well balanced for the current tier, they don't wore too much about where scaling will take them on the next tier. Just as an interplanetary probe will make use of thrusters to adjust trajectory a bit when necessary, the devs use the between-tier patches to make adjustments to spec scaling so that they will be reasonably well balanced for the new tier. This'll knock them further out of balance for previous tiers, but nobody cares much since the content is significantly outgeared at that point.
Reply Quote
90 Troll Mage
13965
Of course the Glyph is used. Your gear and raid buffs provide you more than 20% haste, right? and lust is 30%, plus Troll's racial ability, it's more than 70%. The GCD cap is 50%


The issue is that Haste is not additive; it's multiplicative. And, which Soft Haste Cap are you referring to? Are you looking into the one where Frostbolt falls under the GCD, or the cap where the GCD reaches 1 second? Because the GCD being at 1 second is fine, as long as most of our casts remain above 1 second.

I'll use the Haste from my gear as an example in the following calculation (I'll ignore Frost Armor for now, since I'm not sure if it's multiplied, or if it literally increases all your Haste in the end by 5%):

1.05 (raid Haste buff) * 1.20 (Berserking) * 1.30 (Bloodlust/Timewarp) *1.1326 (Haste from my gear) = 1.8551988 = 85.51988% Haste (not sure how Blizzard would round/truncate this value, so I'll leave it as is).

The total Haste needed for Frostbolt to reach a 1.0 second cast time is:

Base Cast time (2 seconds)/Desired Cast Time (1 second cast time) = 2 =100% Haste, so more gear is needed for Frostbolt to become an issue, and even more gear is needed for those who don't have Troll as their race.

As for Mage bomb, that ultimately depends on which bomb you're using (bombs are considered for all specs). Under these conditions, Frost Bomb definetely falls under the GCD, Living Bomb has a 1.0 seconds GCD regardless of Haste, and since NT is an instant, I don't think it's affected. Ice Lance and Frostfire Bolt, when used with Brain Freeze, wouldn't be affected either.
Reply Quote
MVP
90 Human Mage
10015
Instant casts are affected — they're limited by the GCD. At 50% haste, your instants stop benefitting. Frostbolt is the only spell in our rotation that benefits from more than 50% haste.
Reply Quote
90 Troll Mage
13965
11/29/2012 02:24 PMPosted by Lhivera
Instant casts are affected — they're limited by the GCD. At 50% haste, your instants stop benefitting. Frostbolt is the only spell in our rotation that benefits from more than 50% haste.


Oh, I see. Thanks for setting me straight.
Reply Quote
100 Troll Mage
11590
When frost mage do burst damage, most of our spells are instant cast, which means we wont benefit from haste after 50%. Troll+lust = 1.2x1.3=1.56 which is already capped. I can reach 50% haste even without lust (about 30% from gear and raid buff plus troll), which means I have very little improvement under lust (only bomb ticks and frostbolt are affected by lust).

But for most of other DPS, especially for fire mage and balance druid, they have huge improvement from lust.

Dont you think this is an issue?
Reply Quote
100 Gnome Mage
15615
11/29/2012 01:28 PMPosted by Lhivera
Another important thing to remember here is that the devs don't plan for scaling to carry specs along a perfect tier-to-tier trajectory. They focus on the current tier. As long as the specs are reasonably well balanced for the current tier, they don't wore too much about where scaling will take them on the next tier. Just as an interplanetary probe will make use of thrusters to adjust trajectory a bit when necessary, the devs use the between-tier patches to make adjustments to spec scaling so that they will be reasonably well balanced for the new tier. This'll knock them further out of balance for previous tiers, but nobody cares much since the content is significantly outgeared at that point.


Yeah, I'm not addressing this as something that needs to be fixed for right now (any concerns over Frost's DPS right now aside), but the conversation cannot be held too early to get across that there is a long-term problem that will need to be addressed.
Reply Quote

Please report any Code of Conduct violations, including:

Threats of violence. We take these seriously and will alert the proper authorities.

Posts containing personal information about other players. This includes physical addresses, e-mail addresses, phone numbers, and inappropriate photos and/or videos.

Harassing or discriminatory language. This will not be tolerated.

Forums Code of Conduct

Report Post # written by

Reason
Explain (256 characters max)

Reported!

[Close]