Frost Cataclysm Discussion [4.0 thread] - Page 13

Incoherent · 10/24/10, 5:56 AM

Related to the prioritisation of haste, I have a question for Beta testers.

On Live, Haste now improves FFB DPS a little bit by giving us more DoT ticks within the DoT uptime window. I see up to 3 extra ticks of the higher damage at high haste levels. (2 with 19.2% haste, 3 with IV, 19.2% and a 522 haste trinket proc) It is a little unclear (the low damage with rounding makes for high variability, but the total DoT damage done increases proportionally to haste.

Does this behaviour also effect Frostfire Orb? (Or ignites, would need a lot of haste for that) I can extract the effect for FFB myself but I am limited to level 80 on live and I can't access the US Beta thread.

This has implications for how I am simulating it.

Edit: Seems to have been altered in a hotfix. Haste is not affecting FFB dots anymore.

Zaldinar · 10/24/10, 7:44 AM

Originally Posted by Zeldyrr

However, the comments that the combatlog is a bad means by which to test timings made me doubt the other numbers determined in the link, comparing the two different waterbolt macros. Zaldinar, is there a way to write a macro to more accurately test the time between casts of a spammed spell? (As opposed to the GCD time, since Waterbolt cast time > GCD time.)

It's doable, but not as simple as the one I posted for the GCD unless you as a human don't mind doing some math. What exactly are you trying to determine? The modified cast time of a water elementals waterbolt spell and whether or not is pauses between casts?

Lhivera · 10/24/10, 9:13 AM

Stat priority seems to be changing dramatically as stats increase. I've got my sim handling gems and (passive) enchants now, so I'm getting closer to a realistic set of T11 stats. I'll post some updated numbers later today.

Zeldyrr · 10/24/10, 10:48 AM

Originally Posted by Zaldinar

It's doable, but not as simple as the one I posted for the GCD unless you as a human don't mind doing some math. What exactly are you trying to determine? The modified cast time of a water elementals waterbolt spell and whether or not is pauses between casts?

More the later, as I don't think the actual cast time changes. If you set your WE to attack with /petattack, then there are X seconds between casts. If you spam a macro that says "/cast Waterbolt", then there are Y seconds between casts, where Y < X. I'm trying to determine X and Y, to see if it is worth it to macro in "/cast Waterbolt" to every frost mage attack.

Lhivera · 10/24/10, 10:39 PM

(I had hoped to have new hit/crit/haste/mastery comparisons done before posting this, but it ain't gonna happen tonight. Hopefully tomorrow.)

OK, with the following stats derived from T11 and equivalent gear, using mostly veiled gems (which are now purple) to fill red and blue slots, and a hit enchant on the bracers:

Health: 89633
Intellect: 4086
Mana: 80554

Spell Power: 6393
Hit Rating: 1216 (94.87%)
Haste Rating: 1072 (8.37%)
Crit Rating: 1509 (23.62%)
Mastery Rating: 1406 (15.84)
Meta Gem Active: Yes

The percentage numbers include talents and base values.

I've added a feature to always use the average base damage of a spell rather than a random range to decrease variance a bit. Still pretty subject to RNG; even with 500 iterations I'm getting results varying by as much as 80 DPS. I think to correct that I'll need to implement an option to use average damage based on a given cast's crit chance. This'll also involve averaging the MOE return.

Anyway, point is, it's hard to get extremely precise on the value of individual talent points. But here's my best estimate at present.

Base Test

No Early Frost, no Ignite, no Netherwind Presence.

Early Frost

The GCD change seems to have really evened out the value of the two talent points. You get full value except when running over 30% haste. Value does decrease a bit with haste.

With zero haste, EF takes a Frostbolt cast from 2.0 secs to 1.3 secs, saving you 0.7 secs. With 10% haste, EF takes a Frostbolt cast from 1.82 secs to 1.18 secs, saving you 0.64 secs. So 10% haste effectively reduces the value of Early Frost by 8.6%. This explains why the value I'm getting with gems and enchants is a bit lower than the value I posted a couple days ago.

EF appears to be worth about 1.14% DPS per point
Second point may actually be worth a bit more than the first, 1.21% vs 1.07%
That may just be due to RNG, but I ran several 500-iteration trials and got pretty consistent results

Ignite

This is assuming that it is unintended for Frostfire Orb to be unable to generate Ignites (Flame Orb can). If it turns out to be intended, the value will decrease.

Ignite appears to be worth about 1.4% DPS per point
Sim values were 1.48% for the first point, 1.18% for the second, 1.53% for the third
The values of the individual points was not very consistent across multiple runs, though total value was, so I think the average value is closest to truth

Netherwind Presence

Netherwind Presence is a pretty weak talent for us. It's partly hurt by the fact that it doesn't affect the Water Elemental (though Haste Rating does), and partly by the fact that it cannot help us cast additional Deep Freezes, Freezes or Frostfire Orbs. Those four sources make up a considerable percentage (about a third) of our damage.

A point of Netherwind Presence is worth about 0.7% DPS.

Enduring Winter

As expected, better gear greatly increases our endurance and full points in Enduring Winter becomes increasingly unnecessary. I set the spec to optimal DPS (2/2 EF, 3/3 Ignite, 2/3 NP), and increased the duration to 10 minutes. Over 500 runs, it determined the average time at which the caster was unable to cast any spells at all (spells would be cast later in the simulation as Evo came off cooldown or through regen). I don't now how long fights will be, but I suspect that even if we need 3/3 EW when starting T11 content, we'll be able to drop to 2/3 by the end of that content, if not even 1/3.

Enduring Winter 1/3: 7:00
Enduring Winter 2/3: 8:20
Enduring Winter 3/3: 9:00

Incoherent · 10/25/10, 2:35 AM

Really nice work.

Originally Posted by Lhivera

So 10% haste effectively reduces the value of Early Frost by 8.6%.

It also reduces the value of Netherwind Presence and Icy Veins (and whatever amount of haste you already have) by the same percentage. (9.09%?)

I don't now how long fights will be

That is a big factor for us. We are so dependent on these CDs, Deep Freeze, Freeze, FFO and Early Frost that the Fight length, especially where we Cold Snap early, inversely and adversely affects our DPS. As it has been for a long time, Frost remains very bursty. The (small) integer number of events means DPS vs fight length is a pretty bumpy curve.

Zaldinar · 10/25/10, 3:59 AM

Originally Posted by Zeldyrr

More the later, as I don't think the actual cast time changes. If you set your WE to attack with /petattack, then there are X seconds between casts. If you spam a macro that says "/cast Waterbolt", then there are Y seconds between casts, where Y < X. I'm trying to determine X and Y, to see if it is worth it to macro in "/cast Waterbolt" to every frost mage attack.

Ugly with a macro, I think I can work it into my cast bar monitoring addon though. I'll see if I can make that happen today or tomorrow and get back to you.

Nathyiel · 10/26/10, 5:16 AM

Originally Posted by Lhivera

Ignite

This is assuming that it is unintended for Frostfire Orb to be unable to generate Ignites (Flame Orb can). If it turns out to be intended, the value will decrease.

Ignite appears to be worth about 1.4% DPS per point
Sim values were 1.48% for the first point, 1.18% for the second, 1.53% for the third
The values of the individual points was not very consistent across multiple runs, though total value was, so I think the average value is closest to truth

Netherwind Presence

Netherwind Presence is a pretty weak talent for us. It's partly hurt by the fact that it doesn't affect the Water Elemental (though Haste Rating does), and partly by the fact that it cannot help us cast additional Deep Freezes, Freezes or Frostfire Orbs. Those four sources make up a considerable percentage (about a third) of our damage.

A point of Netherwind Presence is worth about 0.7% DPS.

What about Fire power ?
I don't think it will worth a point. But with 2/3 or 3/3 Ignite, I am just curious about this talent.

Lhivera · 10/26/10, 8:01 AM

So it looks like T11 gear will crit cap Frost (until, of course, T12 when crit gets suppressed a bit, but not sure how much).

Given the following starting stats, which include gems and enchants in mostly T11 gear:

Spell Power: 6393
Hit Rating: 1216 (94.87%)
Haste Rating: 1072 (10.37%)
Crit Rating: 1509 (23.62%)
Mastery Rating: 1406 (15.84)
Meta Gem Active: Yes

+250 Int (+591.97 / +4.30%)
+250 Hit (+385.14 / +2.80%)
+250 Mastery (+225.31 / +1.64%
+250 Haste (+199.15 / +1.45%)
+250 Crit (+101.4 / +0.74%)

The value of Mastery and Haste stay reasonably close together, less than a 15% difference, while the value of Crit plummets. That's expected, of course -- Frost simply treats Crit more like Hit than like Mastery and Haste, very valuable to a point and then not so valuable at all. It means Frost, unlike most other DPS specs, will try to cap two stats and then make choices between the other two, rather than capping one stat and making choices between the other three.

Incoherent · 10/26/10, 9:39 AM

Originally Posted by Nathyiel

What about Fire power ?

This is a really interesting question. I have some (real rough) calculations showing some synergy potential when mixing up FP and Ignite. At this unevolved stage I think it will be better than NP. I have no emperical data on FFO or it's explosion though. If it explodes, if it Ignites, I would definitely include it, if only for the fun factor.

My ghetto Analytical Solution is approximately matching Lhiveras sim on all points in a 360 second fight at the moment, so I would be facinated to see this modelled.

Nathyiel · 10/26/10, 10:23 AM

Originally Posted by Incoherent

This is a really interesting question. I have some (real rough) calculations showing some synergy potential when mixing up FP and Ignite. At this unevolved stage I think it will be better than NP. I have no emperical data on FFO or it's explosion though. If it explodes, if it Ignites, I would definitely include it, if only for the fun factor.

My ghetto Analytical Solution is approximately matching Lhiveras sim on all points in a 360 second fight at the moment, so I would be facinated to see this modelled.

At the same time, we also have to put the glyph of frostfirebolt in question.

Is glyphed Frostfirebolt proc FoF like glyphed Frostbolt in WOLTK?

If yes, it could be a big upgrade with Fire Power and Ignite in place of a plain 5% crit chance for frostbolt.

jak3676 · 10/26/10, 10:27 AM

Originally Posted by Lhivera

So it looks like T11 gear will crit cap Frost (until, of course, T12 when crit gets suppressed a bit, but not sure how much).

Given the following starting stats, which include gems and enchants in mostly T11 gear:

..
Crit Rating: 1509 (23.62%)
..

You lost me a bit here. This seems to be contradicting the "stack Crit to 33%" unless I'm just not reading you right. Does this count, but not show the 5% from Molten armor (glyphed) or the 5% from the Frostbolt glyph? Were you including other raid buffs or the 2pc T11 bonus as well? I know the FB glyph won't show on your paper doll, but the Molten Armor does.

Can you list your talent points and glyphs that you used in your sim?

I suppose with a glyphed FB reaching the crit "cap" 5% before our other spells, we may need to further refine the stat priority so that we hit the first soft cap at 28.333% (for frostbolt), and then the main soft cap at 33.333% (for everything else). When we get 2pc T11 set bonus (+5% for IL to crit) our Ice Lance would use this 28.333% soft cap as well. Of course this distinction would only be significant if there's a change in priority between 28% and 33%.

eduh · 10/26/10, 11:19 AM

Originally Posted by jak3676

You lost me a bit here. This seems to be contradicting the "stack Crit to 33%" unless I'm just not reading you right. Does this count, but not show the 5% from Molten armor (glyphed) or the 5% from the Frostbolt glyph? Were you including other raid buffs or the 2pc T11 bonus as well? I know the FB glyph won't show on your paper doll, but the Molten Armor does.

I suppose with a glyphed FB reaching the crit "cap" 5% before our other spells, we may need to further refine the stat priority so that we hit the first soft cap at 28.333% (for frostbolt), and then the main soft cap at 33.333% (for everything else). When we get 2pc T11 set bonus (+5% for IL to crit) our Ice Lance would use this 28.333% soft cap as well. Of course this distinction would only be significant if there's a change in priority between 28% and 33%.

I'm assuming its the 5% crit raid buff, or is there no such thing anymore?

Zeldyrr · 10/26/10, 11:20 AM

Interesting sim results Lhivera. I'm not happy about the Ignite numbers at all. If 3 points in ignite give roughly a 5% improvement in damage I think that will move them out of the "marginal dps talent" range for most people. If that's the case, here's how I see the build progression. The first 27 points are locked in:

Base 27 points.

We need 3 more points to be able to get Deep Freeze. Logical choices are: Enduring Winter (remaining 2 points), Permafrost, Ice Shards, Improved Cone of Cold. For the sake of argument, I'll put 1 more point in EW, and 2 in Permafrost. That unlocks DF and once you get 31 points in frost, the subspec choice is now no choice at all:

Base build with Ignite.

That leaves you with 2 free points, if you can ignore Netherwind Presence completely. If not, you have zero free points left.

Only having 3 points to move around in frost seems pretty poor to me. Ghostcrawler touched on this topic (in general) during the Blizzcon Q&A. He argued that people over theorycraft and advised waiting and seeing how the talents played out at 85 in live raids and pvp before jumping to conclusions. So I doubt there is much chance, if any, of this changing before early 2011. I just hope that if they look at raiding frost mage talent distributions and see 95%+ with ignite they will act on it.

jak3676 · 10/26/10, 12:09 PM

It's not too hard to justify 3 talent points for +5% dps, but that's assuming we're already fully talented in Master of Elements and Burning Soul. If we aren't takeing those talents for their own benifit, then we're looking at spending 8 talent points for +5% dps and some situational flexibility - that may be a different calculation for different people. Instead of going with 2/8/31, I wouldn't be surprised to see some people raiding with some sort of 3/2/36 or 3/3/35.

On the other hand Enduring Winter may be pretty pointless if someone else in the raid is already bringing replenishment. Even if we lock ourselves into a 2/8/31 build, that means we have 4 points to spread between Imp CoC, Permafrost, Ice Shards, Enduring Winter, a 2nd point in FFO, and Shattered or Reactive Barrier. That's not a lot of choices, but if you take Enduring Winter out of the calculation you can probably get what you want from the other options.