This site makes extensive use of JavaScript.
Please enable JavaScript in your browser.
Live
PTR
10.2.7
PTR
10.2.6
Beta
Arcane Mages and Haste
Post Reply
Return to board index
Post by
354213
This post was from a user who has deleted their account.
Post by
243007
This post was from a user who has deleted their account.
Post by
354213
This post was from a user who has deleted their account.
Post by
299735
This post was from a user who has deleted their account.
Post by
243007
This post was from a user who has deleted their account.
Post by
Squishalot
Yay, dataz! \(^o^)/
The reason I asked on mana is simply because I'm not game enough to try to theorycraft it without access to Wowwiki at work ;p
And my mage is 79 (and 9.5 bars...), so I've never known what the values actually are.
Ok, let's see how this looks:
DPS (while casting):
Haste AB4 MBAM AB4 AM AB4 ABarr
0% 3375 2631 2667
10% 3712 2894 2934
20% 4144 3158 3200
30% 4490 3421 3467
40% 4835 3684 3700
50% 5180 3947 3925
60% 5658 4210 4146
As percentages:
Haste AB4 MBAM AB4 AM AB4 ABarr
0% 100% 100% 100%
10% 110% 110% 110%
20% 123% 120% 120%
30% 133% 130% 130%
40% 143% 140% 139%
50% 153% 150% 147%
60% 167% 160% 155%
So in a very rough nutshell, we can see the impact of haste at the key breakpoints highlighted earlier, valued at approximately 3%. So what we could say is that by crossing over a haste breakpoint, we can 'add' that much additional haste to the calculator, in effect.
As an example, using Gusiedon's gear example before, we have a cloak with an extra 100 haste, or a cloak with 130 crit. Each haste is worth 2.2, each crit is worth 1.8. So on the face of it, with no breakpoints, each the first cloak is worth 220, and the second cloak is worth 234.
Now, knowing that achieving the breakpoint provides an extra 3% haste, we can confidently state that *IF* the cloak takes us over that breakpoint, the value of the haste isn't an extra 100 haste, but in fact, an extra 198 haste (being 100 base on the cloak + 98 haste rating representing the extra 3% DPS using AB4 MBAM). So the first cloak is actually worth 436 weighting, because it crosses a haste breakpoint.
Mind you, that's only if we could guarantee that our MB will proc within 4 ABs each time. Let's consider the times when it won't proc.
------------------
If AB3 doesn't result in a MB proc, then we need to make the decision of whether to cast an un-procced AM after 4 (with the 40% chance that we'll get a proc), or cast ABarr, and suck it up.
From the tables, we can see that at lower levels of haste, ABarr is the higher damage option, simply because you're not wasting 5 seconds sitting there channeling. Practically, there are the other benefits of being able to reposition while the GCD is up, and the fact that you're less likely to be interrupted (read: pushback, not actual spell interruption) while casting.
This changes when your haste increases. At some reasonably high level of Haste (between 35-40% after talents and raid buffs), ABarr no longer gains any benefit from Haste, on account of the GCD being floored at a minimum of 1 second. Looking down the % table, you can see that each 10% of haste no longer improves DPS by a flat 10%, as Haste improvement is only relevant to the AB portion of the damage.
So does this mean that we should use ABarr? Not really. ABarr is more mana consuming (on a per-second basis), so the higher DPS is offset by higher mana usage. If we compare the mana costs, for the MPS of an AB4 ABarr rotation (366mps at 0% haste), you could use AB5 AM instead (379mps at 0% haste), which would yield you higher DPS (2722 vs 2666), as well as giving you another 40% chance of getting an MB proc and saving you the 953 mana cost of the AM in addition to the extra damage.
So what's the optimal rotation?
ABx3, start casting AB4
-> If you have a MB proc, MBAM
-> If you don't have a MB proc, AB5 then AM.
Does this apply at all haste levels?
Yes.
So what's the value of our haste?
Earlier, we said that the 98 extra Haste Rating (equivalent) from hitting the breakpoint was conditional on an AB4 MBAM rotation. The Haste Rating equivalent from hitting the breakpoint on AB5 AM is
zero
. This is because the breakpoint benefit only comes from squeezing a 5th spellcast into the Pushing the Limit buff. By contrast, the Haste Rating equivalent from hitting the breakpoint on AB5 MBAM is about 2.5% or 82 Haste Rating - a bit less, because you're wasting time on a 6th unhasted spell.
Combining it all together, looking at the first breakpoint:
AB4 MBAM will occur 78.4% of the time
AB5 MBAM will occur 13.824% of the time (being "no MB proc after 3, and MB procs on either 4 or 5")
AB5 AM will occur 7.776% of the time (being "no MB proc after 5")
How long will we spend in each (@ 0% haste)?
AB4 MBAM runs for 10.203 seconds
AB5 MBAM runs for 12.384 seconds
AB5 AM runs for 15.265 seconds
Average rotation time is 10.203 * (0.784) + 12.384 * (0.13824) + 15.265 * (0.07776) = 10.898s
Average time spent in each is 7.999s, 1.712s and 1.187s.
So average Haste Rating equivalent is (7.999 * 98 + 1.712 * 82 + 1.187 * 0)/10.898 = 84 Haste Rating
N.B. This assumes no usage of Bloodlust, Black Magic procs or any other 'on-use' haste effects. If you look at Icy Veins, for example, the 84 Haste Rating equivalent will not apply during the 20 seconds, because the Icy Veins proc will take you over the first breakpoint (400 as calculated by Gusideon) anyway - the extra haste has no value. The corollary is that the second breakpoint (1100 haste) is *only* meaningful with Icy Veins, and as such, you can only apply the extra haste rating during the 20 seconds when theorycrafting.
So as an example, in a 5 minute fight, you might use Icy Veins 2 times and Bloodlust once. As a result, there are 80 seconds in the fight where the breakpoint is achieved by non-gear buffs, and so, the Haste Rating on the gear is only worth the base amount, as it's not required to reach the breakpoint. So the 84 bonus haste rating equivalent for reaching the breakpoint should be scaled back by uptime of 220/300, to 62 Haste Rating.
So at the end of the day, the cloak, if it tips you over the 1st breakpoint, will give you 162 Haste Rating equivalent over the 5 minute fight. Edit: Relative to the 130 Crit Rating on the other cloak.
Post by
200565
This post was from a user who has deleted their account.
Post by
243007
This post was from a user who has deleted their account.
Post by
Squishalot
Assuming that Gus's numbers on the haste breakpoints are correct (I'm not sure):
Optimal efficiency / DPS rotation (assuming T10 2pc):
ABx3 and start chaincasting AB4
If you have a MB proc: MBAM
If you don't have a MB proc: AB5 AM
AB spam DPS < AB4 MBAM DPS
Based on that rotation, if your gear takes you over 700 haste or 1100 haste, it's worth a bonus 62 haste rating (roughly) in additional DPS because you can squeeze an extra AB into your Pushing the Limit buff (but only if you time it well enough).
If you don't have T10 2pc, then there are no breakpoints to speak of.
Edit: Rephrased, and used the correct 1st haste breakpoint =_=;;
Post by
Squishalot
In terms of mana costs for rotations....
Evocation lasts ~6 seconds raid/gear hasted. This would suggest that needing to use a single
extra
evocate across a 5 minute fight would cost you 2% of your activity time.
In that respect, blowing your mana on rotations that require you to evocate an extra time would need to be justified by an extra 2% DPS.
Given a 20k mana pool, evocation will give you 12k mana, which is approximately 40 mana per second. Given that the jump from AB3MBAM to AB4MBAM is more than 40 mana per second, for a ~2% increase in damage, this would suggest that it's not optimal as far as whole-of-fight value is concerned, assuming a 5 minute fight.
This is probably how EJ comes to the conclusion that you should be aiming for AB0234MBAM, and just starting the new stacks as soon as possible.
Thoughts, anyone?
Post by
303016
This post was from a user who has deleted their account.
Post by
243007
This post was from a user who has deleted their account.
Post by
Squishalot
Well, at 34k, it's about 68 mana per second over a 5 minute fight. That's still less than the increased mana cost of maintaining a 4stack rotation. And the same again between 2 and 3 stack rotations.
I think EJ's called this one perfectly.
Post by
200565
This post was from a user who has deleted their account.
Post by
243007
This post was from a user who has deleted their account.
Post by
Squishalot
This is because of time lost when evocating, and small diff between 3 and 4 stack rotations in terms of dps gain. As far as I understood it - You're better of with casting MBAM earlier on non-burst periods of fight, than having to evocate more often.
Yep, great summary.
Post by
535364
This post was from a user who has deleted their account.
Post by
Squishalot
4 Seconds is what I try to aim for.
If you're getting 4 second Evocations, then stop wasting your Bloodlust / Icy Veins buff on Evocation ;)
Anyway, see my follow up:
Well, at 34k, it's about 68 mana per second over a 5 minute fight. That's still less than the increased mana cost of maintaining a 4stack rotation. And the same again between 2 and 3 stack rotations.
We're talking DPS of 4800 vs 4840 unbuffed. Across a 300 second fight, at a 1% difference, the maximum you should be willing to give up is 1% of that time evocating for a 1% DPS gain. 3 second evocates should be the 'target', for it to be worth shooting for 4 stack AB all the time.
Post by
404185
This post was from a user who has deleted their account.
Post by
243007
This post was from a user who has deleted their account.
Post Reply
You are not logged in. Please
log in
to post a reply or
register
if you don't already have an account.