Friday, April 18, 2008

The Immolate Effect and Rotation Down-time


For Affliction Warlocks, Immolate is a giant, raging, burning pain in the rear. Since the release of The Burning Crusade, it has been one of the most uncertain of spells in the Warlock's DoT arsenal. Questions abound about where in a spell rotation it should be used, and like a hemorrhoid that just won't go away, those questions remain. Sure, we 'locks have got ideas about when the spell is most useful, but unlike the Corruption/UA issue, nothing has ever been burned in stone. The data just isn't out there, and as a result, the spell still displays an aura of mystery that has many still scratching their heads. Just check the Warlock Forums and see for yourself.

But if you've ever specced Affliction, where Immolate is considered viable at certain gear levels, you already know exactly what I'm talking about. I'm sure you've asked yourself when you should cast the spell, and if it's poor damage-to-mana ratio is even worth casting it in the first place. And I'm also sure you've noticed that you must refresh it every 13.5 seconds, faster than any other 'lock DoT, which regularly cuts into Shadow Bolt spam.

And perhaps worst of all is the fact that Immolate is responsible for a majority of lost GCDs in a DoT rotation. This is usually because the spell ends up competing for refreshment with another DoT. Again, this is an after-effect of it's short 15-second duration, but it can result in some major issues with DPS maximization. Not a good thing.

Unfortunately, even given all of this poor design and it's interrupting character, the fact still remains that for some (mostly those who have just hit 70), Immolate is a DPS boost that should still be used in their spell rotation. And luckily, I'm beginning to understand what it'll take to figure out where exactly it should be.

While compiling some numbers regarding Affliction raid specs, I noticed something. The MaxDPS spell rotation, which puts Immolate at the end of the DoT application cycle, results in two lost GCDs just 12 spells in! What happens is that both UA and Imol end up requiring a refresh at the same time, the exact kind of situation we do not want to find ourselves in:

1) Agony* -> Corr -> UA -> SL -> Immol -> SB spam (41/0/20)
2) Agony* -> Corr -> SL -> Immol -> SB spam (40/0/21)
3) CoS -> SB spam (0/21/40)

Stats: +1000 Spell Damage, 202 Spell Hit, 17% Spell Crit**

*Avg damage will be used.
**Random number generator used to calculate critical hits
________________________________________________

UA

1) 0.00... Agony… 24 secs… 234 dmg every 2 secs… 2805
2) 1.50... Corr… 18 secs… 421 dmg every 3 secs… 2525
3) 3.00… UA… 18 secs… 412 dmg every 3 secs… 2475
4) 4.50… SL… 30 secs… 179 every 3 secs… 1793
5) 6.00… Immol… 15 secs… 531 DD, 253 every 3 secs… 1265
6) 7.50… SB… 2.5 secs… 1574
7) 10.00… SB… 2.5 secs… 1574
8) 12.50… SB… 2.5 secs… 1574
9) 15.00… SB… 2.5 secs…1574
10) 17.50… SB… 2.5 secs… 2361 (Crit)
11) 20.00… Corr…18 secs… 3030 (ISB)
12) 21.50… UA… 18 secs… 2970 (ISB) >UA and Immolate Compete for Reapplication

Surprised by this result, I moved Immolate to the very front of my DoT rotation, hoping that would minimize lost GCDs. This sequence resulted in the equivalent loss of two GCDs after 17 spells or about 30 seconds:

1) 0.00… Immol… 15 secs…
2) 1.50… Agony… 24 secs…
3) 3.00… Corr… 18 secs…
4) 4.50… UA… 18 secs…
5) 6.00… SL… 30 secs…
6) 7.50… SB… 2.5 secs
7) 10.00… SB… 2.5 secs
8) 12.50… SB… 2.5 secs
9) 15.00… Immol… 15 secs
10) 16.50… SB… 2.5 secs
11) 19.00… SB… 2.5 secs
12) 21.50… Corr… 18 secs… -0.5
13) 23.00… UA… 18 secs…
14) 24.50… SB… 2.5 secs
15) 27.00… Agony… 24 secs -1.5
16) 28.50… SB… 2.5 secs
17) 31.00… Immol… 15 secs -1.0
18) 32.50… SB… 2.5 secs
19) 35.00… SB… 2.5 secs
20) 37.50… SB… 2.5 secs

Pretty interesting stuff for sure. And the most important part about these observations is that I'm beginning to think that all of these lost GCDs eventually even out in time, no matter where you place Immolate in your initial sequence. Of course, I can't be absolutely sure without actually testing that hypothesis, but the data seems to hint that is the case.

Be careful though! If this idea pans out to be true, it still doesn't mean that you should cast Immolate anywhere in your rotation.

You want to settle down with a sequence that minimizes early losses of GCDs
. That way, you are more likely to see the mob you are working on go down before ever having to deal with a lost GCD! Keep that in mind the next time you're clearing your way through trash mobs. ;)

Anyway, I'll work more on these rotation calcs this weekend. Hopefully they'll help us find something more conclusive about these effects on rotation downtime. In the meantime, however, use Immolate wisely. It can burn you too. ;)

9 comments:

Upas said...

While I know you didn't say to put Immolate first, I want to warn any up and coming locks NOT to open with immolate. The direct damage it deals will make threat much harder to manage, whereas corruption and UA only tick after the tank has had sufficient time to build aggro.

At least... that's what I think.

Jagoex said...

Hi upas, thanks for commenting.

You're right in that casting Immolate first is not considered ideal. I believe that is most applicable at low-levels, while a Void is tanking, for example. However, it's still manageable if the warlock approaches his threat generation with care.

At level 70, I don't think Immolate's place in a cast sequence is much of an issue, threat-wise. Any good tank should be able to maintain aggro even with a 'lock casting Immolate at the beginning of a DoT rotation.

That said, there are plenty of bad tanks out there, so it's all about DoT'ing with care. =)

Jesyka said...

If you're affliction - you should be focused on stacking mostly +damage and +hit, and forget about crit really. Further more, with the ability to stack tons of school specific + damage, an affliction lock should focus on shadow damage. With things such as the Soulfrost enchant and Frozen Shadoweave armor set... as well as shadow specific off-hands, it's really easy to stack an incredible amount of just shadow damage, over fire - with that said:

If you have very low +crit, the inital burst damage of Immolate probably won't be significant damage, also, it's a continous tick of damage based on your bonus spell and fire spell damage. If you have a weak 700 +fire damage versus 1200 shadow damage, the GCD should certainly go to the shadow DoT over Immolate, inspite of it's ability to have a 'burst' of damage.

Really though, even when I raided as an UA lock... I used Immolate. I don't really recall ever having conflicting GCD wars between Immolate and any other dot... or interrupting my shadowbolt spam. I mean I had things like the half cast meta at the time that allowed for a lot of variance between rotations anyways... because if that proc'd a quick cast Immo, and I had a nightfall proc afterwards... etc.

So, the point of this was to simply state if you have even enough +damage between your shadow and fire... then Immolate shouldn't be forgotten about. Especially if you have a little bit of crit to support it. If you severly lack fire spell damage, though, then maybe consider only using it on longer fights where you will constantly be able to reapply all of your dots... or fights that require tons of movement, since dots will always rule in that situation ;)

Enough theorycrafting and more Dr. Boom testing, imo. :p

Anonymous said...

Its interesting to see it written out. Thanks for posting.

Anonymous said...

I cast Immolate after Corruption and before Unstable Affliction. Will that work out okay?

Anonymous said...

Are you kidding; you're running affliction and casting CoA? If you're affliction and not casting a maledicted CoS you need to delete your character.

Jagoex said...

This is a post about maximizing GCDs; some situations call for CoA (a composition and even raid-specific issue), and so the effect it has on GCDs should be considered.

Think of it as a limited-case scenario.

Anonymous said...

Thanks so much for this information

Anonymous said...

penis