Optimal opening spell order for bosses?

[Deleted]

I actually start with a prepot, followed by a precasted SF, while at the same time popping SB pretty much ASAP at pull. Even if I pop SB, I don't have to use SF for 15 seconds (at which point SB fades), which means I can use it to refresh ISF from my first SF which is about to run out.

Other than that the opener is usually Immolate -> BoD -> Conflagrate -> Corruption -> Incineratex3 -> Chaos Bolt.

Any comments to this, perhaps, unusual opener? Yay/nay?

I think you would be better off by casting BoD a bit later, as the DPS that BoD does is actually quite low (less than 2k) and as such delaying it is not a big dps loss. Conflagate does much bigger DPS and corruption slightly bigger DPS.

[gakpad]

I think you would be better off by casting BoD a bit later, as the DPS that BoD does is actually quite low (less than 2k) and as such delaying it is not a big dps loss. Conflagate does much bigger DPS and corruption slightly bigger DPS.

DPS is irrelevent when talking about individual spells like this, DPET is what matters. And the reason he uses BoD before conflag is that there's a small delay between when you cast immolate and when the server allows you to cast conflag that isn't worth sitting there doing nothing for to get a conflag off as soon as possible. You're better off filling that small void with your next highest spell in the priority list.

---------- Post added 2012-02-28 at 09:05 PM ----------

i do: Pre-pot>start casting soulfire 2 sec before pull>Use BoD in GCD>Immolate>Corruption in GCD>CoE>Demon Soul>Chaos Bolt>conflag in GCD>start spamming incinerate. I use my DG any time i have my Corrupted mind trinket proc up , Power torrent , Pot and lightweave proc up and heroism if we use that in the begin of a fight ( Like on Ultraxion ). If there's a certain point where im going to need the dps instead ( Like in ph3 of rag hc ) and it'll be within 6 min of the pull i wont use and wait for the required point and use my pot again and then summon it , preferbly with other procs and buffs

Why use demon soul after you cast immolate, and why wait so long to cast conflag?

[Deleted]

DPS is irrelevent when talking about individual spells like this, DPET is what matters. And the reason he uses BoD before conflag is that there's a small delay between when you cast immolate and when the server allows you to cast conflag that isn't worth sitting there doing nothing for to get a conflag off as soon as possible. You're better off filling that small void with your next highest spell in the priority list.

Why would DPET be the one that matters in this case, and not the DPS of the individual spells? He should get (very slightly) more dps by opening with immolate->corruption->conflagate->BoD anyway.

[gakpad]

If you look at the spells in terms of the total damage they do for the GCD you waste, it would look something like

70k immolate -> 90k bane of doom -> 40k conflagrate -> 30k corruption

versus

70k immolate -> 30k corruption -> 40k conflagrate -> 90k bane of doom

Why would you do the latter? (those numbers are not completely accurate but the general damage ratio is about right)

If you did what you said, you'd never cast anything but incinerate because it does the highest dps of your damage breakdown. Herp, can't look at things like that.

[The Trunk Munky]

Gakpad/Ignominious/Mithbrandir have the right idea. Although if your tank/raid is extremely reliable you could go one further and pre-pot after your Soul Fire hard-cast has finished but before combat has started, since it has quite a long travel time and this would save you wasting some of the time on your potion on a fairly low DPCT spell.

[Deleted]

If you look at the spells in terms of the total damage they do for the GCD you waste, it would look something like

70k immolate -> 90k bane of doom -> 40k conflagrate -> 30k corruption

versus

70k immolate -> 30k corruption -> 40k conflagrate -> 90k bane of doom

Why would you do the latter? (those numbers are not completely accurate but the general damage ratio is about right)

If you did what you said, you'd never cast anything but incinerate because it does the highest dps of your damage breakdown. Herp, can't look at things like that.

You would do the latter because delaying corruption by 3 seconds, by your numbers, would yield on average in a damage loss of 3/18 * 30k = 5k. Delaying bane of doom by those 3 seconds again, would yield an average damage loss of 3/60 * 90k = 4.5k. And because both spells are going to have a 100% uptime after the opener, opening with a immolate->corruption->conflag->BoD would give you 500 more damage in this scenario. And certainly, you are going to have more procs up for a bigger bane of doom if you cast it a bit later.

[gakpad]

If corruption lasted 60 seconds, sure, but again you are completely ignoring DPET. There's a reason BoD is higher on priority list. Again, if we followed your logic we'd never cast anything but incinerate because it does more dps than anything else of our rotation. You're still not looking at this correctly.

---------- Post added 2012-02-29 at 05:27 PM ----------

To further expand; changing the priority list in simc to place corruption above BoD results in a ~240dps loss for me. Until you have any actual mathematical evidence that isn't obtained by completely disregarding everything that has been shown to be true I'll continue operating under the assumption that you're wrong.

[Deleted]

What does DPET stand for?

[gracfuldeath]

What does DPET stand for?

Damage Per Execute Time

And I guess I should change what I say for openers. I always just said Corruption > BoD simply because that's how I felt like typing it :x

[smegdawg]

I have my 4pc T13 and am wondering why we cast SB:SF so early. The obvious reason would be for the ISF buff but what about lining up your procs with doomguard?

When i SB:SF off the start my temporal ruin buff goes away before my IotCM Power Torrent and lightweave all end up proc'ing. So usually i delay SB:SF until after immolate and before conflag.

I'm an affliction whore, destro has all ways given me issues...

[Deleted]

If corruption lasted 60 seconds, sure, but again you are completely ignoring DPET. There's a reason BoD is higher on priority list. Again, if we followed your logic we'd never cast anything but incinerate because it does more dps than anything else of our rotation. You're still not looking at this correctly.

---------- Post added 2012-02-29 at 05:27 PM ----------

To further expand; changing the priority list in simc to place corruption above BoD results in a ~240dps loss for me. Until you have any actual mathematical evidence that isn't obtained by completely disregarding everything that has been shown to be true I'll continue operating under the assumption that you're wrong.

I don't see how the duration is relevant on the opener. You will want to have as much uptime on both dots as possible, and after the opener you probably should use DPET to determine spell priority. I don't know where you got this "logic" also, I never said we should use spell DPS to determine what we use past the opener, merely to determine which spell you rather delay by 3 seconds at the start. Also, changing the spell priority list to corr>bod is not what I'm proposing, I'm just saying that using immo-corr-conflag-bod opener will yield a miniscule amount of extra damage. Simulating that is useless as no amount of simulations will be big enough to reliably test it.

[gracfuldeath]

I don't see how the duration is relevant on the opener. You will want to have as much uptime on both dots as possible, and after the opener you probably should use DPET to determine spell priority. I don't know where you got this "logic" also, I never said we should use spell DPS to determine what we use past the opener, merely to determine which spell you rather delay by 3 seconds at the start. Also, changing the spell priority list to corr>bod is not what I'm proposing, I'm just saying that using immo-corr-conflag-bod opener will yield a miniscule amount of extra damage. Simulating that is useless as no amount of simulations will be big enough to reliably test it.

So you're saying that what you're arguing for can't be proven, right or wrong? If something can't be proven correct, it is best to not assume it is correct until there is a way to prove it, and much better to go with what is already proven.

[gakpad]

Let me try a different approach to explaining this to you.

Let's suppose, for the sake of this topic, that Bane of Doom does the same damage that it does over it's full duration right now, except over the same duration as corruption, and has a cooldown that would keep it doing the same damage over the course of a fight as if it had a one minute duration. Would you still want to cast corruption first, even though you'd be doing less damage than bane of doom by casting it over the same duration? If you answered no, what possible logic can you use to justify that decision, assuming away any external factors in this scenario?

The correct answer would of course be to cast BoD first. There is nothing that would suggest we should be deviating from our normal priorities for our rotation at the start of the fight and then go back to "normal" as the same math and mechanics are applied at any point in the fight regardless of when combat started.

What you're saying is that at the start of the fight it is worth delaying BoD for three seconds and cast corruption instead; if your reasoning is correct, what changes later in the fight that you would go back to prioritizing bane of doom? I'm confused, as you're basically making an argument against yourself by not fully explaining your reasoning behind this >_>

[Deleted]

So you're saying that what you're arguing for can't be proven, right or wrong? If something can't be proven correct, it is best to not assume it is correct until there is a way to prove it, and much better to go with what is already proven.

I'm not saying it can't be proven, where did you get that? Just because simulating it might not give reliable information on the issue, it doesn't mean it cant be proven (not saying I'm the one to prove it). Also, I did a similar simulations with 50k iterations and it showed a 14 DPS increase when placing corruption above bane of doom.

Gakpad, well that situation would still be the same as the DPS Bane of Doom does is still the same, its just more frontloaded. I think it's still better to lose less damage by delaying the bane of doom and get more procs up for the bigger-DPET-BoD.

Lets assume that corruption is what it is and it tics for 8000 (8400 with PT up), and bane of doom is a 1 minute dot that tics every 3 seconds, for 5000 (5250 with PT up) and you have 0% haste so they tick every 3 seconds. You open up with soul fire, immolate, corruption, conflag, bane of doom. Power torrent happens to proc within the 3 second window between BoD and Corr. At the 1 minute mark corruption has had 57 sec uptime for 152k dmg and BoD has had 54 sec uptime for 94.5k dmg. Total 246.5k.

Switch Bane of doom and corruption the other way around, Bane of doom gets 57 sec uptime for 95k damage, Corruption gets 15sec+39sec uptime for 42k+104k = 146k dmg, total 241k damage.

Opening up with immo->corr->conflag->BoD yielded more damage in this scenario.

[gracfuldeath]

I'm not saying it can't be proven, where did you get that? Just because simulating it might not give reliable information on the issue, it doesn't mean it cant be proven (not saying I'm the one to prove it). Also, I did a similar simulations with 50k iterations and it showed a 14 DPS increase when placing corruption above bane of doom.

"Simulating that is useless as no amount of simulations will be big enough to reliably test it." -> No amount of sims will be able to reliably prove it -> Cannot reliably prove it -> Cannot currently be proven -> Right now, can't be proven.

Hence my "it is best to not assume it is correct until there is a way to prove it". I apologize if the first statement of "can't be proven" and the later statement of "can't currently be proven" were too conflicting.