Well, thing is, fire mages will stack every single cooldown known on earth during that 20%. Its arguably a lot more than just 2.7% fire damage, given that typically I will use:
- molten fury (+20% damage)
- drums of speed (+5% haste)
- combustion
- icy veins (+20% haste)
- skull of gul'dan (+11.15% haste)
- hex shrunken head
- bloodlust
- destruction potion
- flame cap
(in total: 1.5s fireballs, combustion, +411 spell damage, +2% crit, +20% damage)

I am sure you can see why COE would be awesome for the last 20%.

Remember that all those cooldowns simply multiply the effect of those 2.7%. So even if you do double DPS during the last 20% (which is probably not realistic considering your cooldowns won't stay up for the entire 20->0 period) the effect is only 5.4% dps increase for the mages at the cost of 2.7% dps to your shadow users (btw isn't it additive to scorch and misery making it less than 2.7%?).
Swapping malediction to the CoE is a lot less straight-forward than you make it seem. Not saying you should or shouldn't, but it's definitely not an obvious thing, especially if u run with 3 locks, 3 mages and 3 shadow priests in your standard raid, as the lack of malediction on CoS will cause 6 people to lose dps compared to 3 peopel that gain, so if the gain is doubled during the last 20% you only gain as much as they would. It just doesn't look like swapping malediction to CoE would be a good idea, but there's a lot more you need to calculate to really tell for sure what's worthwhile, evev the trinkets SPs/locks have and when they used/will use it can make a difference, nontheless their actual dps etc.

Indeed, all factors must be considered in any given raid, but the point is simple: fire mages get a benefit from Molten Fury, which is further amplified by CoS (or CoS is amplified by Molten fury, however you like to think of it), that shadow users don't. Each guild will need to examine their set up, the DPS being pulled by each member (remembering to account for differences in who got Malediction, etc.) to accurately model wether it's more useful to trade 2.7% of the raids shadow damage at 20% and below for an increase of 3.36% (2.7*1.2, for the molten fury effect) in the raid's sub-20% fire damage.

Keep in mind it's most likely to not be the case, though, as just the fact that there are generally twice as many shadow users as mages, and that you're much less than doubling your dps with molten fury+saved cooldowns. While it's best to just run full numbers, it's just not likely to be worthwhile to swap the malediction curse without an extremely wierd raid composition.

What is the exact cast time of your fireballs with all of that & your passive +haste? I've never bothered to look to see if I'm hitting the GCD on fireball during everything being popped. Since you have such a large increase of damage during this point...the value of haste would be 0 once you hit the GCD...and lower the overall value of haste on all fights...but more noticable on sub 3'30" fights.

Oh that's an easy one.
Using [Zhar'doom, Greatstaff of the Devourer][Pattern: Bracers of Nimble Thought][Ring of Ancient Knowledge][Pattern: Drums of Battle][The Skull of Gul'dan]
(or [Mana Attuned Band] ideally - it won't really affect the numbers much)

[top] 369 spell haste

~ 23.50% haste

3 / (1 + 0.2350) / 1.3 / 1.2 = 1.55s per fireball

(in other words, the max haste you can use with bloodlust/icy veins is 28.21% =~ 443 haste)

EDIT: Also, I don't think the debate is about swapping maledictioned COS/COE on execute range. It's more when you have 1 mage and 3 locks (1 using COD), the last COD/COA should be swapped for COE in that case I think.

Do the skull and the hex head not put a cooldown on the other? Would you possibly get better dps out of the 225 dmg from the [Serpent-Coil Braid] stacked instead of the haste from the skull? I doubt it, but I wonder how close it would be from a stacking perspective.

I'm assuming you are using molten armor and have a shadow priest. How long is that sustainable w/ using flame cap/destro pot at the final 20%

As was said many times over, if you activate the skull before any +dmg activable trinkets then it stacks. In other words, its as if all +dmg increasing activable trinkets will trigger a cooldown on other trinkets, but the trick is that some trinkets, like skull of gul'dan, aren't +dmg. In this case, for as long as you pop skull first, it stacks.

As for serpent-coil braid, using it implies you lose on flame cap (and use a 'bad' trinket assuming best gear avail).

Yeah I suppose with a 3 lock 1 mage setup you would most likely use CoD for most of the fight but be better off with CoE for the last 20%, probably even regardless of whatever cooldowns you're stacking. Isn't that a rather odd raid setup though?

Well, when I made that statement, I was thinking more along the lines of a 3 mage/2 destro/1 afflic/1-2 spriest raid, and curse swapping.

Your criteria here is probably more applicable to a typical raid. Regardless, the math should be similar, and it depends on how much your mages outdo your locks in that 20% range.

You warlocks just dont quite understand the dps gain we get at 20% when we pop all this. We're talking 10,000+ fireballs BEFORE ignite every 1.55 seconds for 20 seconds then a slight taper off for the rest of the fight but still high. I wish WWS would model it, but during that 20 second time period i was topping 8000 dps with PI.

On a side note, with 83 passive haste, drums, and heroism i'm getting into rolling ignite range. I wonder if not using icy veins here is a dps nerf though.

As far as I know rage is a short fight. Since it's such a wash it'd probably be best if he swapped for CoE after his last CoD tick as it would be a combination of <20% HP and <1min fight length, wouldn't it? This would shift matters even more in the favor of "CoD until 20% then CoE". Especially considering you did 307.7k before the 20% and 92.3k after 20%. Depending on when his CoDs ticked it could be anywhere between a small to huge benefit if he'd swap to CoE at last 20% or at least after his last CoD if he casted it well before the last 20% (or even CoA so he can swap to CoE at 20% although that's harder to calculate).
Bottom line is CoD->CoE at the end would've netted more DPS than just CoE as well as more DPS than just CoD, assuming done correctly taking fight duration into account.

Yeah typically Rage is a short fight for us too, just had a lot of DPS die.

But anyways, I was curious how you guys are modeling CoE. Is it just a flat 10% damage boost or does the 88 less resistance to cold/fire effect anything?

The only true difference the 88 less resist makes is on Supremus (fire) and Rage (frost), so I'd assume it's not even worth trying to include in the modeling.

It also helps a lot on Council due to the resistance aura. Have a heck of a time getting our locks to use it there due to one mage taning and leaving only 1-2 dpsing, but without it I usually see around 15% partial resists. With CoE it is usually around 5%, so CoE is a 20% boost on that fight. That makes it viable with only one mage dpsing.

Ah yes, I'm usually the tank for council, I think I've only dps once so far. Totally forgot about that aura.

I don't know why you would use kill times using darkmoon faire buff (Thottbot World of Warcraft: Sayge's Dark Fortune of Damage) with 15 leatherworkers as your baseline assumption of what to expect.

Yeah, looking at the top 50ish worldwide parses is not a reasonable assumption of what most guilds are capable of doing. Moreover, whatever the threshold for double IV is (if it exists), there's still plenty of fights that will never be realistically short enough qualify, and even for the encounters that might qualify, there's no guarantees you'll be able to output the needed RDPS. Sometimes it just isn't your night, whether it's sloppy play or just bad luck or whatever else, and you struggle even with fully farmed fights.

To add on to previous talk about PI, I certainly don't believe it's worth speccing out of CoH, but man is it fun. Last night on Winterchill (was getting CoE there), one of our priests was PvP specced and PI'd me when our bloodlust went just before 20%. I pulled ~8500 DPS for the duration of my cooldowns, all crits above 10k. The DPS from that sort of stacking is just insane.

Assuming Molten fury range is approximately 15% of the actual kill time, you need a 4:30 kill to use double veins entirely within MF range. That kill length coincidentally would be long enough to get the first veins in after stacking scorches. The reason I believe it still not better above 4 minutes would be that the second veins is not nearly as powerful as the first since trinkets / destro pots would be on cooldown. Although having veins up through an entire lust all during molten fury range sounds powerful at least.

I must remind people that the ultimate goal of this theorycraft is to lead your guild to victory in whatever progression they're attempting to win.

On that note, when learning progression, there's 4 scenarios concerning molten fury.

1) You kill the boss just fine and dont get the full duration of molten fury. A dead boss is a dead boss you've acheived your progression goal.
2) You're dead, in which this theorycraft doesn't matter.
3) You're one of the last people standing, and you're doing whatever you can to beat enrage timer. In which good theorycraft can make or break the kill (Ouro anyone?).
4) You and the entire raid's up but the fight is so intense you're struggling to kill it and good theorycrafting makes or breaks the kill (Ouro again anyone?).

My point is if you kill a boss except for epeen what does it really matter if your guild gradually kills him faster and you get less molten fury time?

Using CoD over CoR is absolutely retarded. CoR is the highest DPS curse on most fights, CoS second, CoE 3rd assuming multiple mages, then CoD.