I would like to point out two things... One, is that live testing confirms that a Lava Burst Overload proc is affected by Glyph of Lava Burst while Lightning Bolt Overload procs are not. Two, is that this is a recent change. A while back (six months ago) I ran most of the variables that might affect an Overload proc through the gauntlet, and the result was that neither glyph affected either cast's Overload proc. [Elemental] Cataclysm Discussion - Patch 4.3

I'm not sure what to do with this information, or what I expect anyone to do with this information, but there it is. And I am 99.9% sure of my tests way back when. This was a change.

Your armory indicates that you're using a Stat weight that favors Mastery instead of Haste, that would be the reasoning for you to have a higher up-time with the Elemental 4pc.

As for me I'm seeing the same as Aggixx, a fairly low up-time on the Elemental 4pc and me creating a sadface whenever it falls due to how underwhelming it makes the bonus feel when comparing it to just how awesome other classes set bonuses feel.

Edit: After reading a bit I got curious and ran a few sims just to see what swapping to a Mastery setup instead of my Haste setup currently would do, Mastery came out a whole 23 DPS above my current setup... I'ma thinkin' that's a bit too little of a dps increase to bother messing with it.

Hi friends, this is my first post on Elitist Jerks, so please be kind
I just reached 4pc elem, and wonder how it will change my rotation.
Indeed, when we are under "time rupture" buff, is the priority of ES 6-7-8 stacks decreasing ? Because ES cannot refresh this sweet 750 haste buff (No Elemental OVerload from ES), and is keeping it up for few seconds more better than loosing a Fulmination Charge ?
Thanks
Ayö

It does not change your rotation.

That said, your rotation as noted is wrong. You should be aiming for ES at 8-9 stacks, preferably 9 if you can manage it.

To add an additional question to the 4p bonus: is there an optimal amount of haste and/or mastery you want to get in order to statistically have a 100% uptime on the haste buff during pure LB spamming?

3 LB's per 4th second
With a cast time of 1.33 of LB with 33% mastery I will every 3. cast or 4th second get a proc.
But to get 1.33 cast time I need 50% haste or 6402.9 haste rating which is going to be hard to optain.

2 LB's per 4th second
Since mastery has shown to be pretty equal to haste, at some points a bit better, could stacking mastery to 50%>Haste be optimal since you, statistically, get 1 proc from 2 cast per 4th second meaning 100% uptime during LB spamming.

Is that a possible way to do it?

Or is the right way to calculate the average amount of procs/4th second? With this method I would simply just have to found out whether haste or mastery averagely results in more procs.

Base haste: 5% (WoAT) (cast time 1.9047619 seconds)
Base Mastery: 16%
Average amount of casts/4th second: 4/1.9047619 = 2.10000001 casts
Average amount of procs/4th second: 16% of 2.10000001 = 0.336000002 procs

Adding 1% of haste, then mastery

+1% haste
Haste: 5% (WoAT) +1% (cast time 1.88679245 seconds)
Mastery: 16% (still base)
Average amount of casts/4th second: 4/1.88679245 = 2.12 casts
Average amount of procs/4th second: 16% of 2.12 = 0.3392 procs

1% mastery
Haste: 5% (WoAT) (cast time 1.9047619 seconds)
Mastery: 16% (base) +1%
Average amount of casts/4th second: 4/1.9047619 = 2.10000001 casts
Average amount of procs/4th second: 17% of 2.12 = 0,357000002 procs

Additional procs from +1% haste: 0.3392-0.336000002=0.003199998 procs
Additional procs from +1% mastery: 0,357000002-0,336000002=0,021

More procs from +1% mastery than +1% haste: 0,021-0,003199998=0,017800002 procs

Correct?

The first way is the proper way to think about it (you calculated incorrectly) because procs have different values. A proc with 1 stack is worth 500 haste for the next 4 seconds while a proc at 2 stacks or 3 stacks is worth 750 haste for the next 4 seconds. However, the uptime of your 3stack depends not only on your chance to proc a stack with 2 or 3 stacks, but also on your chance to have a 2 or 3 stack (which is dependent on your chance to proc at 0 stacks and 1 stack). Since all of your casts that can proc the buff have the same travel time, you can model it as a moving 4sec cast window.

WRT maximizing 4pc buff uptime, you want to stack mastery once you are at a haste point that maximizes the # of casts per 4sec window that you can achieve in your gear set. There are many haste "breakpoints" for this, but anyone with decent gear should be able to hit 2LB+LvB/4 at 0 stacks in a raid.

Whether haste or mastery gives you better overall performance is something that you should sim for each boss fight with your current gear set, spell priority, etc. 750 haste is pretty valuable, and investing points in order to maximize the uptime of your free haste buff seems good so I would guess that with 4pc mastery will generally win once you hit that 2LB+LvB/4 point.

Your calculations under the first section are incorrect because you have not used probability correctly. The probability of getting a proc after 3 casts with 33% mastery is not 100%. It is the inverse of the probability that you will not get a proc on 3 casts or 1 - (2/3)^3 = 1 - 8/27 = 19/27 or ~70%.

There is definitely optimal stats for maintaining 4pc uptime with any gear set, but whether those stats are also optimal for damage output is a different question answered only by simulation and more simulation.

Speaking of Damage output, what, if any, changes to rotation/ CD usage needs to be made for the Madness fight. I feel we as elemental are a a big disadvantage due to us not receiving full benefit from the Spellweaving mechanic. I have heard that some Shamans are doing 'fine', but i felt that even pushing the limits of what i could do, i just could not keep pace with the rest of the DPS on that fight.

Doing 'fine' is a delicate way of saying you're not going to beat out the classes that can use an instant cast AoE spells see: Arcane Explosion. I attempted using Earthquake a few weeks just to see how well it worked out and it just didn't work out at all, it only procs spellweave on the initial cast. Chain Lightning even though not the best it did much better as long as you don't target the blood in the front of the pack (every one targets that one, it dies way too fast) target the blood in the backish and you'll be able to CL without having to swap a target.

On the Green (Ysera) platform if you stand on the far left of the platform you will be in range of both arm tentacles, this lets you Dual-Flame shock for more then just when the Mutated Tentacle is up. (Note: I found trying to keep up three Flame shocks was a loss of DPS because of all the target swapping to do so, not to mention the camera spinning)

"Since the Elemental takes a snapshot of your stats at the moment it's dropped, it will gain substantially from all your procs being up at once."

Does this work with searing/magma totem as well?

No, those totems update on every shot/pulse with whatever stats you currently have. They are also affected by buffs you gain that don't help the Elemental at all (Concentration on heroic Majordomo, for example).

very simple question, can't seem to find an answer. Lvb>7-9stack fulm, but in the rare occurrence we reach 4-5 seconds remaining on flameshock, with LvB coming off CD that same second, what should we do?

1. Fulm so that FS can be refreshed in time and no LS stacks are wasted but possibly lose 1 lavaburst by missing a surge proc.

2. Lavaburst and lightning bolt spam until FS needs to be refreshed so that we miss 0 lavasurges and keep a 100% uptime on FS - but lose many lightning shield charges.

3. Lavaburst and then fulm at ~3 (or less with lucky surges) seconds left on FS so we get the lavaburst, get the 9 stack fulm but lose 2-4 seconds on flameshock uptime and possibly more lavabursts as a result.

This spec gives more dps than the simulationcraft one, and far more mitigation to boot. I've simmed my own character, and the bis profiles for a moderate gain.

Talent Calculator - World of Warcraft

PS to the following posts: simulationcraft = maximum dps on paper. Opinions will vary on usefulness, enjoy.

What is your definition of 'moderate' dps gain? I used your spec and simmed myself, and it was only a 200dps increase. Also, i am somewhat peturbed at the skipping of the improved shields. I can understand it to a degree seeing as it only makes up about 7 % of your dps, but all you seem to gain is Fire/Earth ele and searing time as a tradeoff of 15% of fulmination's damage.

Worse than that is the loss of Ancestral Swiftness with that spec. On paper 3 extra points to dump in Ancestral Resolve or Spark of Life seems like a survivability gain. But aside from Ultraxion, being able to move out of fire 15% faster is more significant in practice.

In my experience the fire ele spec is situational and also quite gear dependent. I was seeing a simmed dps gain of almost 1500 at one point while gearing this tier but the benefits are much smaller now. You will have to sim your own gear and decide if you feel it is worth it because there are some quality of life issues with the spec.

1. If you are wiping quickly and rerunning, 1/2 of your progress attempts will have you doing garbage dps without the elem - pretty frustrating.

2. 15% movement speed is as Jessamy says far better than it looks for surviving and positioning yourself well to dps effectively.

I was confused that Jessamy mentioned Rolling Thunder in this case both in this thread and in Simple questions, because the wording of the talent seems to be exact about 60% chance for both LB and CL. So I decided to run some tests on a target dummy. Here are the results:

Spell name	# of casts	Overloads	Rolling Thunder	Rolling Thunder %
CL	40	2	26	62%
CL	40	6	32	69%
CL	40	4	30	68%
CL	40	5	29	64%
CL	40	10	34	68%
Total	200	27	151	66%
LB	40	17	37	65%
LB	40	11	34	66%
LB	40	16	31	55%
LB	40	12	31	60%
LB	40	16	35	62%
Total	200	72	168	62%

Each row represents separate attempt. In each attempt I casted 40 times, then checked actual number of spell hits using Recount (thus calculating number of Overloads) and then counted Rolling Thunder procs in combat log.

Judging by results CL does not have a lower chance to proc Rolling Thunder on single target. It even ended up to be higher for CL than for LB but this can probably be attributed to RNG.
Chance to proc an Overload is indeed noticeably lower.

This should not change the overall conclusion about LB vs CL but I was curious about reasoning.

Not sure where the RT information came from, it's only ever been overload that had a reduced proc count.

I'm using my amazing moderator privileges to bump this thread and keep it on the front page.