Bid on cloth. The more you do so, the more clothies will whine, the higher the chance that people will post about "elemental shamans stealing my loot" on official forums and thus the more likely that Blizzard will realize it's a pressing concern.

Ideally, it will push Blizzard to redesign either elemental or resto and make us use same gear.

I picked up the new totem today after we ran ToC, I didn't do too much testing, but I found a few things.

First, it looks like there might be an ICD. After 140 Lightning Bolts (note that this is including LO procs, since it's just off recount. I don't know if the totem can proc off of them, although I didn't see any) I had 20 procs. I never had a proc within 6 seconds (not necessairly 6 second ICD, but something around there), the first 19 were with 5-3 seconds left on the duration, #20 being at 1s left. This was done with a LB spam.

The proc rate looks like it might just be 70%. Whatever it is, uptime is very high on training dummies. I would expect it to fall off a decent amount for any boss with lots of movement.

Running my actual rotation, still on a training dummy, a refresh generally didn't occur until ~2 seconds, and it fell off once, for about 2 seconds (out of a 128 second fight, not very long, and only one test)

Time for more Haste Stuff.

So people don't seem to understand haste very well. Is it too complex? Am I not explaining it well enough? Who knows. So what I've done is make some pretty graphs to illustrate to people what happens with your cast times under certain conditions, and divide the rotation time into "Time spent on 1.5 second casts or GCD use" and "time spent casting LB".

First off, some basics.

• 1% Haste = 32.78998947 Rating
• Rating buffs stack additively with rating from gear
• Percentage buffs stack multiplicatively. eg: Cast/(1+Rating%)/(1+%buffA)/(1+%buffB)

Haste %	No buffs	WoA	WoA+Aura
20%	656	469	360
30%	984	781	662
40%	1312	1093	966

Ok, so now we've seen the sort of haste rating values you need to get to 40% haste without BL, lets see how haste affects spells/rotations.

Firstly, cast times.

Here we can see how cast times are affected by our 20/30/40% haste figures, with and without BL. As you can see, the GCD based spells (FS, LvB) are already GCD capped under BL with 20% haste, but remain uncapped outside of BL. LB is still not GCD locked at 40% with BL (which requires at least 966 haste rating to achieve).

Next, lets take a look at the overall rotation.

Here you can see a comparison between how much time is spent on those GCD spells and LB. Obviously the impact of the GCD spells on the rotation is fairly minimal (~18% with 20% haste), which means that even once you've hit 1 second GCDs, haste will still be a valuable stat (ie: you'll still prefer it over crit, but spellpower will generally beat both hands down).

Lastly, lets take a look at the resulting dps of a rotation as you scale with haste. Previously my automated rotation calculations would have dips around where you'd add in extra LB casts, but earlier in the week I figured out how to fix that. Here is the resulting graph (my calculations are Method 1. Method 2 was Helmchens as per the Elemental Rawr thread).

I haven't done any detailed investigation yet, but it suggests that for certain small haste values, it may be better to gem for haste rather than spellpower (however, since this is a small 0.25% range, it will probably not be worth worrying about).

Just a quick question Bink why is your graph insinuating that 40% is the cap? I was under the impression that 50% character sheet (~38.7% or ~1270 rating with raid buffs) was the true GCD cap as modeled by
1=1.5/(#CumulativeHaste) or roughly 1=1.5/(1.03*1.04*1.387)

The equation seems accurate given my in-game testing observing the cast time changes in Astral Recall.

Just wondering what I'm missing here.

I just used 20/30/40 as example figures, mostly because I didn't want to produce massive graphs. Plus 50% haste is 1269-1639 haste rating, which I think is fairly impossible at this point in time. Any cast times sitting at 1 second are the "WoA+Aura+BL" conditions (ie: has an extra 30% haste)

I have to say I am still very confused on the Haste issue,

I dont understand how the scaling between haste and dps on your graph is mostly linear when the CD on Lavaburst should be limiting it more. For example if I use a FS, LvB, LBx5, LvB, LBx4 rotation even if I increase my haste my dps shouldnt increase unless I gain enough haste to add an additional LB into the rotation which I assume would require a figure at which the Cast time on LB is short enough to add another in or a point at which the LB Cast Time vs the LvB CD results in LB being profitable to use at the expense of not instantly using LvB when its off CD.

How come there are no flat plateaus on the graph? I might be missing something here but if I cant cast any more spells with 33% haste than with 32% haste during the 8s CD of LvB then the only thing my dps will scale with is my crit and spell power, that 1% haste is useless no?

Edit: Shouldnt this result in haste being extremely useful at certain % and useless at other % i.e. you have increases as soon as you can fit in extra spells which might be linear if you are clipping LvB ( more haste would result in less clipping)or just straight jumps if not clipping it but entirely static dps at other times purely with regards to haste as in the areas where Methods 1 and 2 are the same should be flat and not resulting in a dps increase.

If you're waiting for the cooldown to be up rather than casting another spell, then your assumption would be correct. However, it's been pretty well shown that even if you only have 0.5s or so left for your LvB cooldown, you're better off casting another LB and not hitting LvB exactly when it's available rather than waiting for the cooldown. In effect, your effective LvB cooldown is somewhat higher than 8s. Haste would shorten the casting time of the between-LvB period, reducing that effective cooldown so that you're getting closer to the ideal 8s interval.

I take your point that movement makes haste more valuable as you could squeeze in extra casts before you have to move again.

But what does it matter for a situation where there isnt much movement if my DPS per Cast ( DPSC?) is higher if I cant cast additional spells due to a static CD thats unavoidable. I still can only cast 2 LvB and 1 FS every 16s roughly the fact that it takes less time to cast them does not result in an increase in dps unless I can cast more LBs in the interim. At certain points its just additional "dead time" with no gain.

The issue I have is that the graph above seems to indicate a complete successful rotation so I do not understand where the DPS increases are coming from. Even if you get a higher DPSC from haste, without additional spells there is no personal DPS increase which is what the graph is labelled as tracking as I understood it.

Am I misunderstanding it or can someone show me how my dps would increase while not gaining any additional spells being cast ignoring the whole movement situation for now?

True but for that first 0.5s that would be the flat are Im refering to where haste has no benefit. Once you hit the 0.5s then it has benefit, I suppose that its likely they overlap between the first half and the 2nd half of your rotations where you could be 0.5s off on 1st half and then 0.5s off on the 2nds leaving a I dont know 0.3s point at which neither are available but thats what I expected to see on the graph, an amount of haste that would not provide an increase in dps, yet the graph shows that any increase in haste results in an increase of dps for a continuous rotation. Maybe Im wrong about the overlap and that one they start overlapping it would continue to overlap as 1 gets closer to the ideal and then the other ones does and thats why there is always a dps increase.

I think you're misunderstanding me.

I'll make an example. Let's suppose your LB takes 1.25s to cast and does 5k damage, and your LvB takes 1s to cast and does 5k damage. Let's also suppose that FS does 5k damage total.

Your rotation would be FS - LvB - LB x 7 - LvB - LB x 6. You would do 90k damage over 19.25s, for 4675 DPS.

Now suppose you increased your haste so that your LB takes 1.2s to cast. Your rotation would remain the same since you couldn't fit in another LB, but your overal time to complete that rotation would be 18.8s, for 4787 DPS. Part of this is due to the effective LvB cooldowns of 8.4s and 8.2s, rather than 8.75s and 8.5s.

Basically, Higher DPS does not always equal out to be more total damage done.

If you look at the graph, the parts where the blue line is on it's own is the point where you've just added another LB cast into the rotation (~23%) and the point where the gradient flattens out is where you end up fitting 5 casts inside the 8 second window. That 25-29.5% haste window is where the only increase comes from reducing the overall cast time rather than adding additional LB casts in.
If you want to look at it in further detail, click here to see the sheet.

Here is the cast time sheet (the bar graphs). I've added 50% & 60% in, but it doesn't really change much (as I've said, it's impractical to get that much haste).

Also, remember that I'm not doing any LvB vs LB comparisons here, but rather comparing complete rotations. Even though LvB and FS have a higher DPSC value than LB, since you cast them less frequently they only account for 800-1000 dps.

*bzzzt*

DPS is Damage Per Second. You get that from taking Damage, and dividing by time. Increasing D while keeping T constant = higher DPS. If you're talking about the total vs effective dps (where your value of T is either total time or the time spent casting), then yes, that may be correct, but making sweeping statements like that isn't very helpful.

Didn't see this mentioned here... but the changes to wind shear have made it so elemental are absolutely amazing at interrupts.

Removing it from the shared shock cooldown on top of it being off the gcd means that with decent timing you can interrupt with very little dps loss.

Personally I have it set to a /stopcasting macro. (just /stopcasting, next line /cast wind shear) and on fights where normally a rogue rotation was needed for kicks just one shaman can take over the job. With things that need to be interrupted and have a cast time, nearly every one of our spells will have a faster cast time. So you don't even have to /stopcasting to get it in and I don't see it dropping dps at all. A for instance would be the Lord Jaraxxus fight... One elemental can pretty much reliably interrupt every single fel fireball without losing dps.

At 6 seconds cooldown, it's the interrupt (with school lockout 2 seconds) with the lowest cooldown in the game. That and it's mana cost is minimal. It's nearly good enough to think about the 1 second reduction of shock and shear cooldowns talent.

I still find it useful to use a /stopcasting macro for fights like Iron Council hard mode, where the whirl needs to be interrupted as soon as possible.

In saying that, we have an enhancement shaman who handles interupts as they are even better than we are.

The downside is that Wind Shear has a 25 yard range, and everything else is 36+

Delete Plx, Already posted.

My guild doesn't have an enhancement, but yeah they are a better choice.

Yah I find myself readjusting range a bit now, just because timing wise it's more or less free interrupts from us.

It's nice to have an actual reliable interrupt now.

From what Ive seen I lose very little dps.

Anything with a cast time is invariably going to have a longer cast time than I do.

Being that Wind shear is off shock cooldown and off the gcd... you can just hit it right after a cast and move on to your next one.

I've stuck with using the stopcasting macro for it just so I can stop a cast and interpret something, but by timing it with quartz (I have main cast bar in normal place and enemy cast bar right above it) I tend to not lose any dps and still get off interrupts.

Sadly there wasn't much else in 3.2 for elemental shammies ;p Other than gear additions, this is really the only thing that was a positive in it. We run with 2 elementals and just having us switch to general interrupts has definately been a positive thing. I find it to be extremely useful on Auriaya (instead of fear warding rogues and/or having them use pvp trinkets) backing up warrior tank, Iron council (lightning whirls), Lord Jaraxxus (lock down all fel fireballs) and just in general versus any caster or healer. I'm guessing the fact that I can actually run my dps cycle with an reliable interrupt worked in now is really nice, as opposed to not having it at all when wind shock was on shock cooldowns (or very rarely).

There are almost no boss fights where a spell with less than 2 sec cast time is needed to be interrupted *.
Thus. You can create macros for all your spells like:
and if you see Evil Spell is beeing cast, just push down alt, and continue your usual cast sequence. Evil Spell will be interrupted within ~1.7 sec. DPS loss is exactly zero.

Provided, if you got lags of 300 ms, you won't be interrupting at all, unless you are the only one capable, but then again, DPS loss for interrupting casts is no problem.

*: Gundark mammoth boss if you go for the achivement.

Wind Shear obviously makes it easier to contribute interrupts and that's fine for the odd fight in 10-man, but honestly it's sub-optimal to have anyone but a melee class performing interrupts. Even if that melee character is sacrificing some resources (and thus damage) to do it, the design of a melee class is so much more conducive to solid interrupting than any ranged class that it's best to push the responsibility onto the melee. Also consider that Wind Shear is a relatively weak 2-second lockout and thus will likely push interrupting responsibilities onto the melee anyway (and in a way that's inconsistent with the timing of the other interrupts).

I have seen a lot of discussion about the interrupt effect of wind shear, but I have not seen much on the threat reduction. I personally love having a spell that is off global cool-down that does not trigger my shock cool down to lower my threat. I know from testing that by using a macro to chain Wind Shear with my Flame Shock whenever I would Flame Shock, I have seen my threat on a target cut by ~50%.

I guess this leads to the obvious question. Is the mana cost worth the threat reduction?

I personally think so, but I guess it would depend on how good your tank is.

I don't find it particularly useful to macro in threat reduction or even use it at all for that purpose. Unless I'm doing something drastically wrong, I don't find our threat to be a problem. Perhaps I'm just blessed with good tanks, or my damage is particularly bad.

Even when I've tried using it with PuGs, I don't find the threat reduction high enough to be useful.

The problem I see with that is that the only times that threat reduction might come in handy at endgame content (under normal circumstances, i.e. good tanks), there is no practical way to work in Wind Shear effectively. There are two fights in Ulduar that have threat considerations: Vezax and Hodir hardmodes. On Vezax, being in range for shear means we're far too close to the boss, and on Hodir there is no guarantee at all where we will be at any given moment in the fight. Unless Anub'arak has some wonky threat mechanics, nothing in Coliseum gets us close to threat capped at all. It just doesn't seem that macroing Wind Shear in with any spell is practical or even necessary.