Got it down to 0.05% jumps.
The spike occurs at 18.75%, and is a -1.24% "gain". Everything else tracks at the normal 0.04% gain, so the 18.75-20.3% haste range has a negative haste impact. Suffice it to say I think that the current rotation mechanics will be hard to master for most players, mostly in the way haste negatively impacts on output.

This statement and your recent charts lead me to believe you are reversing your earlier statements about haste being twice as good as any other stat? Am I reading the charts right when I see that spell power is going to be the most dominant stat for DPS?

I agree with this statement too, working 3 spells with various cool-downs around haste is going to be dumb. So to sum up what I read in your charts:

Spell critical won't affect 1/3 or more of our DPS.

Spell Haste not only won't affect a large portion of our DPS, it's effectiveness is reversed by all of the lengthy cooldowns involved in our rotation.

Spell power has been castrated as a source of scaling for the spec, and yet due to the pervasive weakness of the spec it remains the best by default.

Sound about right?

That's the nature of betas, things change.

However, there are some incorrect things in your post.
Crit won't affect approx 20% of our dps (Lava Burst and Flame Shock dot)
Haste affects our dps just as much as it used to, but there are some devalue points as you increase haste.
Spellpower is mostly fine, but the problem is the best scaling spell is also the one they want us to use last.

Based on 32.79 haste rating = 1%, that would put Shaman "haste capped" at 614 haste rating or 18.72% haste. That unfortunately does not include Wrath of Air, Bloodlust/Heroism, Drums, or Trinkets. The "cap" is released at 666 haste rating or 20.31% haste.

Assuming Wrath of Air 100% uptime

1.1872 = 1.050 * X

X = 1.1306 (4 sig figs rounded down for safety)

13.06% haste = 428.2374.

So 428 haste rating would be the cap for the lower bound.

1.2031 = 1.050 * X

X = 1.1458 haste (4 sig figs + 1 for further calculations)

14.58% haste = 478.0782 rating, so essentially one would need 479 haste rating to hit the upper bound on the "negative" area.
1.

Thus our new "negative" zone would be between 428 and 479 haste rating. Interval length of 51 rating is fairily severe considering our +damage scales poorly compared to Locks/Mages/Druids and Crit only affects part of our rotation.

Bloodlust/heroism should not be a severe issue as if you are at the 18.72% max point you would be pushed to 54.336% (1.872 * 1.3). Unless there's another "dead zone" between 54.336% and 56.403% (1.2031 * 1.3 it's not an issue (Bink, you may want to check).


I'm a bit too tired to bother considering effective uptime on drums, trinkets, etc.

We want to avoid the following ranges, either base or via buffs/procs:

Start %	End %	Rotations
18.75	20.20	LB & CL
25.00	26.45	LB
37.55	39.00	CL
43.75	44.90	LB & CL

This will change with latency as well.

Hmm, the last time I saw a lengthy damage assessment in beta, Lava Burst alone was making up more than 20% of the total damage, and the flame shock DoT was another 5-6%.

Since then Lava Burst base damage has been buffed 10%, a Lava Burst Glyph now adds 10% extra spell power, and a Flame Shock Glyph + Flame Shock talents now increase the DoT duration by 50%. I can't see how the Lava Burst + FlS DoT won't make up at least 30% or more of a rotation. I guess I'll have to check it out.

edit: My testing is showing approximately 40% of total damage comes from Fire now (~30% from LvB and ~10% from the FlS DD+DoT). It is also indicating that spell power is likely the best dps stat to stack by wide margins when using the "18 second FlS rotation" of FlS > LvB > Lightning Spells > LvB > Lightning Spells; repeat. The test indicated that critical is the worst stat to stack (I want to say largely because of 30-35% of our DPS gains no benefit, but even for Nature spells it was lackluster). Haste did not appear to be all that great, though it was preferable to critical.

Yes, those are pretty close to the numbers I was seeing before the Lava Burst and new Flame Shock glyphs were introduced. I take it you are not calculating those changes yet?

Has anyone given thought to Relic swapping? Since relics do not initiate a GCD, it would be possible to tie each one into a macro (like I did for restoration with healing wave and chain


Venture Co Lightning Rodin a macro with Lava Burst

Totem of Hex in a macro for CL and LB

But more interestingly would be the gladiator trinket: Deadly Gladiator's Totem of Survival - Item - World of Warcraft.

After each Flame Shock, you'd have 2-3 casts with the extra 70 spell power. If that's the case, wouldn't you want to use LB and Lava Burst over CL in that time frame since they will scale better?

Uhm.. they DO, since 2.4. Is that change reverted?

Edit: My mistake, I was testing out of combat. Silly me.

It causes a GCD in combat. And you cannot swap after you cast. So no, the change is not reverted and things are as they have been since 2.4

Not only is the dps going to drop in theory but also by missing more you loose out on elemental focus procs, elemental oath procs, and have a naturally depressed crit rate due to the 2 roll system. If you need a better explaination of whats going on with your numbers you will need to post how you arrived at them in more detail. I see no reason that the best damage per item cost would stop being hit however.

Edit: Also I thought you only get 3 hit from external sources needing 14% hit on gear. Are you using hit food or something? Also Erdluf showed what I meant why talking about the naturally depressed crit rate due to a 2 roll system

check my theorycrafting:

for those of you doing spreadsheets

Assuming 2000 spell power (on gear), 45% spell critical (gear and talents), 14% spell haste (gear and talents), hit capped

Compare Elemental Shaman (Flame Shock Glyph, Lava Burst Glyph) using a rotation of FlS LvB CL LB LB LB LB LvB CL LB LB LB (~18.8 seconds)

to a Balance Druid (no glyphs) with equal stats using only Starfire.

I am showing DPS within ~1% at those gearing levels / glyph assumptions.

However I am showing that Druids will scale ~9% better with Haste (their best DPS stat) than Shaman will with spell power (their best DPS stat).

Also, Druid Glyphs and adding other spells into the rotation give them additional upside.

That said, the scaling gap appears to no longer be nearly as huge as it was a few weeks ago. A "10% spell power bonus to all spells" talent would pretty well cover the gap, according to what I am seeing.

So I just stuck numbers into this and got ~114 Mp5. That seems like a lot for one talent point, plus that's only one shock every 18 seconds. Jumps up to ~342 if you're enhance and shocking every 6 seconds. Am I failing 5th grade math here or am I not remembering what 'a lot' means for Mp5? Or is it just a really good 1 point talent?

Your equations are off.

BM*FS*.45 is the mana saved via the talent. There's no need to do a subtraction; we're only interested in the mana being saved, not the remaining cost. Plugging in the 4396 base mana and 17% cost of Flame Shock we get 336.3 saved mana every 18 seconds. That's ~93 mp5. Enhancement would see ~280 mp5.

That appears to be a lot for a single talent point, but remember that Enhancement shares itemization with hunters, mp5 is not a stat they'll be getting much of. Elemental Focus is even more powerful.

Remember also that thinking of it in terms of Mp5 isn't entirely accurate. These talents yield zero benefit if you aren't casting the spells they affect, while the true Mp5 granted by Unrelenting Storm is always there.

Figured =\


Thanks for clearing that up.

For scaling, did you factor the druid receiving full benefit from +crit while FlS benefits weakly from crit and Lava Burst not at all?

I'm curious what rotation you used as you have 14 more chain lightnings than lava bursts on patchwerk and only 7 more CL's on Anub'Rekah.

They were run at equal DPS statistics from gear and buffs -- otherwise it wouldn't be an apple to apple comparison. All spells were treated properly with regard to dps statistics.

Also, to Loi, thanks for posting those screenshots. However, I am not sure how much we can read into those numbers. In most high-end level 70 raids 3400 DPS won't get you into the top 5, and in many raids won't even get you into the top 10. So being #1 at 80 with 3400 DPS really just leads me to believe that pug was really bad. Like the level 80 Mage that barely broke 2000 DPS. Mages haven't been nerfed at all yet, and they have been consistently in the 4000-4500 DPS range.

I simply tried to prioritize spells, and use them as they were off CD. On Patchwerk, open with FS, then LvB, CL and fill LB until CL or LvB is off CD. If LvB is off first, use that, if not use CL etc.. refresh FS everytime it's about to wear off. It's not as easy to manage three separate CDs, and try to not to stop casting at any point in time. There were a few times I had to use another LB when LvB was not quite ready yet, same for CL etc. I tried to keep rotation as close to perfect as humanly possible, but had a few clippings here and there. Mostly it was on the money though. on Anub, the fight is not static, as I had to move, deal with scarabs, adds, being thrown in the air etc, so rotations were not as close to perfect as I would want them to be.

Yeah, I stated in previous posts that pug was pretty bad as far as other player's performances. Unfortunately, with beta servers being nearly empty, I couldn't pick and choose with whom to raid :P I also posted before about my raids pre LvB buff and glyph change - in 25 man with some of the top dpsers (players from Crimson, Forlorn Legacy etc), I managed to get 3.2K while top mages, hunters, arms warriors etc were in 5K or more range. Considering back then I had also the benefit of SP, lock with CoE, and boomkin, my dmg in the current raid (if we pretend the last 10 man patchwerk had those classes present with corresponding buffs/debuffs), my dmg could probably go up by another 15% (about 12% from both SoE and SP, and another 3% hit from boomkin's imp. faerie fire), which would put me in 4K dps or so range, which is still way below the top dps classes.

I don't think you'll be seeing 5k from other classes. There have been quite a few changes on that. (Armor increase, hunters no longer getting windfury, etc). Most look to be under 4k now if this parse is a decent indicator Wow Web Stats

No idea of what the gear the people in that parse are using though

So with current skill/talent set the classes will hover on Patchwerk at around 4k in T7 gear? And they will be spread within 30% of that value? Is that an correct assesstment of the impact of recent changes ro armour etc?

Do we have any theorycraft with current skill/talentset for the various dps speccs/classes at something asembling T8 and T9 gear? How many asumptions do we have to make to make mockups of values at that lvl? (i.e. impact of setbonuses, something which was large on T6 gear)

My point being is that it would be nice to have some sort of guidelines to how much elemental will fall behind even if we were close to competitive at lvl 80/T7. The value of 9% less scaling than a boomkin (a number if iirc based on scaling for pure spelldamage) is mentioned as one comparable example.

I haven't seen any longer calculations comparing us to warlock,mages etc. (and is unable to do so from work)
While the argument that "pure" dpsers need to be slightly ahead of us is valid for Blizzard it is still relevant to know how their scaling is compared to our. If they are "slightly ahead" at T7 but they will be dominating us at t8 and T9 we will be in the exact same position that we were in T6 endgame when we fell further and further behind.

With that rotation you would be wasting mana on the Flameshock right to Lavaburst, you would want to flameshock to either a CL/LB or LB spam till flameshocks almost out then do lavaburst, getting the most out of your flameshock and the mana used for it.

Edit:Spelling error