Enhancement Cage Match: Two specs enter, One spec leaves
 

RE: http://elitistjerks.com/593203-post6074.html

Thank you for joining us for today's battle royale! In the home team corner our returning champion, the original spec, the People's Spec, the ever popular 0/45/16!
And in the opposing corner, up and comer, ever the downtrodden bad guy, the spec you love to hate: 16/45/0! (note that you can obviously have some variation of how the first 10 points are spent in order to get the final 5 points in Reverbration)

The overarching question: Which spec produces maximum personal DPS for the enhancement shaman?
Sub question - If the answer to question one is Elemental sub spec, does the benefit of the increased DPS out weigh the intangible raid benefit of 10 yards added to (cheaper) group totems?
Group consensus is that 10 yards to totems is irrelevant beyond 10man and T5 content. BT/Hyjal and beyond make almost no use of the talent

Rules and Regulations of the fight:
Keep it above the belt, no flaming.
Absolutely positively no bullshit anecdotal "I spec'd XYZ and did better DPS" posts. Provide evidence that utilizes proper testing techniques. In other words, a real comparison point would need to have a shaman with no elemental or rest points vs a resto/elemental sub-spec. The "no sub-spec" case needs to be our neutral control point I think. (Please note that a WWS parse is not necessarily evidence unless you had exact, replicated conditions for both sub-specs)

Serious part - We are branching this discussion out of the enhancement thread because it comes up way too frequently and there's no good way to reference discussions in the thread, and because no conclusive answer has really been determined such that I can updated the shaman post. Therefore, keep the elem vs resto talk here and out of the main thread. We'll post conclusions in the main thread, discussion of the issue goes here.

Edit -
Further Guidance
People are posting way too many variations of stats using the sims. We need these to be consistent or its impossible to analyze data. From now on everyone should discuss these things from a similar starting point. Lets take a moderate amount of AP/Crit say 1500 AP, 30% crit. For the Elem/Enhance test, assume *exactly* enough hit rating to cap yellow attacks, no more. For the resto side, assume *exactly* the 9% from talents, and no more. No haste, no -armor. Assume that with a 6 sec shock CD you can max the #of shocks in 2 minutes, and same with a 5 sec CD. Assume Battleshout, Kings, Might, food and a flask, all the normal spell debuffs. Uncheck the box for "someone steals my SS debuff."

Baseline Scenario
The following info should be used as our base assumption, 45 points in the Enhancement tree, no points spent in resto or elemental, the 9% hit comes entirely from gear in this situation. (To test resto with this setup, we'll pretend that we can just remove all the hit from gear and use only the 9% from talents)

We ran across this last night. We were attuning a new Enh shaman and our regular Enh shaman was also in the raid. We only run 1 melee group. We put the new Enh shaman in the hunter group, seeing as that would be the greatest benefit to all involved (and really there was no other place for him to go that would benefit anyone). The new Enh shaman was Elem/Enh and, since he was lacking Totemic Mastery, the buffs were very hard to keep on the entire group. Obviously this wouldn't be such an issue if we didn't have 2 Enh shaman in the raid at the time since he'd be with the melee group. The answer is clearly just to switch the 2 shaman, which we did, but further down the road when we only bring one of them (probably tonight) I can see the spec potentially being an issue.

No testing here. Just adding fuel to the debate.

Well, foul one is putting a shaman whom you KNOW has 20 ft buffs in a situation where he's likely to spend most of his time 30 ft + from the group he's buffing.

Obviously if you only bring one enhancement shaman, he should be in either the melee group or the tank group, never the hunter group.

To start with sim results, using my own gear for a basis:

16/45/0 scores in at 1558 dps, manually adding in call of flame since this doesn't appear to be in the sim we end up with 1570 dps. (16% as shocks)

0/45/0 scores 1483 dps. (11% shocks)

and

0/45/16 gets 1527 dps. (11% shocks)

Based on the sim we get a 2.8% increase going from resto/enh to elem/enh, clearly this relies on using shocks as frequently as possible. While I won't be respeccing each week and maintaining the same group buffs to show different specs I am willing to get a few parses done to atleast show that a full shock cycle with totem twisting is maintainable for the elem/enh spec.

Number of hours for those sims? How many times did you repeat?

For purposes of the sim and showing how close (DPS wise) the two builds are, try this spec for resto instead: 2/45/14

It puts a couple leftover points into concussion for shock damage rather then cheaper totems, especially as the sim is doing full shock rotations rather then twisting exclusively.

And if you're going to run twisting, see what the difference is in twisting with both enh, and twisting without resto, if sustainable. To be thorough, sustainable includes using mana pots if you go OOM, as I know with my current setup, I never have to use anything except small combat pots to sustain mana, if that. But it's perfectly reasonable to use supers when needed in a raid situation.

I'd run it myself, but I honestly don't have time to do the trials to a reasonable extent, and I personally don't twist unless we really need to push the DPS (usually watching the raid and calling things out for the fight). I do sometimes twist TA on aggro sensitive fights though. Although I'm sidetracking the discussion now. :p

Well, I guess we can avoid a discussion of whether Elemental makes sense in the end game -- 1820 unbuffed AP is insane.

Question: did you alter the "Number of shocks per minute" value at all? If not, I think you may be actually low-balling your performance.

I had that build for awhile. It's not really worth it -- 2% additional damage to shocks is 2% to 10% of your damage, so your overall damage may go up by .2%

On the other hand, if you stuck those 2 points in totem talents, you're improving your mana usage for Twisting by 10% of the cost of totems, the equivalent of adding 35 mp5 (2 totems per 9s @ 635 = 352 mp5 cost to twist).

The biggest benefit offered by elemental is the addition of 20% more shocks over time. All the talents leading up to this are merely icing on the cake. (Any elemental spec that eliminates this talent isn't likely to result in appreciative additional DPS.)

Ran the sim a couple of times too. 15k hours ran it 8 times for each spec, with the following stats:

AP: 1712
Crit%: 34.13%
Hit%: 21.37%
Haste%: 0
Boss Armor: 2190 (6200 - 2600 - 610 - 800) (Teron armor - 5 Sunders - FF - CoR)
Armor Penetration: 126
MH DPS: 100.2
MH Speed: 2.6s
MH Expertise: 6
OH DPS: 100.3
OH Speed: 2.8s
OH Expertise: 6
Spell hit%: 4%
Spell crit%: 6%

Procs:
Mongoose MH & Executioner OH
Trinkets: Tsunami Talisman & DST
Relic: Stonebreaker's Totem
Setboni: 4 piece Skyshatter
Shocks: Earth + Fire
Meta: Relentless Earthstorm

Buffs:
BoK
BoM
BS
MoTW
Haste Potion
Roasted Clefthoof
Flask of Relentless Assault
SoEarth
0 Resistances
CoE
Someone stealing SS charges
Scorch
Misery

--

I had Enh/Resto averaging on 1855 DPS, whilst Enh/Ele ended up on an average of 1927 DPS.
I had a 3.9% DPS gain. I haven't actually tested the spec, but I might just do it this reset.

Where did you get 4% spell hit?

I ran with the number of shocks I could realistically put out with each spec (pretty much every 5 sec with the elem sub, less than every 6 sec with resto sub due to gcd clipping vs cooldown on shocks).
Sim ran for 10,000 hours 5 times per spec.

I did leave the boss armour box on default value however, as I was in a slight rush and didn't have time / was too lazy to check for more realistic values. I also only used buffs I get or use 'every week', haste potions and flasks on farm content are not something I see much point in. I think that covers why Illundai's sim gives noticeably higher numbers, but the differences are still in the same direction.

The sim atleast does agree with what I said in the main thread when this first came up, that since our recent patch changes went in elem is the better option for personal dps, it is however more involved. If you twist with a resto sub, and then just shock with an elem one you will lower raid dps. But if you do a full shock/twist rotation with elem you will gain a small boost, along with taking slightly less fire/frost/nature damage, but also losing totem range (a non issue in my opinion with a melee group, the number of times you need to move far enough to need to redrop with elem, but wouldn't have needed to redrop with resto is very small) and ankh bonuses.

Alliance shaman have a 1% spell hit aura.

I'll pop in here with some quick math for shock rotations shortly, working on a little spreadsheet for comparisons. Should be up for you guys later this afternoon.
However, I'm not sure what damage co-efficient Yo! used for Earth Shock, as there is a 10% penalty for the interrupt effect.

Co-Efficient
Earth Shock: 0.386
Flame Shock: 0.670 (0.150 Direct, 0.521 DoT)

[e]calc finished.
Using the stats from above, and ignoring ES getting a 20% bonus from SS (at this stage anyway), an elemental build has ~100dps more than Restoration, using a 2/1 rotation rather than a 1/1 rotation (1/1 results in a dps loss, but 2/1 may only get a SS debuff for ES 2 times in 3: SS-ES-FS-SS-ES-ES-SS-FS-ES-SS-ES-FS)

Gentlemen: working on a long treatise on shocks and twisting (spending a lot of time building my application today).

Thought i'd throw up some prelim results:
- Complete twisting cycle with SS & shocks costs 557 mp5 with a resto spec and 648 mp5 with an elemental convection (678 without), an increase of 91 mp5.

- With a 1.5s GCD, it is impossible to synchronize a 9s WF cycle with a 9s Goa cycle, a 5s shock cycle and a 10s SS cycle. There's enough GCDs/minute, but you encounter a race condition within the first 25s and something has to slip.

- I believe this is true of a 6s shock cycle as well -- you're only doing 2 more actions per minute with elemental. Folks who tell you they maintain a shock cycle while twisting are definitely de-prioritizing one or the other of twisting and shocking.

- Got an idea for a 2 macro twisting system (one for totems, one for instants) that de-prioritizes GoA, resulting in slightly less GoA uptime in exchange for a close-to-ideal shock & SS schedule with no slippage of WF. Will be able to enumerate how much GoA is lost and how close to perfect the shock & SS schedule is. You should be able to use this information, along with the mana utilization issue, to tell whether you could maintain shocks & twists in an elemental spec and whether your DPS group gains more from prioritizing shocks vs. prioritizing totems.

I used the stats exactly as you have them listed, with 10 shocks per minute and Yo's sim is saying it's only:
Enh+Resto: 1679 DPS.
I then changed the stats to 18.37% hit, and 1% spell hit, with 12shocks per minutes which netted the following DPS. . .
Melee stuff: 1454
Shocks: 128 *1.05 (for the concuss talent) =134.4
Searing Totem: 79 *1.15 (calling those flames) = 90.85
Enh+Ele totals: 1679.25

I don't know how you got 1800s and 1900s for DPS. But I certainly got no where near that high. My results show the two specs to be essentially identical for personal DPS. Obviously I missed something else on the sim. I only used the exact things listed all other effects I disabled.