Chaostheoryx

[Ward of the Violet Citadel] instead of [Voice of Reason] -- 1.12 SPower
+8Hit/+12 Stam -- 1.13 SPower
+40 Hit Food -- 1.15 SPower
+16Hit Gem -- 1.19 SPower
Old Helm Enchant -- 1.24 SPower
Surestrike Glove Enchant -- 1.4 SPower

So, if one is using that off-hand to reach hit-cap, you'd be about as well off to just gem it or use food.

 

Chaostheoryx

That sounds like a fair summary, Mmootimus. I would add a general rule of thumb, "The most expensive source of Hit is gems, followed by food, then gear."

Contrary to conventional wisdom, I don't think the Benefactor gloves are worth breaking the 4pc. I think the error comes from figuring only the DEP difference between the two gloves but ignoring the fact that the gloves are one of the best sources of +Hit we have (at .36SP). If you skip the T7 gloves you're forced to make less desirable choices for your Hit afterwards.

As an exercise for the reader, try taking this build: chardev.org - a World of Warcraft Character Planner - x

Copy the stats into SEIC:
Int: 1008
Spell: 1989
Hit: 291
Crit: 569
Haste: 512
T7 2pc: 1
T7 4pc: 1

Without trinkets that should be 4733 DPS (v3.0.8.4).
With trinkets that's 5137 DPS.

Now tweak items/gems until you find something higher, then post your build here if successful.

 

Mmootimus

Has anyone suggested that? Most people seem to be hating on T7 gloves, its true, but I don't think anyone has suggested that Benefactor's are so much better they are worth ditching the set bonus for on their own. Unless I am missing something, it seems to be more a case of people aiming for 4-pc + Benefactors or cloth+T7+Benefactor's.

 

Chaostheoryx

I just can't imagine anyone not wanting the Wanton leggings I guess.

With Wanton: chardev.org - a World of Warcraft Character Planner - x DPS w/o trinkets: 4733

With Benefactor's chardev.org - a World of Warcraft Character Planner - x DPS w/o trinkets: 4664 (This is Starcrossed's list)

With Benefactor's+4pc+Footsteps of Malygos chardev.org - a World of Warcraft Character Planner - x DPS: 4672

So the Wanton leggings w/4pc give about 69 more DPS over Starcrossed's gear, albeit at the expense of 500hp and 1.3k armor.

Swapping out only the Wanton legs/Spellshock Ring for the Malevolent/Benefactor combo results in a loss of 61 DPS.

(all SEIC values from v3.0.8.4)

 

Mmootimus

Yes, the Wanton Leggings are probably part of any absolute BiS set. They are an truly excellent piece of cloth. Unfortunately being cloth I imagine many will never get a chance to roll on them. I feel this conversation is going over a lot of old ground, or maybe its just personal bias towards upgrades I might actually get to see .

Also, is SEIC really showing Spellshock ring as 2nd best for pure DPS? I would have thought there were several marginally better.

 

Ezareth

I don't know how useful it really is to take these minor upgrades that are going to net us maybe 100 more DPS from going from BIS Mail to BIS Cloth/Mail and then gemming/enchanting it all.

Part of the strength of shaman is our ability to take a couple of hits in a raid setting without getting 1-shotted. Offhands and cloth take away from that ability.

Last night on 3 drake Sarth I had to keep Thunderstorming to get whelp aggro after they spawned to keep them off the healers while the paladin picked them up from me. I'd take quite a bit of damage, but I guarantee I'd die without a shield.

If there is noone else in your guild that needs the items then yes take them if they are an upgrade. Part of the problem here is we only have access to a few Ilvl 226 items and none of them are Elemental Oriented Mail. I guarantee after Ulduar we'll have access to a full set of Ilvl 226 mail as well as several pieces of the next level.

 

Shammyskyle

Highest DPS 10 Man Naxx WWS; Patchwerk DPS 3474

Latest 10 Man Naxx WWS; Patchwerk DPS 3382

I assumed the slightly lower DPS from the more recent Patchwerk kill was due to the reduced time it took our group to actually kill him (2.38 vs 3.08). This resulted in less procs of Forge Ember and Sundial of the Exiled which reduced my DPS slightly. Nothing to worry about. I'm averaging about 3.4k DPS on this fight.

Armory Link

3.4k DPS on 10 man Patchwerk (last by a large margin, but that will be fixed soon I hope!) with my current gear. I'm stacking crit over haste. 905 Crit Rating vs 121 Haste Rating. 1669 SD with 0 buffs (no FT weapon, ToW, food or flasks).I noticed other elemental shaman stacking haste over crit.

1. Why? It would seem stacking haste over crit would produce more spells hitting for less damage. I understand spells like LvB and FS (only first tick can crit) don't benefit from crit (due to the 100% chance for LvB to crit from FS) but since a crit = 206% damage I would think 1 point of crit > 1 point of haste. In my Patchwerk kill, 64% of my damage came from LB and CL, both of which benefit more from crit than haste (my total damage % which came from spells that benefit from crit > haste is even higher once you consider the first tick of FS and a few Thunderstorms thrown in). I understand the value of haste, but I'm not stacking it in lieu of crit. As I aquire more Heroic Naxx gear, I'll be aiming for the gear with crit and haste on it (not eiither/or).

2. Do people have some Patchwerk WWS to support haste > crit? I understand the community has theorycrafted haste being more valuable than crit, but theorycraft spreadsheets dont equal in-field practice. I can't seem to find links to 10 man Patchwerk fights in which an elemental shaman is doing more DPS than myself, and stacking haste over crit.

3. With the 3.0.8 patch won't crit be even more important? Keeping Elemental Oath up as much as possible will be a priority. Also, the spells that benefit most from crit (LB and CL) are being buffed.

Thanks for your help in advance. The introduction of Hit and Haste have made Elemental Shaman much harder to theorycraft. It's no longer as easy as 'stack crit and SP to infinity'. I would really like to see some same-fight comparisions to other shaman who have decided to stack different stats.

Right now I feel like I'm trying to squeeze the last few IQ points out of a retarded kid; hopefully in 3.0.8 we will again be a force to be reckoned with.

 

Mmootimus

@ Shammyskyle

I'm probably just in a grumpy mood here, but I am really not sure what the point is of your post. If you want to personally assume for whatever reason that crit > haste, then the best of luck to you and have fun, but bringing your own stat priorities here based on nothing more than a personal hunch doesn't help anyone much.

Just to correct two misconceptions:

This in incorrect. For pure LB spam, or any combination of LB and CL haste will give you more DPS increase per point than crit.

Crit becomes a lot better in absolute terms and moderately-marginally better in relative terms. Regarding EO and CC the issue is diminishing returns. As has already been discussed on these forums, the 2 charge system for clearcasting means that any decently geared Ele will already be at 80%+ uptime for CC. An extra 1% crit will give noticeably less than 1% additional CC uptime. Whilst crit does therefore give us one of our best buffs, stacking it is inefficient in terms of DPS increase.

 

Shammyskyle

@ Mmoot:

I have to disagree with you. I don't think the theorycrafting math that has been done is correct. I believe stacking crit > stacking haste. I am using my 10 man Naxx Patchwerk kill as an example. But this simply my opinion, and I am not married to it. Not trying to start a flame war or argument, just a simple discussion What I really came looking for is comparision WWS. I want to find a WWS for the same kill and group size where another elemental shaman stacking haste over crit is clearly out DPSing me. I'm not finding much.

Can someone produce a WWS that has greater DPS than 3.4k with stacking haste over crit? I have read the threads and reviewed the theorycrafting spreadsheets. However, I don't agree with them, and I believe they use flawed math and assumptions.

What I am looking for is the WWS for a 10 or 25 man Patchwerk kill with an elemental shaman to compare my WWS to. I would like to see a player who has stacked haste > crit and see how their DPS is.

I understand the concepts being delievered by the theorycrafting, but I don't think they are optimal for DPS. I still think stacking crit is better than stacking haste, and will be even better then it is currently in 3.0.8.

 

LITTLER0TT

In a raid you will have stupidly high crit. I have 29% crit on my character sheet, which is 52% in a raid.

So the difference between you and me is, You have 120~ haste and 55% crit, and I have close to 500~ haste and 52% crit.

though I'll give you that you have less gear then me

Oh and as far as "stacking" goes

Spell power is the best talent by far

 

Mmootimus

Right, so to make a fair comparison, you would need a group with the same raid buffs, same gear as you and the same treatment from the RNG in terms of crit %, LO procss, etc. Good luck with that. I can link you a WWS of me stacking haste and doing 3.8k+ on Patchwerk 25 if you want, but it proves nothing.

What concepts are wrong? Where? What are these flaws? You have SEIC and simcraft publicly available, and with the option to dig in to the detail and tell us where they are wrong. Please explain why and how the current TC is wrong, and/or provide your own. Otherwise all you are doing is the equivalent of standing in the street yelling "the world is flat" and expecting people to believe you.

 

Shammyskyle

That doesn't make sense to me, can you please elaborate further or link me to a page explaining this? A crit does 206% of our normal damage, an increase in haste lets me cast the same spell I normally do a little bit faster. If I sacrifice haste for crit, in the long run I have more spells cast in a fight, but each spell hits for less.

Try not to explain to me with absolute numbers; I'm more interested in the concept. Is the idea that I can stack haste and increase my cast speed by say 25% (thus casting 25% more spells) while only sacrificing 5% of my crit % (haste being 'cheaper' to aquire per point of ilvl thasn crit)? Also, does haste decrease the cooldown of LvB, so with massive haste I can reduce the cooldown of LvB from say 8s to 6s?

Sorry to ask you to hold my hand, but I'm just not understanding the underlying concept of haste > crit.

 

Ghando

I don't remember who first said this (it's somebody's signature on these forums), but anecdote is not the singular of data. Saying "my WWS parse is higher than this guy's and I stack crit so that's the way to go" is silly when every simulation for Elemental damage ever created has haste outperforming crit. Remember that before WOTLK, every Ele Shaman stacked haste almost exclusively over damage or crit. It just scaled better than anything else for LB spam and it has no diminishing returns to speak of. The only thing that has changed since then is the implementation of Lava Burst and its associated crit bonuses, and that catapults Spellpower to the top of the pile. As far as ratings go, if you're hit capped haste is better. The jury is in.

I write a humor blog: http://tpalumbi.blogspot.com

 

Kishkegelt

Wanton with T7 gloves and a high end ring is slightly better than Wanton with Benefactor's and Malevolence.

However if you are committed to using the T7 gloves, then I think you would find the gemmed Leggings of Voracious Shadows from Four Horsemen would be slightly better still, and you would have less competition from clothies in getting it.

 

MatsT

Looking closely at the difference on your two WWS links gives us:

3474 dps fight: 212.5 sec unhasted casts in 181 seconds = 17.4% average haste
3382 dps fight: 180 sec unhasted casts in 155 seconds = 16.1% average haste

Ironically, this is the big difference between the damage in these two cases. You had more haste (or less lag) on the one with high dps.

That being said, two very short fights say nothing about your dps, as the dps can vary so much depending on lag, randomness with crits, length of fight (shorter fight typically means higher dps).

 

• Binkenstein

You should probably grab 3.0.8.5
I think you're looking for my sig quote.

There are a couple of reasons why Haste > Crit.
Haste affects everything aside from the FS DoT, while Crit will only affect CL, LB, and the DD portion of FS
Haste also has a lower rating requirement for 1% (~32.7 vs ~45.9 for crit)
Lastly, we already get a high amount of crit. I can be sitting around the 45% mark, so 1% more is a 146/145=0.689% increase, while going from 14% to 15% haste is an increase of 115/114=0.877%

"That young girl is one of the least benightedly unintelligent organic life forms it has been my profound lack of pleasure not to be able to avoid meeting"
Marvin, The Paranoid Android, from Life, the Universe, and Everything by Douglas Adams

<Birdemani> I feel dirty. I wanted to run a pug voa tonight and used a gear score in /1 to get an invite. But I have no idea what my score is so I just made up some number above 6000 and got 10 invites.

 

Shammyskyle

Thanks, Blink, this helps me quantify why haste > crit. Also, a few people PMed me and helped me understand the concept using real examples (thanks Kishkegelt!).

 

Ezareth

Stacking Crit is not better than Stacking Haste, but before I get into the *why* of it, where do we have the option to "stack" either stat? Our first stat is hit, once we max hit, we go for spellpower. I can't think of a single instance where someone should use either haste or crit in any gem, and there are no real haste enchants that are worth it outside of cloak, no crit enchants that are worth it outside of the boots.

We only have a few set pieces of gear. The only piece of gear that I can think where you'd make the choice between haste and crit is [Hammer of the Astral Plane] versus [Wraith Strike]

[Sundial of the Exiled] versus [Embrace of the Spider] is another possibility to chose haste or crit but neither are BIS and in most rotations with my numbers they are in a dead heat unless you are unlucky with procs or end a fight with a long cooldown up in which case Embrace wins from its latent spellpower.

The rest of our choices are pretty much made for us already.

Now for understanding haste and crit scaling modeling remember these three points.

1) Haste does not increase Elemental Oath uptime, crit does.

2) Crit due to the abundancy of it in raid buffs, gear, and talents is subject to a much higher diminishing returns penalty than Haste due to the relatively small amount of haste. Crit does not affect Flame shock dots or Lava Burst.

3) Haste has points where it receives zero DPS increase due to our Lava Burst Rotation. Haste does not affect flame shock dots or Lava Burst meaningfully.

So optimally we'd stack hit until we're hitcapped, stack spellpower in all other circumstances, when faced with a choice between haste and crit you'd chose haste in all circumstances except when it would put you into one of those "dead" zones of zero DPS increase.

In other words all this back and forth of haste over crit is just wasted breath, our path to progression has been laid out for us.

Haste does not affect Lava burst Significantly and affects chain lightning less than crit does due to the fact it doesn't change cooldowns, only casting speed.

Good point on the rating requirements, I hadn't updated myself on those numbers yet.

 

• Binkenstein

There is no L in my name.... B I N K is now it is spelled.
Re: 3. Using a priority rotation there are no black spots (using only FS, LvB and LB), although the work there was done using the r3 rotation models from SEIC. I should probably generate another spreadsheet to look at various CL settings as well.

Also, haste does indeed affect Lava Burst as much as any other spell. You're confusing damage done over total time with damage done over casting time. (I'm fairly sure we've had this argument before somewhere) It important not to confuse count scaling with cast scaling. I think I've shown previously that the only reason haste scales slightly lower than the flat percentage is because the FS dot is not affected.

"That young girl is one of the least benightedly unintelligent organic life forms it has been my profound lack of pleasure not to be able to avoid meeting"
Marvin, The Paranoid Android, from Life, the Universe, and Everything by Douglas Adams

<Birdemani> I feel dirty. I wanted to run a pug voa tonight and used a gear score in /1 to get an invite. But I have no idea what my score is so I just made up some number above 6000 and got 10 invites.

 

MatsT

I'm not really getting this part. Just using hypthetical numbers here, consider a situation where you can cast 4 Lightning Bolts in exactly 8 seconds, making them fit perfectly between two lava bursts. You then gain 0.1% haste, allowing you to cast those 4 lightning bolts in 7.992 seconds. Surely it must be worth waiting for Lava Burst 0.008 seconds instead of casting another Lightning Bolt. This means the only benefit to the rotation was the faster LvB cast, meaning you take the rotation from 9.5 seconds to 9.4985 seconds, making it a dps increase of only 0.015%. It's not a total black spot since it adds dps, but it's pretty close. In reality, due to the chance of mixing in chain lightnings in the rotation, there will be very many very small such spots. Varying latency probably removes the small effect remaining, but in a simulation you should definitely see spots where small amounts of haste helps very little.

Lets consider another hypothetical scenario. Let's say LvB did an average of 1 million damage while Lightning Bolt only did 1 damage. Virtually all of your damage comes from LvB, but you can only cast it every (8+casting time) seconds. Yet again, an increase in haste with 0.1% would only increase your overall dps with about 0.015%. This would cause shamans to scale terribly with haste. In reality, the damage from LvB is obviously not 1 million that of LB, but you will still see an effect in that direction.

 

Ezareth

Haste black spots are always apparent in any rotation especially if you count in latency. There are always points where casting another spell instead of waiting for lava burst will cost you DPS until you can cast that spell even faster.

For all intents and purposes Lava Burst is a 9.5 second spell except haste only affects 1.5 seconds of it. It also has a haste cap at 50% haste which is easy to reach during bloodlust. This is why I said haste affects Lavaburst marginally, because it is applied to a "9.5 second spell".

Chain Lightning is the same and is a 5 second spell.

Lightning Bolt however scales perfectly(although equally with diminishing returns) because it is applied to the entire spell, has no attainable "cap", and it scales precisely along the curve with diminishing returns.

Because Lava Burst takes up a proportionally large amount of our damage compared to it's casting time, this means that haste scales with Lava Burst only slightly better than it does with crit(which is not at all).

I think the poster above me addressed this perfectly as well.

 

• Binkenstein

You're looking at the addition of haste in terms of the cooldown and overall rotation.

If Lava Burst had no cooldown, it would scale exactly the same way it does now. If it was lower damage, provided a buff, and had no cooldown, it would scale exactly the same way as it does now. By looking at an overall rotation and saying "well, we only get 1 lvb off every X seconds, so because the cooldown can't be affected by haste we don't get to use it much more than we would normally, so it doesn't scale well with haste" we distort the view of the spell.

Lets look at it another way. Scorch is a spell that is only cast every so often. It scales normally as you'd expect for a 1.5 second spell. Saying that it doesn't scale well because you only use it once every 25-30 seconds is incorrect. Scaling has nothing to do with spell frequency, so while you are looking at a single spell compared with the overall rotation time, you need to either be looking at the spell compared with its cast time, or the overall rotation damage compared with its time.

"That young girl is one of the least benightedly unintelligent organic life forms it has been my profound lack of pleasure not to be able to avoid meeting"
Marvin, The Paranoid Android, from Life, the Universe, and Everything by Douglas Adams

<Birdemani> I feel dirty. I wanted to run a pug voa tonight and used a gear score in /1 to get an invite. But I have no idea what my score is so I just made up some number above 6000 and got 10 invites.

 

MatsT

Well ok, i can settle for that instead of saying that LvB scales badly with haste because of it's cooldown, Elemental Shamans scales badly with haste because of LvB's cooldown. Semantics... the point still stands, haste increases our dps less now than if we would be spamming only LB or LvB. basically.

Another way to look at it: 0 haste gives you 1.5 sec of good dps followed by 8 seconds of lower dps, letting you do good dps 1.5/9.5 of the time. 50% haste means you do good dps 1 second followed by 8 seconds of lower dps, meaning you have good dps 1/9 of the time. Let's say LvB is 6k dps and LB 3k dps unhasted. With 0 haste we would do 1.5/9.5*6k+8/9.5*3k = 3474 dps. With 50% haste we would do 1/9*6k*1,5+8/9*3k*1,5 = 5000 dps. This is a 43.9% increase in dps. If we had just been spamming one spell we would have gotten 50% dps increase.

 

Ezareth

I don't agree. It is semantics you are arguing about.

Let's ask ourselves:

What is Haste? Haste is the stat that allows you cast twice as many spells in the same amount of time.

Now what is Crit? Crit is the stat that allows your spells to do twice as much damage in the same amount of time.

Haste scales with only 15.8% of lava burst which considering it's own diminishing returns as well as the haste cap is marginal. It is because for every other spell with haste(save chain lightning for the same reason) we expect 100% haste to double our relative damage. This is not true with Lava Burst any more than we can expect crit to give it any more damage when we are already at 100% crit. They don't scale for very similar reasons.

Sure I realize cast time is reduced but because cooldowns are not, haste barely scales with it.

Crit however now has two additive scales. One is the normal crit scales for the DD of FS, as well our lightning spells.
The second scale added to this is the uptime increase on Elemental Focus.

I think you pointed out the correct reason for haste being better than crit earlier when you showed the point ratings to equal 1% of each. If they were equal Crit would likely be the better stat.