The One Stop Elemental Shop - Page 21

Daidalos · 12/18/07, 5:08 PM

I saw a downgrade since I had less crit and therefore less clearcasting uptime and DPS

Haste is better than crit (in ideal situations). Did you have ideal situations? I have no idea I can't view wws logs while at work but statements like "I had less dps because I crit less" aren't neccessarily so. If you are mana constrainaed then yes that will be true if not (and spamming LB not hitting gcd) its better.

If you are arguing that you do not have the mana regen to sustain a haste build vs dmg / crit build for LB spam thats perfectly valid and is one of the draw backs to haste. For me, we always have spreist in the caster group and they are in the 1300-1400 shadow dmg range unbuffed. Its diffucult to dump my mana fast enough at times so lack of clearcasting doesn't invalidate hastes worth.

Originally Posted by Tejs

Perhaps I should re-phrase then. My 'perception' was based on WWS reports, and I didnt notice any benefit from switching from T5 shoulders to Pauldrons of the Furious Elements, and in fact I saw a downgrade since Unless you stack spell haste, spell haste on one or two items is mostly worthless for the item budget it takes. That's the argument. Your reaction time, latency, and perhaps even hardware make haste less useful until it becomes more 'noticeable'. Even if you are mashing your keyboard button for Lightning Bolt, can you say that each mash comes 0.04 Seconds after the previous one, or in enough time for the client to queue the spell / cast it?

Theorycrafting in general has a difficult issue when matching calculated timing against real life timing. And my real timing experience shows that the Haste just wasnt worth it.

I still disagree with this logic. if you arguing that for x fight I can't chain cast therefore spell haste is not valuable ok I agree. If you are saying that in 1% haste isn't as good as 1% crit because of timing I disagree. your reaction time does not affect the vlue of haste much assuming that it is a small faction of total casting time and you are mostly chain casting.

Lets take an example of the .04s reduction.
0 haste
2.05s cast lb
2.10s cast lb
2.12s cast lb
2.02s cast lb
etc etc
avg cast time 2.0725

The time inbetween is due to reaction time or latency etc etc.
now assuming its not harder to hit the button for a 1.96s lb than it is for a 2s LB we get
2.01s cast lb
2.06s cast lb
2.08s cast lb
1.98s cast lb
avg cast time 2.0325

actual reduction .04s

2.0325/2.0725 = .980699
1.96/2.00 = .98
so there is a slight reduction in the value of haste due to lost time between casts but as you can see it does very little to reduce the impact of the value of haste.

Personally I just spin my mouse wheel and I have very little latency between casts.

• Binkenstein · 12/18/07, 5:40 PM

Originally Posted by Tejs

Perhaps I should re-phrase then. My 'perception' was based on WWS reports, and I didnt notice any benefit from switching from T5 shoulders to Pauldrons of the Furious Elements, and in fact I saw a downgrade since I had less crit and therefore less clearcasting uptime and DPS.

I would never go to PoFE over T5 anyway.

I'm working on (and reworking) my haste calculations, with feedback from Daidalos & Skyhoof, but it's mostly around de-valuing haste due to the extra mana use (and, in a cleaver calculation, that de-valuing is based on the current regen/use ration that mp5 is valued on).

Kasi · 12/18/07, 6:18 PM

Why does the mana use matter though? Currently in a raid I run with a shadow priest, and I am sure the majority of ele shamans do as well. I can now run Cl/LBx3 virtually every fight without fail. I don't see how haste. no matter if you push LB down to 1.5 seconds will make a pure LB spam more mana inefficient than having every 4th spell be a Chain Lightning. With the new water shield plus just BoW I don't see how I'd ever run out of mana on any fight even without a shadow priest if I just spammed LB. Add in there the new viability of ret pallies has made uptime on JoW much much higher.

Comparison (no haste 3 LBs, 1 CL = 300x3 + 760 = 1660 mana) vs (enough haste to make LB take 1.5 seconds 5x300 = 1500) Not to mention it would be hard to get that much haste anyway so I would be a faster than 1.5 second LB.

Right now over pms with Daidalos I'm trying to work out with top of the line end game gear whether it is better to go with full dmg/crit and the 3/1 rotation or go with best in game haste items, getting about 200 haste and sacrificing roughly 100 dmg/crit and going pure LB spam. I have a feeling the first will be better, but I don't know. Any questions about threat or mana are meaningless really since without a SP CL in your rotation is untenable. All I want to know is which does more actual damage. Whether 4 piece T6 plus haste will be able to overcome the pure damage/crit numbers of a CL/LB3 rotation is imo the biggest question the raiding elemental community has today, and something that for some reason has not been answered.

• Binkenstein · 12/18/07, 7:40 PM

Part of the calculations is an option to select Spriest/BoW/Super Mana Pots/Judgement of Wis as mp5 buffs, so the extra regen will be taken into account.

Once I've got ShamStats re-done and up to scratch, ShamSpells will be next under the knife, so hopefully that will be able to highlight the differences

Tejs · 12/19/07, 12:31 AM

Originally Posted by Binkenstein

I would never go to PoFE over T5 anyway.

I'm working on (and reworking) my haste calculations, with feedback from Daidalos & Skyhoof, but it's mostly around de-valuing haste due to the extra mana use (and, in a cleaver calculation, that de-valuing is based on the current regen/use ration that mp5 is valued on).

Unfortunately for me, the ShamStats (or w/e) spreadsheet said PoFE > T5 Shoulders when I purchased them, so the version I had obviously favored spell haste a bit more =D

RK · 12/19/07, 9:06 AM

The problem with haste calculation in (previous versions of?) the spreadsheet is that it didn't take into account global cooldown, especially global cooldown on CL. It should have been pretty accurate for straight lightning bolt spam. PoFE possibly are better for straight lightning bolt spam than t5.

There's also the practical issue, which has been raised, that particularly with minor amounts of haste you may not be able to take advantage of it on a practical level, so dropping other stats for 1-5% haste would be a potential downgrade even for straight lightning bolt spam.

There's also a second practical issue- if you have 50 spell haste rating (roughly 3.33%), then you need to be casting for a full minute before you actually get an extra lightning bolt out of it at all. You'll finish your 30 lightning bolts in 58 seconds instead of 60, allowing a 31st lightning bolt to get in. And if you're working chain lightning into the rotation, as you should be, then it's even longer before the 31st lightning bolt gets in. Until then, your total damage output hasn't increased at all.

Again, at higher values of haste you'll see that "extra" bolt come around sooner.

Nonetheless, this has to compare with spell damage and crit which boost every bolt, even if you're only getting off 2-3 in a row before having to do something else. There's a lot of fights where this is the case.

Kasi · 12/19/07, 12:14 PM

I have been looking at spell haste as the idea of getting to pure LB spam instead of using CL because of the 4 piece T6 bonus plus the GCD issues of CL, but I suppose one with enough haste could do a 4/1 rotation instead of a 3/1 which would be a hybrid of the two systems. Daidalos is working on that now and hopefully we can see the numbers today. If not I can work out the math myself this weekend.

Buanna · 12/19/07, 12:42 PM

Originally Posted by Tejs

Unfortunately for me, the ShamStats (or w/e) spreadsheet said PoFE > T5 Shoulders when I purchased them, so the version I had obviously favored spell haste a bit more =D

It overrates all haste.

*I had an explanation here, but Binkenstein already figured it out in post 502*

Wodi · 12/19/07, 1:24 PM

You can calculate your DPS increase from HASTE by (Avg. LB dmg. / Cast time) * (Haste% / 100) = DPS increase AVG. You will also notice that the more your LB's are hitting for on average the more bonus you get from +haste.

Nez · 12/19/07, 9:05 PM

Haste is cheap on the budget and scales with both spelldamage and crit.

RK · 12/20/07, 9:01 AM

Originally Posted by Wodi

You can calculate your DPS increase from HASTE by (Avg. LB dmg. / Cast time) * (Haste% / 100) = DPS increase AVG.

But for the reasons in my previous post, in practice it doesn't work that way because you don't see the benefit at all until the haste "Advantage" has built up enough that you fit in one more bolt than an unhasted person. Haste is "spiky" in the dps it adds. It doesn't nicely average out for spreadsheets or formulae.

You will also notice that the more your LB's are hitting for on average the more bonus you get from +haste.

Same for spell crit or any other scaling stat.

RK · 12/20/07, 9:23 AM

Originally Posted by Nez

Haste is cheap on the budget and scales with both spelldamage and crit.

Haste doesn't reduce the global cooldown on chain lightning or the cooldown on chain lightning, and therein is why it doesn't actually scale well for shaman at all unless you have lots of it. To take advantage of haste at all you must go to a 4/1 rotation instead of 3/1, but a 3/1 rotation is more DPS than 4/1, hence haste starts off in the hole.

And it's cheaper in the budget than spell crit, which has always been overpriced, but it isn't as far as I can tell cheaper than raw spell damage, which is still our best value stat (if by a lesser margin than before the co-efficient nerf).

BTW, before doing a calculation for someone with 100 haste rating using 4/1 vs someone with 0 haste rating using 3/1, which is pretty close to a break even point for theoretical DPS: remember that that 100 haste rating came at the expense of something. If the 3/1 gets given +100 spell crit rating vs the 4/1 with their +100 spell haste rating, the 3/1 wipes the floor. And in reality, the alternatives to spell haste gear are not just stacking crit- it's generally spread around between spell damage and spell crit, which is more efficient anyway.

I still think it's that point that haste advocates really miss. Got to take into account the opportunity costs of getting haste. I could say "+300 int improves my dps!", and it would be true too, but to stack int you're losing a big opportunity costs of better stats. Similarly, stacking haste loses too much in the way of spell damage and spell crit.

HOWEVER, spell haste is clearly the best stat to stack when you are lightning bolt spamming and for various reasons aren't using chain lightning at all. I will give you that. So don't just DE haste stuff or anything

Daidalos · 12/20/07, 12:32 PM

Originally Posted by RK

Haste doesn't reduce the global cooldown on chain lightning or the cooldown on chain lightning, and therein is why it doesn't actually scale well for shaman at all unless you have lots of it. To take advantage of haste at all you must go to a 4/1 rotation instead of 3/1, but a 3/1 rotation is more DPS than 4/1, hence haste starts off in the hole.

And it's cheaper in the budget than spell crit, which has always been overpriced, but it isn't as far as I can tell cheaper than raw spell damage, which is still our best value stat (if by a lesser margin than before the co-efficient nerf).

BTW, before doing a calculation for someone with 100 haste rating using 4/1 vs someone with 0 haste rating using 3/1, which is pretty close to a break even point for theoretical DPS: remember that that 100 haste rating came at the expense of something. If the 3/1 gets given +100 spell crit rating vs the 4/1 with their +100 spell haste rating, the 3/1 wipes the floor. And in reality, the alternatives to spell haste gear are not just stacking crit- it's generally spread around between spell damage and spell crit, which is more efficient anyway.

I still think it's that point that haste advocates really miss. Got to take into account the opportunity costs of getting haste. I could say "+300 int improves my dps!", and it would be true too, but to stack int you're losing a big opportunity costs of better stats. Similarly, stacking haste loses too much in the way of spell damage and spell crit.

HOWEVER, spell haste is clearly the best stat to stack when you are lightning bolt spamming and for various reasons aren't using chain lightning at all. I will give you that. So don't just DE haste stuff or anything

I agree. I'm not sure if I am a haste advocate or not, I try to just do the theorycraft and let people decide how to play from there. I do think is a good point to bring up that the gains from haste are only if you get additional spells in.
Now in most situations a boss fight will last 5-10 mins where (if you are spamming LB) you will see gains (the 100th LB) after a little over 3 mins of spamming or less (depending on how much haste you have). Another thing to note is that there are situations like on council where constant movement is required and having enough haste to get a cast off before moving or having to interrupt can greatly increase dps.

Our frost mage stacked haste until he had a 2.2s frost bolt and he noticed a very large increase in dmg on council because he was able to squeeze in casts between moving and CS'ing where previously he and the other mages would end up just canceling the spell all together. This same principle should apply to all casters. Note it could also cut the other way in a worst case situation. Say that after a min or so of casting you are about half a cast ahead of the other casters when silence, aoe, or whatever happens and your cast stops. The non hasted casters would have just finished their cast and then begin to move while you have to cancel your cast and move. In this worst imaginable situation haste would basically lower your dps with 0 gains from its itemization. The amount of haste traded for spell dmg and crit would end up returning 0 dps.

Basically people need to think about the theorycraft and how it will apply to them and not just look at DEP or X is better than Y and run off and always use X. Theorycraft is averages over time in ideal situations. This does not make it wrong but failing to understand what is actually being theorycrafted can result in poor choices.

• Binkenstein · 12/20/07, 2:00 PM

Originally Posted by Daidalos

Basically people need to think about the theorycraft and how it will apply to them and not just look at DEP or X is better than Y and run off and always use X. Theorycraft is averages over time in ideal situations. This does not make it wrong but failing to understand what is actually being theorycrafted can result in poor choices.

I think I need to put a big "This is a guide! It is not 100% accurate" sign onto my calc sheets.
If you read Skyhoof's resto thread, you'll note that I completely ignore the Scarab, even though it gets rated highly.

Wodi · 12/20/07, 4:11 PM

Had a question. What DPS #'s are you seeing in BT and MH for elemental shamans? Nihilum and other guilds like them do not take elemental or enhancement shaman to BT or MH becuase they say we do not bring any thing to the raid. They do take resto shaman for heals though. The also said Elemental shaman and Enhancement shaman do not scale the dps like the other classes. Looking at there DPS charts they are between 1800 and 2600 DPS in there top 10 on recount, none are shaman. I am confused, becuase I thought blizzard had pushed shaman as a DPS class in the last expansion. I would like to stay Elemental for in game. I am burned out for healing pre TBC.

• Binkenstein · 12/20/07, 4:46 PM

I vary between 5-10 on DPS atm, depending on the fight, but that's partially because I'm not geared optimally yet (too much hit, still got some blue items).
It terms of DPS figures, 1-1.1k atm. I expect that to go up another 300, minimum, as I replace some of these sub-T5 items.

Daidalos · 12/20/07, 5:24 PM

Originally Posted by Wodi

Had a question. What DPS #'s are you seeing in BT and MH for elemental shamans? Nihilum and other guilds like them do not take elemental or enhancement shaman to BT or MH becuase they say we do not bring any thing to the raid. They do take resto shaman for heals though. The also said Elemental shaman and Enhancement shaman do not scale the dps like the other classes. Looking at there DPS charts they are between 1800 and 2600 DPS in there top 10 on recount, none are shaman. I am confused, becuase I thought blizzard had pushed shaman as a DPS class in the last expansion. I would like to stay Elemental for in game. I am burned out for healing pre TBC.

If Nilhlum thinks that enh don't add TONS of dps they should really try it out. Take a look at Blood Legion's dps on WWS. They usually bring 2 enh and they have alot of the top 5 WWS parses (My guild also has a few top 10 WWS and we're still gearing up somewhat) out there for every boss in BT Hyjal. No slight against Nihilum they are obviously amazing, but I think plenty of other world class guilds have shown the viability of enh shaman perhaps they have other reasons they prefer not to.

For ele shaman take a look at elitest jerks WWS. They often bring ele shaman (I haven't checked for a month or so but they used to anyways) but its sadly true ele shaman cannot really compete with other dps classes. For group buffs, the 3 hit 3 crit, 101 spell dmg, and 50 mp5 are nice, but doesn't scale as well as enh shaman 10% ap + str totem and windfury/agi twisting. Since both UR and windfury scale amazingly well.

I don't mean that a ele shaman cen never be num1 dps (I'm sure there will be people who post saying they are) but in comparable gear and skill with well thought out synergistic groups I don't see ele sham being top 5 unless its a very CL friendly fight.

In my guild rogues hit 2200dps or more
dps wars 1500-1800
hunters 1400-1700
enh shaman 1400-1700
locks 1400-1600
mages 1400-1600
spriest 1200+ (but return 300+ mp5)

Really its up to your guild. If you are a skilled player who is always ready with the appropriate mats and have done your homework on properly gearing yourself, you analyze every fight to minimize movement and maximize casting time, etc etc, and your guild rewards that with a spot I doubt you will be disappointed with your dps. However, if dps meters are your primary concern then you might not be so happy as ele or enh.

Skyhoof · 12/20/07, 6:03 PM

Originally Posted by Daidalos

Basically people need to think about the theorycraft and how it will apply to them and not just look at DEP or X is better than Y and run off and always use X. Theorycraft is averages over time in ideal situations. This does not make it wrong but failing to understand what is actually being theorycrafted can result in poor choices.

We should get this framed and hang it over the entrance to the class mechanics discussion forum.

Wodi · 12/20/07, 6:19 PM

Blink I did some comparing of our Elemental shamans.
Mine:
http://www.wowarmory.com/character-sheet.xml?r=Mal'Ganis&n=Wodi
Yours:
The World of Warcraft Armory

Excluding gems in ShamStats 12-13-07
Mine is at 2317 DEP
Yours is at 2292 DEP

My stats From Armory (not including Totems)
____Nature damage 880
____Hit 16.05%
____Cirt 32.59%
____Haste ~9.9% (ShamStats)

Yours stats From Armory (not including Totems)
____Nature damage 914
____Hit 16.21%
____Cirt 33.04%
____Haste ~2.35% (ShamStats)

I am pulling between 800 DPS to 1300 DPS depending on Boss or Trash Pull. Have Peaked at 1956 DPS on a Underbog giant in SSC.

Was wondering if there is something I can do to improve my DPS pre BT, MH.

Belteshazzar · 12/20/07, 6:50 PM

I'm curious about the broad statement that Heroism/Bloodlust (and haste in general?) does not improve GCD. I've seen explicit declarations both ways, and the hunter testing steady shot is as solid evidence as I've found one way or another.

Here's the hang up: one of the priest set bonuses shaved .1 seconds off the cast time of fheal. I didn't do any theorycrafting when I was healing MC, and it could have all been in my head, but I always assumed I could chain-cast fheal faster because of it. Is it possible that spells with a cast bar can have a dynamic GCD based on haste, while instant-cast spells (and hunter shots) would have the standard 1.5 second GCD.

So here's my plan: my priest and my wife's shaman (Hiswife; Ursin) both have the Pendant of the Violet Eye. Plain and simple, I'm going to see if I can get extra procs while it's active from 1.5 second-cast spells (fheal, r1 smite, LHW, r1 LB, etc) while BL is active versus normal casting times. If the results seem earth-shattering to me, I'll start a new post with screenies; otherwise, I'll just report it here.

• Binkenstein · 12/20/07, 7:07 PM

Originally Posted by Wodi

Blink I did some comparing of our Elemental shamans.
TL;DR

I'm not going to give advice on how you can improve your dps. If there is something that is not covered in the first posts, then point it out and I will update it.

Originally Posted by Belteshazzar

I'm curious about the broad statement that Heroism/Bloodlust (and haste in general?) does not improve GCD. I've seen explicit declarations both ways, and the hunter testing steady shot is as solid evidence as I've found one way or another.

Here's the hang up: one of the priest set bonuses shaved .1 seconds off the cast time of fheal. I didn't do any theorycrafting when I was healing MC, and it could have all been in my head, but I always assumed I could chain-cast fheal faster because of it. Is it possible that spells with a cast bar can have a dynamic GCD based on haste, while instant-cast spells (and hunter shots) would have the standard 1.5 second GCD.

So here's my plan: my priest and my wife's shaman (Hiswife; Ursin) both have the Pendant of the Violet Eye. Plain and simple, I'm going to see if I can get extra procs while it's active from 1.5 second-cast spells (fheal, r1 smite, LHW, r1 LB, etc) while BL is active versus normal casting times. If the results seem earth-shattering to me, I'll start a new post with screenies; otherwise, I'll just report it here.

Interesting plan. I'm not sure if Instants and Casts-Time spells will have different GCD mechanics, but it's something worth considering anyway.
Might pay to hit the PTR, grab some haste gear, and then test it.

Belteshazzar · 12/20/07, 7:48 PM

Ok, GCD is definitely GCD

R1 smite was casting really freaking fast, but the GCD was definitely consistently giving me 13 procs on the trinket. *Shrug* It was a thought, but now it's good to know.

Tejs · 12/20/07, 9:55 PM

I actually decided to 'do something about it' and made my own Elemental Shaman DPS Combat Simulator in ASP.NET (if someone has free hosting somewhere, I can mail you the source code for you to compile yourself).

It factors Spellpower, Spell Crit, Spell Hit, Spell Haste, Spell Rotations, TLC, LO, and Skycall Totem, depending on what you choose. I'm at home right now, but I'll post some Screenshots of it tomorrow if anyone wants to see what it looks like and/or wants me to model their stats.

You set the number of simulations to do, the spell rotation you want, and then your stats (nothing spectacular like piece by piece, more of a 'what is your total') and then it computes a sample rest run of data using that. So you can look back and see what each simulation did and then I display an average over all test runs to see what kind of average DPS, Damage Done, Spells Cast, procs, etc there were.

Nez · 12/20/07, 11:09 PM

The only difference I've noticed with haste is that it gets harder to time your next lightning bolt. (~1400dmg 41 crit 101 haste and skycall totem)
I'm using pure lightning bolt spam.

Hardbody · 12/22/07, 1:14 AM

Originally Posted by Wodi

Had a question. What DPS #'s are you seeing in BT and MH for elemental shamans? Nihilum and other guilds like them do not take elemental or enhancement shaman to BT or MH becuase they say we do not bring any thing to the raid. They do take resto shaman for heals though. The also said Elemental shaman and Enhancement shaman do not scale the dps like the other classes. Looking at there DPS charts they are between 1800 and 2600 DPS in there top 10 on recount, none are shaman. I am confused, becuase I thought blizzard had pushed shaman as a DPS class in the last expansion. I would like to stay Elemental for in game. I am burned out for healing pre TBC.

I am regularly top 5, and average anywhere from 1300-1500 on fights like teron. Our warlock GM and I duel for #1 on Illidan. Really it all just depends on your gear level and skill as compared to other classes.