Rhaegal

Ah! Interesting. I was going off what the people said in the wowhead comments and hadn't seen the Arcane Strike spell. Only the Aldor comparison is at all relevant then.

I think the Pendant of Resolve is going to be fairly useless for us, given that the expertise cap is ~90, and we'll have 44 from the trinket + whatever else you have from other gear. That trinket plus just the three new T6 pieces is well over the cap, which would make the proc do absolutely nothing. I'd much rather have passive expertise than to rely on a proc, or even an on use effect, for it.

Stand back! I'm going to try SCIENCE!

 

Kahdrick

Saw this older post and thought I'd add in a modification that would help to make sure you have Stonebreaker's Totem equipped before casting a shock:

set up both of the shocks as described, with

/cast Flame Shock
/equip Totem of Astral Winds

and

/cast Earth Shock
/equip Totem of Astral Winds

and then set up stormstrike and watershield to equip the stonebreaker's totem:

/cast Stormstrike
/equip Stonebreaker's Totem

and

/cast Water Shield
/equip Stonebreaker's Totem

The reason for this is because water shield is not on any cooldown other than the GCD, it costs no mana, and it's something that I usually refresh all the time anyway. It also figures that the only time your casting macros wouldn't equip the stonebreaker's totem before a shock was being cast was when stormstrike was on a long cooldown, and you had just used a shock (thereby equipping the totem of astral winds). That's the perfect time to refresh water shield and equip your stonebreaker's - when the gcd is not going to interfere with your rotation.

Of course, you could just equip the stonebreaker's outside of the macro, but why not maximize the utility of a single button press?

Shamrogue FTW!

 

Othieus

So basicly you're saying that it works like Armor Penetration does, except opposite.

 

Omniro

Concerning haste I just steal the graphics from the Rawr-Thread (thx to the auther):
wf procs vs. haste rating


flurry uptime vs. haste rating


1% haste does increase dmg by roughly 1% (all that has no cd, so basicly white dmg) . White dmg is about 40% of our dmg. 1% haste = 0.4% dmg increase. But as you can see in the first image there is a point where windfury procrate is breaking in by 2.5% (WF being about 30%) so at this point 1%haste does lead to : +0.4% dmg - 0.75% dmg = -0.35%

I think this is about right, ain't it?

 

CeNo65

This is going to sound bad. I know it shouldn't work. I didn't think it would work. Yet, it does.

/castsequence reset=12/combat Flame Shock(Rank 7),Earth Shock(Rank 8)
/equip Stonebreaker's Totem

I tested swapping from no relic to Stonebreaker's and from a healing relic to Stonbreaker's. Latency might be the cause. In the first zone i tested it my latency was at 150ms. In the second zone it was 200ms. Any fios folks want to try this out? I am west coast and i believe my server is located in Boston.

 

Omniro

Just a quick sidenote: It always DID work.. for any kind of "range" slot item. Librams, Idols, Totems.
Just as starting to cast a heal and swapping to healer weapon/shield does work. But not the other way around.

 

• Binkenstein

I thought it was the blue expertise/flat ap buff item that you asked about?

That said, it spawned the whole proc rate discussion, and finding out that the rogues had a formula all along!
I'll see what I can work out for the stacking buff in an "average AP value" based on weapon speeds.

[e]there's nothing on wowhead, so I can't check the trance duration/chance either

 

Malan

Post said formula please.

Shitting up every single thread on EJ since '06

 

• Binkenstein

D = Duration of the buff
p = Chance the the buff procs on any given attack
v = Number of attacks that can trigger the proc that occur in a given second, and
c = Internal cooldown of the trinket
U = uptime, then

u=D/(c+(1/VP))

It's in the proc mechanics wiki entry (Malan can read it, the rest of you probably can't). Only difference is I went V/P rather than 1/VP, but it works out the same anyway.

DW V = AB/(A+B) where A & B = weapon speeds

 

kijik

Question in short, I have some badges saved up, but I am also looking to stack armor penetration to an extent, and thus I am pondering buying Dory's Embrace, BUT, Vanir's Right and Left Fist of Brutality are going to become available in 2.4, and I am wondering whether I should save badges for those, or go again and burn my dkp on Syphons in BT. Also, keep in mind that I do not have DST and do not forsee myself obtaining a DST in the near future, rather I will be picking up the Shard of Contempt from Heroic Magister's.

Thank you for any light you can shed on the subject.

Edit: Of course when I enter SWP with my guild I'll be picking up Mounting Vengeance ASAP.

 

Tulaugu

I've been studying the main post and the calculations now for a few days and first off all I'd like to say that it's impressive work, since it also seems to really work like mentioned above.
I only have two questions that keep bugging me a little bit.

1. Is there already any progress on "the most effective talent build"?
2. What's the EP value of a meta socket. Take for instance RED. It's not hard to calculate the EP value of 12 agility, tho the 3% crit damage seems a bit hard for me. I hope there is a solution for that.

My thanks

-Tulaugu-

p.s. Maybe it has already been asked before, if there is an awnser, please put it in the main post. I was not really motivated to dig through over 200 pages.

 

Khlysti

Before I go any further, the important thing to always remember with the '1.5 speed point' is that while haste may have smaller returns in terms of dps when crossing this point, any meaningful amount will still be a dps increase.

Now, to give a full run down on the set I used for the sim, (as standard for me, I use my own char as a base).

AP: 1932 Crit: 31.48 Hit: 17.48 Haste: 4.76 Armour Penetration: 252
MHDPS: 100.3 MHSpeed: 2.8 MHExp: 0 OHDPS: 100.3 OHSpeed: 2.8 OHExp: 0
Spell Hit: 1 Spell Crit: 5 Strength: 216 Agility: 328

I used Dual Mongoose, Madness of the Betrayer, Stonebreaker Totem, 24 shocks per minute and no waiting for SS use.
5/5 Concussion, 3/3 Call of Flame and 5/5 Reverberation. (other talent tick boxes as default)
Default for Buffs/Debuffs except: Improved Hunter's Mark, Expose Weakness, CoR, Solarian's Sapphire.
Boss Armour was set to 6600 (yeah this number is odd, didn't notice till I was done, but it doesn't affect the result).
I also had no haste pots drums etc on the sim as these I would model myself.
Also note that Heroism is an extra 30% multiplier on weapon speed and cannot just be added as +30 to haste% on the sim.

With that listed on to the simming.

This setup gave me a baseline dps of 1853 on the sim.

For now lets assume the shaman is the only leatherworker in the group and as such we can have:
A: Drums + Hero for 30 secs, Hero for 10 secs, nothing for 30 secs or
B: Hero for 40 sec and Drums for 30 sec.

In this case A gives a 2055 average and B gives a 2068 average.

This would appear to suggest that option B (use seperate) is the best idea.
This however is not always true, option B wins here because the mongoose procs on top of drums flurry and hero are just enough to push us fractionally below the 1.5 point, while using drums seperate from hero means we never hit this point. With mongoose off A wins by ~15 dps, with it on B wins by ~13 dps. (the dps with mongoose on is vastly higher than that with it off).

Modifying haste rating to the point that we just clip the 1.5 point without needing mongoose procs sees option B pull out to ~20 dps ahead of option A.

This shows two basic points,
1) haste stacked with hero is better than the two separated
2) crossing the 1.5 point can cause haste to give lower returns

In our example above the clipping is enough that combined with how small the haste gains from the drums is we are better off seperating, we do however still gain dps from drums used during heroism. (I have no numbers here, but I assume that as a group drums + hero is still better than seperate since the other group members don't have the 3 sec WF cooldown to contend with).

We will now assume that drums have 100% uptime (by which I mean that for the peroid at and near to heroism use drums are up full time, this only needs 2-3 leatherworkers, easily achievable in a melee group wanting to max dps).

With haste potions we have a similar choice to before.
A: Hero + Pot for 15 secs, Hero for 25 secs, nothing for 15 secs
B: Hero for 40 secs, Pot for 15 secs.

Staying with the same details from earlier, but with the extra 80 haste rating from drums always on we get:
A = 2210 dps and B = 2203 dps. A having a 7 dps lead.

With no mongoose we see however that B is 9 dps ahead.

Once again this a 1.5 point factor, if Hero + mongoose + haste rating means you are already sub 1.5, then using the pot with the hero is better, if hero + mongoose + haste rating still leaves you just above 1.5, and using the pot on top of hero will take you under it may be better to use them seperately.

Other clickies used during hero will help sway to using a haste pot then too.

The conclusion then.

Whether or not to stack on use haste effects with heroism depends on whether the extra effect is what pushes you through 1.5 speed, if the on use does this it may be better to seperate, if you would cross it anyway then using together is better.

As to haste rating 'getting worse the more you have, as a sort of reverse of Armour Penetration', that is just pure silly talk. If you could hit enough passive haste to reach 1.5 with just passive / flurry / mongoose, then a piece taking you just over may be bad, but with the need to use very slow weapons that amount of haste is not realistic.

Once again (since this really does need saying as often as possible), passive haste that causes hero+flurry+mong to go over the 1.5 point is still a dps increase, just a slightly smaller once, any more passive haste past that point goes back to normal value again, it is just that one piece with a reduced return, and even then its a small reduction. Claims like 'Haste AEP goes downhill AFTER your 'Heroism-Speed' is below 1.5' are plain wrong, as once you have already crossed that point the value is the same as it would be higher than it, only the act of crossing is relevant.

 

Grekk

You're right, i just tried the numbers myself. :P

I guess that's what
was about. Sorry to have wasted your time :P

Anyway even though the DPS loss from crossing the 1.5 point is small, wouldn't it still be better to 'invest AEP budget'
into other stats and prevent this happening altogether?

 

◊ Shalas

Hitting 1.5 with just flurry and passives isn't unrealistic -- it's impossible. Once 2.4 comes out, the absolute maximum passive haste possible for a shaman will be +29%, and +33% is needed. However, if you have 0 haste on your gear and no haste procs other than flurry/mongoose, 2.6 speed weapons will be hasted to almost exactly 1.5 under BL. As a result, the sim might be giving overinflated AEP values for haste. Depending on how it calculates AEP values, it could easily also be giving underinflated values for the first point of haste too, as adding 2 HR is enough to push you under 1.5 seconds with BL/Flurry/Mongoose x1.

With a DST, there's probably a small drop in haste value at 150 haste rating, but that really isn't worth worrying about.

If you have any haste at all, you're already spending part of the fight under the 1.5 level. Adding more haste merely increases that time, which means that haste has a linear (upward) effect on your dps which should be accurately predicted by Yo!'s simulator. It's really not worth second guessing the sim -- if it says an item the best choice, it probably is.

 

• Binkenstein

Right, updatery time.

Basic uptime assuming just white hits (I haven't even tried getting a WF model working yet)

A & B are the weapon speed values for calculating V from above.


But for simplifying the calculations, I will use V in place of the above.



Now, the trick with a trinket like this is that we have to calculate the "ramp up" effect of the buff.
This is done using the following, where c is the per-charge buff value



This should give us an overall buff ap value, which we simply multiply with U to get our average AP value for trinket comparison.

Might want to double check the math though.

[e]D is the duration of the "trance" plus 20 seconds after it.

 

• Disquette

I'm a bit confused by the basic "hits that can proc per second" part:

putting in 2.6 and 2.6 for the weapon speeds, I get 1.3 hits per second. It seems to me both from that result, and from doing the algebra, that it should be the multiplicative inverse. Shouldn't it be 1/1.3 hits per second = 0.77 hits per second when dual wielding non-haste/flurry 2.6 weapons?

Mainhand / second = 1/speedMH
Offhand / second = 1/speedOH

Total hits / second =

[top] MainhandHPS + OffhandHPS

1/speedMH + 1/speedOH
= (speedOH + speedMH) / speedHM*speedOH

Or, in your terms:


I apologize if I'm just being obtuse. I don't try to be dumb, but sometimes it just comes natrually.

 

• Disquette

Now I'm feeling *really* dumb, because there's another part I don't understand. Please feel free to PM me if you don't want to clutter the board. Let's take the formula in question...

So, uptime% = all that jazz.

OK, so let's put some example numbers.
D, duration = 20 seconds
c, cooldown = 60 seconds
V, hits per second that can proc - this is going to be my example variable.
P, proc% per hit = 10%.

So, simplifying your equation, we have:




So, all other things being equal, if you have a total of 1 attack per second, this is...






Or, all other things being equal, if you have a total of 2 attacks per second, this is...






Why is it that if you have more attacks per second, all other things remaining equal, that your uptime decreases? Well, I know the answer is "it doesn't", but I guess my real question is... where am I going wrong in understanding the formula?

Thanks much, and again, my apologies for cluttering the board, but I'd really like to understand this.

 

• Binkenstein

Ah yeah, sorry.

The variable description was copied from the Proc Mechanics wiki.

My biggest "problem" with algebra atm is handling all those "nested" divisors, but yes, I think you're right.

V = hits per second, so 2 hits would actually be 0.5/0.1 = 5 or 65 seconds rather than 70.

That's what I get for doing everything in my head >.<

 

Grekk

One last try to contribute something useful

Let us assume that this 30% multiplier increases the value of haste by 30%.

(all stat weights are for my gear, I use HR, DPS, AEP and HR interchangeably)

Using Yo!s simulator I found out that crossing the 1.5 point will cost me 22dps = 63 AEP = 34.8 HR.
To counter this loss, the 30% increase to haste with Heroism up needs to be greater than said 34.8 HR.

• Using drums that take you below 1.5 when hasted increases their value by:
  80 HR * 0.3 = 24 HR, 24 HR - 34.8 HR = - 10.8 HR
  => You lose 10.8 HR compared to using them while unhasted.
  
  
• Using a haste potion that takes you below 1.5 when hasted increases it's value by:
  400 HR * 0.3 = 120 HR, 120 HR - 34.8 HR = 85.2 HR
  => You gain 85.2 HR after compensationg for the loss.
  
  (strength should have it's value increased by 30% too, or am I mistaken?)
• Using an insane strength potion when hasted increases it's value by:
  120 STR * 2.2 = 264 AP / 1.81 = 145.86 HR, 145.86 * 0.3 = 43.76 HR
  => You gain 43.76 HR by using insane strength potions during heroism.

Conclusion (for personal dps):
If drums take you below 1.5 speed, they are worth less than when they are used unhasted.
However if you were to have 100% uptime, canceling the buff is not an option,
as you will lose the (80 * 1.3 - 34.8) = 69.2 HR they still give you.
Even with 0% passive haste and a 2.8 speed weapon haste potions will always take you below 1.5 during Heroism.
Therefore they should be used while hasted as the 30% gain is higher than the 'cross-loss'.

So there are basically 4 possibilites:
(highest dmg first)

You have...

1. .. so little haste that drums don't take you below 1.5 when hasted.
   => use drums while hasted.
   (unless there is some synergy with the potions if used at the same time that i am not aware of)
   
   
2. .. enough passive haste that drums will take you below 1.5 when hasted.
   => use drums while not hasted.
   
   
3. .. enough passive haste that you are always below 1.5 when hasted but using drums unhasted will not take you below 1.5.
   => use drums while not hasted
   
   
4. .. (for some reason) enough passive haste that using drums will take you below 1.5 even unhasted.
   Unless you are wearing daggers (which you shouldn't) this should never be the case
   => use drums while hasted

Whenever you are in a group of non-shamans you are probably better off using drums when hasted,
as the other classes don't have 3s-WF-CD issues.
So basically it took me three hours to find out what everyone has probably been doing anyway:
Use everything while hasted only


feel free to pick this apart as it's built on assumptions + I was tired :P

 

Rhaegal

I assume that trinkets, etc., can also proc off of specials like Windfury and Stormstrike, though, which is going to complicate that formula further. Add a term:

A = average number of other attacks per second

and leave V as is, the amount of time between each autoattack hit. The formula becomes:

u = D / (c + (1/(P/V + PA)))

Looking at some old WWS reports for the average frequency of WF hits, I get ~ 0.2 hits per second, and if I Stormstrike once every 12 seconds (accounting for waiting for WF to cool down, and latency), that's 2 hits per 12, or 0.17 his per second. V, as calculated for 2.6 speed weapons flurried to 1.82, is 0.91).

u = 20 / (45 + (1/( 0.1/0.91 + 0.1*0.37))) = 0.44

44% uptime seems kind of high, but then, I get 37% if I leave out WF and Stormstrike, so it's not a huge difference. The units work out, anyway. Do WF and SS not actually proc trinkets, and I'm an idiot for including them? Or did I include them incorrectly? (both reasonable possibilities)

[e] Just cleaned up the algebra a bit so it didn't look as messy. It's a little more clear this way, too, and makes me think I'm less crazy.

Stand back! I'm going to try SCIENCE!

 

Gehenna

The pendent of resolve will not be good, you should have your expertise capped in 2.4, and +100 expertise will basically do nothing.

In Addition, do not bother with swapping totems as it resets the swing timers, you will not gain DPS, you will lose a massive amount.

 

Camplord

Yo, a shaman buddy of mine told me to read this thread, OMG at the amount of pages, to find out whats best for offhand, slowwep with WF or fast with Flame. And im having quite a hard time finding it, read that stuff on the first page but there was just a place holder for WF vs Flame so wondering if someone knows on which page this is written if it is at all.

Right now i use Syphon for MH and Swiftsteel in offhand with Flame. In raids im a healer most of the time but once in a while when we have alot of healers signed i get to spec Enhancement aswell. And our ordinary Enhancement shammy uses what you have written in this thread, Syphon and Rising with WF in both and he has better gear then i do, not alot but better nevertheless. And i say this cause i win dps on 7 out of 10 bosses and on all trash, and still all shamans i know tell me to get a slow offhand and use WF on it, but i cant get myself to do it since none of these shamans has never outdpsed me. So i wanted you guys to explain to me=) my char name is Campsama, so if you check armory you should see my enhancement gear atm.

So from my point of view atm its like slow weps with WF in both hands has the potential to do more dmg but most of the time they dont, while fast offhand with flame gives a more stable but rather high dps. And it feels like a stable dps is better then a maybe high dps=)

 

Mox

I can only assume the shamans on your server/that you play with are horribly bad and plain suck, theres no way that WF/Flame will beat WF/WF given the same circumstances. It's not a case of "potentially more dps" it's 100% certified a hundred times over.

 

automatica

I don't even know where to begin on this because the grammar made my head hurt.

It's an understood that slow/slow with WF imbued on both weapons will always outperform a fast offhand with FT or FB. You can read section 1.1.1 for further explanation.

 

Seidule

Yeah -- I tried the totem swap thing last night and it was disaster. Woops!