TLDR version: Windfury has a 36% chance to proc on an individual swing while Dual Wielding. The cooldown counteracts this, and nerfs Windfury hard to boot. Please fix it.

There has been much speculation on the 3 second Windfury cooldown (don't bother looking for it in the patch notes, they didn't put it in there). There is a lot of confusion (do to the almost complete lack of communication by Blizzard). I hope to clear up a lot of that confusion. This post is current a of 2.1 Live.

A quick history of Windfury and cooldowns. There was a short cooldown added (less than one second) as the first fix for Windfury self-procs. When they recoded Windfuy as yellow damage, they took the cooldown off because there was no need. The only way people knew there was a cooldown (aside from the lack of self-procs), and its later removal, was the ability to get a Windfury off of a Storm Strike and an Auto Attack at the same time. A 3 second cooldown was implemented in patch 1.11 (we don't know why) and removed in patch 1.13 or 1.14. The cooldown we currently have was in a more recent patch, we guess it at 2.0 (though because it's an undocumented change, we can't be certain, and Blizzard doesn't bother to update old patch notes with accidental omissions).

One of the workarounds (and because Blizzard wouldn't talk to us, we didn't know the cooldown was intended) for the cooldown was using staggered ranks while Dual Wielding. This meant that each weapon had it's own Windfury cooldown, instead of sharing the same cooldown between weapons However, there were scattered reports of 28% Windfury proc rates with this setup.

Numerous tests off of the 2.1 PTR (where the staggered ranks workaround has been fixed) show a proc rate between 17% and 19% (slow weapons). This is higher than suspected, based on what we know of the cooldown.

However, there have been calculations made. Assuming a 36% Windfury proc rate (I'll get to why 36% later), and a 3 second cooldown, you get around a 27% Windfury proc rate using staggered ranks. With a shared cooldown, the proc rate drops to around 18%. This seems to show that while Dual Wielding with Windfury on both weapons, you get a 36% proc rate.

Why 36% though? That's because each weapon checks both buffs every time it swings (0.2 + (1-0.2)*0.2 = 0.36)! The bug where a single swing would yield 4 Windfury hits was when both Windfury buffs proc'd (4% chance, 0.2*0.2 = 0.04). The 3 second cooldown was to counter-act the high proc rate, and prevent double-procs.

Instead of recoding Windfury so only the enchanted weapon is affected, they slap a 3 second cooldown on. It's an emergency solution. It also beaks Windfury. Windfury scales poorly with Haste effects, is worse with faster weapons, and has a 17% proc rate while wielding a single weapon enchanted with Windfury (or Dual Wielding with only one Windfury enchant). Heck, with fast enough weapons even assuming the 36% proc rate you get a severely reduced proc count (below 15% for each weapon!).

Data

Screenshot of two double-procs (2.0 Live):
http://geocities.com/californication696969/Test_WF5_WF1.jpeg

Because geocities got Pwned (2.0 Live):
http://img407.imageshack.us/img407/251/testwf5wf1ca0.jpg

I used the Mag'har Fury Brand MH and Hand Axe OH. MH had WF5, OH had WF1. You'll notice on both quad-procs that it jumps from around 90 to around 120 damage, and that they hit twice in a row for each. This confirms that the double-procs are a result of a single weapon proccing both buffs at one time. Which supports the rest of the theory.

Links (2.1 PTR):

http://forums.worldofwarcraft.com/thread.html?topicId=95979842&sid=1

http://forums.worldofwarcraft.com/thread.html?topicId=95983611&sid=1

http://forums.worldofwarcraft.com/thread.html?topicId=96122465&sid=1

http://forums.worldofwarcraft.com/thread.html?topicId=96051650&sid=1

Here is my simulator for non-staggered rates:
http://geocities.com/californication696969/WFSim2.java

And here is my simulator for staggered ranks:
http://geocities.com/californication696969/WFSim4.java

I have done some PTR (2.1) testing, attacking a Servant of Allistarjj for ten minutes while watching my Combat Log for a specific even (which did not occur). The 3 second cooldown still remains. However, when a weapon pocs Windfury, it uses the AP bonus of the Windfury enchant on the weapon. Windfury has been recoded slightly so that it always uses the weapon's own AP, and linked the cooldowns between all ranks.

Addendum: Regarding Tseric's "scary DPS" comment: It is my belief that his comment was made based on the glitched, pre-cooldown Dual Wield Windfury. It should be noted that under those conditions, it would have the equivalent of a 40% proc rate, which would result in ridiculous DPS. So while it isn't fair to mock him for the comment, it is Tseric's fault for not saying "My bad, the statement was erroneous because it was based on a bug."

WF does not scale poorly with haste. The more you swing, the more chances there are to proc WF. People seem to suscribe to this " you must perfect your haste to 1.5 seconds for DW or 3 seconds for 2 hand", but you forger that it is a chance to proc.

That being said, yes, haste will lower your proc percentage - but your overall dps will still go up.

IMHO, either:
* Give the MH and OH separate cooldowns for WF, or
* Reevaluate the other weapon enchants to make them more desirable by someone who actually swings their weapon.

If we had some sort of hybrid talents - relating strength to increased spell damage or some variant - then FT & FB would increase in viability as they'd scale.

Just my 2 cents though.

---

Q u o t e:
WF does not scale poorly with haste.

That being said, yes, haste will lower your proc percentage - but your overall dps will still go up.

---

How could you say WF doesn't scale poorly with haste then admit that it does no more then a couple sentences later?

He didn't say haste reduces your dps, he said WF scales poorly, which is the outright truth. Yes your dps goes up, but 30% haste does not always net a 30% increase in DPS, like it should.

Silence is for the dead!

---

Q u o t e:


How could you say WF doesn't scale poorly with haste then admit that it does no more then a couple sentences later?

He didn't say haste reduces your dps, he said WF scales poorly, which is the outright truth. Yes your dps goes up, but 30% haste does not always net a 30% increase in DPS, like it should.

---

Perhaps Blizzard did not intend Windfury to benefit from haste increase. Haha, what I am kidding... Blizzard used a sh*tty bandaid.

There are a myriad of problems Enhancement shaman face, from Haste not being as beneficial as it should, to completely whacked itemization (a 60dps slow green is superior or equivalent to a 80+ dps fast weapon? What?), and it all has it's roots in the 3second cooldown on WF. It wasn't a smart solution at the time, it completely borks up Dual Wield,and i'ts not needed anymore. Unfortunately, Blizzard does not seem to be willing to budge on this.

The ones I feel bad for are the "non-advanced" shaman, those that don't learn all the details from message forums, etc. When looking at two weapons, a 60ish dps slow green and a 80ish dps 1.5 speed dagger, the intuitive (even common sense) answer would be that the 80dps dagger is a massive upgrade. It's significantly higher damage, it's purple instead of green, etc. Yet it's not the case. We on the forums know the problems, with hidden cooldowns and offhand proc "stealing", but the poor schlubs in game don't.

The three second cooldown is invisible, it's not documented, and it enormously complicates almost every aspect of the Enhancement shaman and interferes with their other abilities. Shaman have the best Flurry in the game, we have Bloodlust, Unyielding Rage benefits from fast attacks, yet our primary source of damage in the exact same build runs completely opposite of that.

Remove shared cooldowns on WF, and problem is SOLVED. If Enhancement shaman are doing way too much damage at that point, find other ways to balance it.

Ugh. Sorry, /rant off.

The 3 second WF cooldown isn't the real reason you want slow/slow for a DW shaman, it's because of the WF buff itself. The nature of fast weapons like daggers is they tend to have lower damage per hit, but a fast rate, so over time, 2 weapons with equal DPS, regardless of speed, will yield the same amount of white damage. The difference though is that WF gives you 2 extra attacks with more AP, so you get the most out of weapons with higher damage per hit, which tend to be slower weapons by nature. The same goes for Stormstrike, you get more out of it with 2 slow weapons with higher damage, than 2 faster weaopns with less damager per hit but might be higher DPS. The 3 second cooldown doesn't change any of that.

It's the exact same thing with rogues and Sinister Strike, only it's only Main Hand then. You want a MH weapon with the highest possible minimum damage regardless of DPS, because you get the most out of Sinister Strike, because it's based off of MH weapon damage. So in that case you're usually better off with a slow MH sword or mace as opposed to a fast dagger, unless you're going Mutilate or some other dagger type spec.

Here's what I don't understand. Even after all of the DW WF nerfs, everyone still claims that going DW WF is still the best setup for a shaman. Much, much better than 2H WF or using any other weapon enchant. Doesn't that suggest that it was probably too powerful before the nerf?

I suspect they will continue to play with WF and the DW specification until the DPS from this setup is more in line with the DPS from other setups. I could be wrong though.

it's not so much the nerf of my dps i hate, its the pigenholing me into using slow weapons.

I begin to wonder why I ever trained daggers.... none of them are good dps.

Just... shut... up...

How many times are people going to post the "history" of WF filled with miles of text about their numbers. All this does is cause others to show up with information that is just as bad.

ALL anyone has to say is that a cooldown on WF is a horrible "fix", it never should have been put in place and the person who decided to give Duel Wield to shaman should have thought of that. It's poor game design.

The corrections:

---

Q u o t e:
Windfury scales poorly with Haste effects

---

Poorly is not quantifiable. Depending on the amount of haste you could see more WF procs, if it took your weapon speeds closer to 1.5 or 3.0 if using a 2h. It's entirely dependent on the original speed. These types of comments also generally lead to the belief that haste LOWERS your overall DPS which is simply untrue.

---

Q u o t e:
.
The 3 second WF cooldown isn't the real reason you want slow/slow for a DW shaman...
It's the exact same thing with rogues and Sinister Strike

---

No, it's not. Sinister Strike is limited by Energy Generation. Slow or Fast MH will not give them more or less SSs. With a fast WF OH for a shaman you will see more WF's. It's a bit more complicated in that it's not equally smaller hits but more often. Since WF provides 2 extra attacks there is always going to be an advantage for higher damage weapons. However, when comparing that to "stolen" WF procs... you'll find that's the bigger source of lost damage. The amount of "stolen" WF procs goes up drastically with a fast OH.

---

Q u o t e:
Doesn't that suggest that it was probably too powerful before the nerf?

---

No, it suggests that the other weapon enchants are horrible.

---

Q u o t e:

No, it suggests that the other weapon enchants are horrible.

---

And it suggest that 2h weapons are horrible, even though they were good enough for over 2 years before DW showed up? Average shaman dps took a huge jump when DW was introduced. Even after all of the nerfs that have hit DW shaman, the dps of a shaman with average gear is still much greater than it was before 2.0.

I agree with you that DW was not well thought out before it was introduced into the game and that has caused a lot of headaches that could have been avoided.

---

Q u o t e:
Here's what I don't understand. Even after all of the DW WF nerfs, everyone still claims that going DW WF is still the best setup for a shaman. Much, much better than 2H WF or using any other weapon enchant. Doesn't that suggest that it was probably too powerful before the nerf?

I suspect they will continue to play with WF and the DW specification until the DPS from this setup is more in line with the DPS from other setups. I could be wrong though.

---

I hope if they do pay some attention to it they don't use say a T6 premade lol ;) Use a few different levels between 40 and 70 with mixed blues and greens.

Regardless which side of this fence you are on, all weapon buffs need to be revisited so that they are equally and uniquely beneficial to Shaman. Also, each weapon should be on their own CD.

---

Q u o t e:


I hope if they do pay some attention to it they don't use say a T6 premade lol ;) Use a few different levels between 40 and 70 with mixed blues and greens.

---

Agreed, but sadly blizz always seems to do just what your suggesting. I guess its because they want to see what the full potential of the spec is. Sadly i dont think but a handful of enh shaman will ever see T6.

As to the subject at hand, i dont think there is a single dps class that has the same issue that we do as they dont rely on Procs from their melee swings. The ONLY thing Haste does for a Shaman in terms of benefits to WF is Attempt to make it proc every 3 seconds. This isnt always the case either.

---

Q u o t e:

Poorly is not quantifiable. Depending on the amount of haste you could see more WF procs, if it took your weapon speeds closer to 1.5 or 3.0 if using a 2h. It's entirely dependent on the original speed. These types of comments also generally lead to the belief that haste LOWERS your overall DPS which is simply untrue.

---

Personally, I think "poorly" IS quantifiable. Long story short, if they say WF has a 20% chance to proc on weapon hit, then it should have just that. There shouldn't be hidden cooldowns, or shared cooldowns, there should simply be the 20% chance. WF DOES scale poorly because, if WF is a large source of damage, haste doesn't necessarily affect it THAT much. Yes, overall DPS is increased, so it isn't like haste has a negative effect, but compare it to other classes: mages cast faster, elem shamans cast faster, rogues attack faster, channels go faster, and windfury might proc a little closer to its cooldown. Relative to the benefits other classes receive, yes, haste scales poorly.

My biggest beef is the fact that Blizzard is doing all of this behind-the-scenes fixing without informing anyone, and changing WF from what the tooltip is into some complex and ridiculous method of balancing enhancement DPS. I completely understand that enhancement DPS shouldn't be at the top. The buffs I provide to my melee group in raids is extremely nice, so I don't plan on being #1 in DPS. Other classes should be above me, and that's fine. What I'm concerned about is the skill working according to what we're told.

Personally I'd like to see some sort of change made to the PvP aspect of the enhancement shaman, because as-is, either: 1. I have to get pretty damn lucky or 2. The opponent has to suck for me to really do anything. I see no reason for stormstrike and WF hits to be able to be dodged/parried/blocked. There is absolutely NOTHING more frustrating than having 0 mitigation, finally procing a WF on a rogue, and seeing both parried, along with your other swing. Then you swing again and he dodges, then you stormstrike again and he dodges. The sheer number of classes that can just play with me unless I get lucky is astronomical, and to be quite honest, it's dumb. I'm glad I can grind things with blinding efficiency, then be worthless when I don't have teammates around against people of equal gear. I think the tree should be somehow reworked to give the spec a little more PvP viability other than "Alright, spam purge on this guy to purge the pally bubble when it comes." And no, extremely high spikes in damage aren't viability.

/end rant

---

Q u o t e:
And it suggest that 2h weapons are horrible

---

Yes it does. For a 2h to scale equally with 1h weapons they would need to either be much higher damage (which would cause havoc in PvP) or they would have to be much faster with more DPS.

Which goes back to my wish of nuking PvP right out of the game. It limits PvE in so many ways and has been the cause of way to many PvE nerfs. I love PvP but wish it was in it's own game or integrated MUCH better.

---

Q u o t e:
Agreed, but sadly blizz always seems to do just what your suggesting. I guess its because they want to see what the full potential of the spec is. Sadly i dont think but a handful of enh shaman will ever see T6.

---

Classes always have to be balanced with end game in mind. It was the reason for the Clearcasting nerf. End game geared Ele shaman could turn off their mana bar, they didn't need it. You can't be hypocritical when talking about proper balance though. Classes should be somewhat balanced at all levels of play and gear. If the only time an Enh shaman is powerful is after they get a full set of T10 gear than there is something wrong. However, I think the bigger problem is simply poor play skill. The game isn't difficult, yet daily we see people who think Shaman or some other class/spec is broken because they can't make it work. You see people running around in green gear "...of the Idiot". It's not hard to be appropriately geared for your level.

This is a HUGE tangent... however it goes right back to WF weapon. There are still countless posts of people wondering if Flametongue is good. Or if DW WF is good. People don't understand the class they're playing, how do you expect them to do well?

---

Q u o t e:
One of the workarounds (and because Blizzard wouldn't talk to us, we didn't know the cooldown was intended) for the cooldown was using staggered ranks while Dual Wielding. This meant that each weapon had it's own Windfury cooldown, instead of sharing the same cooldown between weapons However, there were scattered reports of 28% Windfury proc rates with this setup.

---

WF should only proc 20% of the time for each weapon. If you're getting more than 20% procs, then that's a bug.

---

Q u o t e:
Numerous tests off of the 2.1 PTR (where the staggered ranks workaround has been fixed) show a proc rate between 17% and 19% (slow weapons). This is higher than suspected, based on what we know of the cooldown.

However, there have been calculations made. Assuming a 36% Windfury proc rate (I'll get to why 36% later), and a 3 second cooldown, you get around a 27% Windfury proc rate using staggered ranks. With a shared cooldown, the proc rate drops to around 18%. This seems to show that while Dual Wielding with Windfury on both weapons, you get a 36% proc rate.

Why 36% though? That's because each weapon checks both buffs every time it swings (0.2 + (1-0.2)*0.2 = 0.36)! The bug where a single swing would yield 4 Windfury hits was when both Windfury buffs proc'd (4% chance, 0.2*0.2 = 0.04). The 3 second cooldown was to counter-act the high proc rate, and prevent double-procs.

---

I claim BS. If each weapon is proccing 18% of the time, that does not mean you're getting 36% WF procs. Let's say main hand swung 1000 times with 18% proc rate, that's 180 WF procs. Let's say off hand swung 1500 times with 18% proc rate, that's 270 WF procs. Together you have 2500 swings, of which 450 were WF procs. That's still 18% proc rate.

---

Q u o t e:
http://forums.worldofwarcraft.com/thread.html?topicId=95979842&sid=1

---

This is great. The guy uses WF + RB then shows that WF only procced 8.6% of the time. Well duh! Some %age of the swings were from the hand that had RB!

Until people (i.e. enhancement shamans) learn basic math, this topic is dead.

Windfury does not scale with haste/flurry. That is the point of the stealth cooldown. People with high crit ratings were doing insane WF damage with no cooldown.

The main issue is the way it destroys weapon itemization for shaman. There are no acceptable epic OH weapons pre-Black Temple/Eye except the gladiator 2.6s. Same holds true for blue level 70 weapons.

Yet there well over a dozen, high dps fast OH weapons (daggers, axes) that kill your DPS. (Slow Level 60 green thru epic weapons outperform them).

This is game breaking, yet Blizzard has ben *silent* on this for MONTHS! We have reported this issue on the PTR, BUG and this forum going back a long time.

I am using a Level 60 weapon in my OH i got a year ago, and yet I have a 93dps epic weapon sitting in my bank because the damage sucks.

There needs to be a WF cooldown.

It needs to be a function of the base weapon speed.

There should be a seperate cooldown for the MH and OH.

Special attacks, like Stormstrike should ignore the cooldown and not trigger it.

2handed weapon attacks should ignore the cooldown and not trigger it.

Blizzard should fire the lame-brained dev or programmer who came up with the current implementation. It is an embarrasment to the company.

I ran WindfuryFu for FuBar, and over the course of 7,700+ swings (with WF/WF) the addon registered a 15% proc rate. I'll say here what I've said in a lot of other threads. Make Windfury work like Sword Spec does for Rogues and Warriors, let the off-hand procs do main-hand damage. It'll broaden the weapons you can use, and wouldn't be a game breaking boost in DPS.

---

Q u o t e:
WF should only proc 20% of the time for each weapon. If you're getting more than 20% procs, then that's a bug.

---

That would be the point of this thread...

---

Q u o t e:
I claim BS. If each weapon is proccing 18% of the time, that does not mean you're getting 36% WF procs. Let's say main hand swung 1000 times with 18% proc rate, that's 180 WF procs. Let's say off hand swung 1500 times with 18% proc rate, that's 270 WF procs. Together you have 2500 swings, of which 450 were WF procs. That's still 18% proc rate.

---

36% proc rate on an individual swing outside of the cooldown. One of the links is to a test where someone parsed a timestamped combat log, removed all swings (hits and crits) the occurred within 3 seconds, and calculated the true proc chance. It was 34% on the first test, and 36% on the second.

---

Q u o t e:
This is great. The guy uses WF + RB then shows that WF only procced 8.6% of the time. Well duh! Some %age of the swings were from the hand that had RB!

---

The Mod that person uses only count mainhand swings, so he needed an extra division by two to accommodate the offhand windfuries in the other tests. The undivided value (17.2%) is the correct value for the WF/RB test (and is what is predicted for the cooldown).

---

Q u o t e:
Until people (i.e. enhancement shamans) learn basic math, this topic is dead.

---

I know my math.