Understanding Fury in 4.0

[davidsmr]

Not sure if this was addressed already. With all the changes to fury, is Tiny abom in a jar worth getting now?

[targetman]

So, has anyone that has reached the hard hit cap of 27% had a problem with too much rage generation? or is that just never a problem?

I can imagine a scenario in which as we progress through new content and the gear gets better, we will be able to adjust out hit/haste ratio to maintain the "perfect" rage gen rate.

What that perfect rate is who knows yet.

I am sure that if you work hard enough someone could generate a differential equation to map out the rate of change of rage generation in regard to a variable gear level and hit/haste ratio.

However, and the point is, Fury dps seems to me to have become a completely different animal. In wrath it was a basically a faceroll rotation, while spamming heroic strike. As long as you had a nice big graphic to alert you when you got a bloodsurge proc, you could pretty much top the dps charts while watching Frasier reruns and eating a sandwich.(I apologize in advance for anyone offended by the thought that their class/spec was actually this easy to play, but it is the truth as I see it.)

Now, I find that fury dps is a more fluid entity, relying on paying close attention to your rage generation and enrage uptime. It actually feels(to me) like a more organic, "warrior-y" way to play, and I quite enjoy that. I feel like I am performing a balancing act or swimming in an ever-changing current, and I like the fact that I have to react to it.

So stop asking what the perfect numbers are for your stats, and what the perfect rotation is, because that doesn't exist anymore. We all know how to increase rage generation, and we all know how to avoid be rage starved, so all we need to do is find that sweet spot for the gear we are currently in.

I am sorry if I got a bit meta there, but all the discussion of mechanics can sometimes, for me at least, be a detriment to actually playing the game.

I very much appreciate all the hard work put into this theory-crafting, good luck with your equation, as I am certain the results will be extremely useful to all Furys(and possibly even some rogues). In the mean time, have fun swimming the rage/enrage currents, and try to keep your head above water.

[Xodiv]

Regarding Battle Trance...

I've spent some more time beating on the target dummy, and can confirm there definitely is a delay in Battle Trance procing after a Bloodthirst hit.

[Thegreatme]

it is entirely possible that haste could be better than strength until you get to the 'ideal' rage generation rate, but accurately determining the value of strength means modeling every single warrior ability and mechanic which is something that is better suited for a spreadsheet.

[Loganisis]

For Titan's grip

Over an infinate time scale Hit = more landed attacks than Haste per 1% (I promise I will get around to copying it from laptop soon). So really I can't see any scenario where for pure rage generation you'd want haste > hit.

Hit = fewer misses so rage generation is smoother
Hit = more landed hits so more rage.

I can see trying to get haste stacked to the point where Weapon Speed * (1 - haste %) = <3.0 seconds so you'll never go more than 2 GCDs without the chance of white swings (for 3.5 weapon seed that's 15% haste. 3.5 * .85 = 2.975). But that's it.

For Single Minded Fury I can see where Haste could be better than hit. With a 3.5 second swing timer, haste started resulting in more rage over a given interval (on average) than hit at >27% haste.

That is if you have less than 27% haste/hit, you'd see more rage with all hit and zero haste.
27% hit = 27% haste in terms of rage generation (27% hit 0% haste versus 0% hit 27% haste)
And once haste >27% then no hit and all haste would result in more rage generation.

I think that's because you start getting swings so often than the sheer volume of swings becomes overwhelming. So with SMF having shorter swing timers to start, it may hit that inflection point sooner, and it appears to grow compared to hit in value for every additional point (e.g. 28% haste > 27% haste + 1% hit).


This is ONLY for rage generation unless I screwed something up, so I'm sure what the exact impact it would have on DPS and in comparision to strength.

But at least for TG, all hit until 27% is better than any hit + haste in terms of rage generation if I did my math correct.

[Loganisis]

So, has anyone that has reached the hard hit cap of 27% had a problem with too much rage generation? or is that just never a problem?

I can imagine a scenario in which as we progress through new content and the gear gets better, we will be able to adjust out hit/haste ratio to maintain the "perfect" rage gen rate.

What that perfect rate is who knows yet.

I am sure that if you work hard enough someone could generate a differential equation to map out the rate of change of rage generation in regard to a variable gear level and hit/haste ratio.

However, and the point is, Fury dps seems to me to have become a completely different animal. In wrath it was a basically a faceroll rotation, while spamming heroic strike. As long as you had a nice big graphic to alert you when you got a bloodsurge proc, you could pretty much top the dps charts while watching Frasier reruns and eating a sandwich.(I apologize in advance for anyone offended by the thought that their class/spec was actually this easy to play, but it is the truth as I see it.)

Now, I find that fury dps is a more fluid entity, relying on paying close attention to your rage generation and enrage uptime. It actually feels(to me) like a more organic, "warrior-y" way to play, and I quite enjoy that. I feel like I am performing a balancing act or swimming in an ever-changing current, and I like the fact that I have to react to it.

So stop asking what the perfect numbers are for your stats, and what the perfect rotation is, because that doesn't exist anymore. We all know how to increase rage generation, and we all know how to avoid be rage starved, so all we need to do is find that sweet spot for the gear we are currently in.

I am sorry if I got a bit meta there, but all the discussion of mechanics can sometimes, for me at least, be a detriment to actually playing the game.

I very much appreciate all the hard work put into this theory-crafting, good luck with your equation, as I am certain the results will be extremely useful to all Furys(and possibly even some rogues). In the mean time, have fun swimming the rage/enrage currents, and try to keep your head above water.

If reading about it diminishes your enjoyment then don't read it.

There's going to a 'good' point of haste and hit to generate sufficient rage and land enough hits to stay enraged and after that it's going to be a set rotation in all liklihood. BT - TB - BT - HS repeat. Spam Execute. The only 'wrinkle' in the rotation is using Victory Rush procs to help out the healers. Maybe hitting beserker rage every now and then.

Not that there was much there originally, but visually with the loss of WW, Fury is even less visually stimulating than before. To me it seems like it's become very robotic.

Maybe I haven't played it enough post 4.0.1, but from what I have, that's the impression I get.

Gearing wise, I think it's far more dynamic.

I think the stat values are going to be based on Weapon speed because that will impact Haste's relative value. So even a 3.4 versus 3.6 weapon speed could make a difference in realitive stats distribution (a different amount of haste is needed to get under a 3 second swing timer for limiting the lulls in rage generation). So that part is more dynamic.

And SMF versus TG will have a bigger spreads in optimal stats me thinks than Arms and Fury in 3.X.

****


I don't remember who suggested it on these forums initially, but the idea of making haste affect melee GCDs is probably the only way to ever break warriors from having a set rotation. Until then, at most, you're just going to have 1 GCD or so to think about, which was quite similar to WotLK - can I fit this Slam in before the BT or not (did it proc soon enough).

[Ronninn]

Here are my Herioc Festergut logs from last night. http://www.worldoflogs.com/reports/b...?s=5664&e=5868

I reforged my Haste to Hit, and I eat Expertise food to get that up to 23/26. I've found that so long as my Expertise is above 22 I don't get any dodges at all. I've not Hit capped yet, around 22% in raid, and close to 45% crit and 21% haste.

I've been using the BT>RB>BT>Free rotation and hitting HS when above 60 rage. For the final 20% burn phase it's spam Execute and HS at over 60% rage.

I managed to pull over 21k dps on Festergut this week. If you check out last weeks mess you'll see that I couldn't even pull 10k and the tanks beat the 2 fury warriors. The buffs we picked up and getting used to the rotation have been a substantial gain in dps.

http://www.wowarmory.com/character-s...has&cn=Ronninn

Edit, for Free I've been using Bloodsurge Procs, or shouts. I use my Deathwish at the start of the fight to coincided with DBW proc, then when that expires if I'm not enraged I use my Berserker Rage. If I have a bad streak without any enrages I'll slip in my Recklessness for the 12 seconds of enrage, and the use Berserker Rage on cooldown when I'm not enraged.

I also make a point of saving my last Deathwish for the Execute phase of the fight.

[Quietsch]

Hit/haste:

I still fail to see why people are saying hit is better than haste. for a 1% increase of either stat you get a 1% net increase of attacks landing, resulting in the same DPS gain, and the same rage gain.

I could easily be missing something here, and if this is indeed the case, let me know.

So, i finally got the mathematical, and because of that understandable solution for this. :P

The base is the hit per second formula. 1/(Weaponspeed/(1+haste%))*(1-(0,24-hit%))
You can convert (not sure if the word is right) it into ((1+haste%)*(0,76+hit%))/weaponspeed.
Now you have the two scalingfactors of the hits per second, haste and hit, with have each a constant (1 respectively 0,76 (and in the case that you include buffs, also flurry and the 10% haste buff)) and a variable, which you can influence.
Because of those constants, hit is 'stronger' than haste, because the effect of hit on the hit-part of the calculation is higher, as the constant is lower, so the region (and again, i can't find a appropriate word...) which is hit is greater.
for example you can compare 10% haste to 10% hit.
10% haste = (1,1*0,76)/weaponspeed = 0,836/wpnspd
0,076/wpnspd hits more, compared to 0% hit/haste (10% more)
10% hit = (1*0,86)/weaponspeed = 0,86/wpnspd
0,1/wpnspd hits more, compared to 0% hit/haste (~13% more)

so you see, that you get more profit out of 10% hit, than 10% haste, so hit > haste.

Other question: Is rb normalized, or not? As i can't login atm, i can't test it.

[Sparrta]

I've researched quite a bit and have not found a clear answer on what fury warriors should be gemming in their yellow sockets. It seems to be between +10str +10haste or +10str +10crit. So TGM or any other elitist on this forum, which is better at the present moment?

[Xodiv]

Regarding Enrage...

It is clear from the combat logs that you cannot be Enraged (+10% dmg) while Death Wish (+20% dmg) is up. Casting Death Wish will cancel any current Enrage buff, and Enrage does not proc while Death Wish is up. This either reduces the value of Death Wish, or reduces the value of gearing for increased Enrage uptime, depending on how you look at it.

@Sparrta: nobody knows for sure, and really until 85 it doesn't matter that much. Either would be fine. There is not yet a good enough combat model to accurately answer those kinds of questions.

[Ion]

Regarding Enrage...

It is clear from the combat logs that you cannot be Enraged (+10% dmg) while Death Wish (+20% dmg) is up. Casting Death Wish will cancel any current Enrage buff, and Enrage does not proc while Death Wish is up. This either reduces the value of Death Wish, or reduces the value of gearing for increased Enrage uptime, depending on how you look at it.

@Sparrta: nobody knows for sure, and really until 85 it doesn't matter that much. Either would be fine. There is not yet a good enough combat model to accurately answer those kinds of questions (and it's debatable if it would be worth making one for a rotation lasting just a few months and likely to be tweaked in the meantime by Blizzard).

You don't gear to increase enrage uptime, particularly, it's just a side-effect of gearing to maximize rage gain.

[Xodiv]

@Ion. I think you have misunderstood me. I am not suggesting people are doing so, and I'm not focused on advising people on how to gear for 4.0.1 because until 85 we are in junk time. My goal is to gather evidence and math towards a comprehensive combat model - what is needed to answer questions about what stats and gear produce the best DPS.

In a Fury warrior combat model increased Enrage uptime has a DPS benefit, and stats that increase Enrage uptime have a DPS benefit. As a result of Enrage and Death Wish not stacking that value, that formula, and contribution to the value of those stats from that reason, is reduced.

Perhaps I'm confusing people by posting all this in the wrong thread, and I should start a new one titled "Fury Theorycrafting for 4.0/Cataclysm" which is separate from advice for today.

[Ion]

This thread is very specifically about how things are now.

Things WILL change, speculating about how we're going to gear at 85 is fairly pointless because I'm absolutely positive that they won't let Heroic Strike be our best attack like it is right now. The design goal for Heroic Strike is to be used as a rage dump, not to be the #1 attack in the first 80% of the fight.

Also, my point stands...Enrage's uptime is modified by exactly the same stats that rage generation is. It's a side-effect, not the goal. That, I am fairly sure, will remain the case. The idea isn't for warriors to be in an infinite rage situation anymore...so gearing for rage gen is basically what we're going to be doing. A side effect of that will be increased Enrage uptime (outside of Death Wish).

[Xodiv]

The base is the hit per second formula. 1/(Weaponspeed/(1+haste%))*(1-(0,24-hit%))
You can convert (not sure if the word is right) it into ((1+haste%)*(0,76+hit%))/weaponspeed.

Error in my math, removed.

[Quietsch]

expertise is irrelevant, as dodged swings grant the same rage like landing. (or i have looked wrong at my ragebar)
and what you're writing there is exactly the same, as i did, until you don't to the step from ((1+haste)/weaponspeed)*(0,76+hit) to ((1+haste%)*(0,76+hit%))/weaponspeed
(x/y)*z is exactly the same as (xz)/y
.27 is in fact the same a my .24, as you take the rating from the paperdoll with factored in precision, while i mean only the % over rating.

[Xodiv]

You are quite correct.

[Garbid]

1 thing I don't get, maybe I'm just confused or my gear difference to those of you posting so far has some effect. But the HS > BT > RB prio doesn't seem right to me. Did a few tests on the dummy and this is what I found out:
HS = 359 dmg per point of rage
BT = 384 per rage
RB = 521 per rage
Slam = 236 per rage
These numbers will of course vary from player to player based on gear, but I doubt the order of them changes. I have no extra mastery so RB can only get better.

So even if your doing your highest crits with HS, it's just not worth prio it over BT and RB.

My rotation is therefor:
BT -> RB -> BT -> filler
and by filler I mean heroic throw or battle/commanding shout for rage (use slam only if your overflowing in rage)
And add in HS whenever I can without ragestarving me (and it's usually spamming though I do have to hold back once in a while).

These are my stats unbuffed if you are wondering:
5305 AP
19.99% haste
20.59% hit
34.05% crit
32 expertise (yes I know, I'm usually prot and only get dps leftover gear)
8 mastery (the innate 8, nothing reforged to it)
271 and 264 2-handed weapons.

[Blubbey]

Not sure if it's been mentioned but I saw a post on the WoW warrior board saying the SMF damage buff carries over to TG unintentionally (aka a bug). http://forums.worldofwarcraft.com/th...87359427&sid=1

I'm unable to play at the moment, would anyone like to try it out?

[Kazeyonoma]

while damage per rage is an important statistic, when caring about full on damage output, you want highest damage output priority. so if HS is your hardest hitting, you prioritize your rage for it. and work to maintain a safe haste level to keep up HS, while the rest of your rotation.

[Ion]

while damage per rage is an important statistic, when caring about full on damage output, you want highest damage output priority. so if HS is your hardest hitting, you prioritize your rage for it. and work to maintain a safe haste level to keep up HS, while the rest of your rotation.

Additionally, only Bloodthirst procs Battle Trance, so prioritizing that saves you a fairly significant amount of rage off of your total usage...if you hit BT on every CD for a 4 minute fight you'd get about 12 Battle Trance procs on average, potentially saving you 360 rage (if you managed to use it for HS each time)...which is no joke...18 raging blows worth of rage.