One of the big decisions a warrior has to make in his pvp talent spec is between the flat 4% dps increase of blood frenzy and the 15% white dps increase of flurry.

Now, I don’t know too many people who will argue that BF isn’t immediately superior if you add another physical dps source to your target, so I won’t waste time arguing that. Add a rogue or hunter on your target and BF pulls double duty, surpassing flurry easily.

But if it’s just you doing physical damage, how big is the gap between 3/5 flurry and 2/2BF? Not as big as you might think - in fact its virtually non-existent.

I based my calculations on the following assumptions. Note that the actual figure for “avg dmg per hit” isn’t that important because both abilities scale. Higher crit should favour yellow damage due to impale, but 33% is pretty easily obtained by any warrior.

In order to compare the talents I needed a figure for how much of your dps comes from yellow and how much comes from white,  because flurry only affects your white dps. For reference, I calculated that without using slam/hs/cleave, white damage makes up 44% of your dps. If you insert 1 slam every 10 seconds, white damage drops to 35%. I did not model for execute, hs, or cleave. Hamstring and pummel damage is pretty much negligible.

The scenario:

A 1 minute fight, using 16 white hits, 10 MS, and 5 WW (I dropped the 6th ww because GCD for MS and WW collide every 30sec, thus you could not actually do 10ms and 6 ww in 60 seconds)
A 33% crit rate (1 in 3 hits are crits)
A normal white hit of 700 damage.

The Shakedown:

3/5 Flurry provides 6.6% more overall DPS.
2/2 Blood Frenzy provides 4% more overall DPS.

BUT

Flurry costs an additional talent point. There are 2 ways to look at this:

Theoretical: BF is 2% dps per talent point, Flurry is 2.1% per talent point.
Practical: With 1 more point you could put it in 2h wep mastery for an additional 1% to your 4% of BF (5% total for BF vs 6.6% for Flurry).

You might not put it there, but that’s effectively what its worth since you could actually do that if you want max dps.

BUT WAIT! THERE’S MORE!

The original calculation does NOT include the use of slam or cleave/Heroic strike. I did not model the test to use hs/cleave as they are hugely rage inefficient and would be unfair comparisons. Everytime you use them you lose 1 of your white hits, so they quickly make flurry look terrible.

I did however re-model it to see what happens if you throw in some improved slams. The number of slams you can use in a minute is entirely dependent on your rage generation. If you had infinite rage, you’d never even do a white hit, instead slamming everytime MS was on CD. So flurry would do 0% dps in that case.

The fact of life is we don’t have infinite rage, and rage generation can vary wildly. If you’re rage starved, you might never slam. If you’re critting and getting second wind/mace procs, you might slam every other white swing. I chose to see what would happen if I slammed 6 times during that 1 minute fight, i.e. 1 slam every 10 seconds, which seems pretty reasonable.

The result?

Flurry would drop to a 5.25% overall dps increase (1.75% per talent point vs BF’s 2% per talent point).

Why is that you ask? Because flurry only increases my white DPS, while BF increases both white and yellow dps. The more my damage that comes from yellow sources, the less flurry is doing for my overall DPS.

Also, each slam takes up 0.5s of my dps time, so right away if I do between 1-7 slams, I lose 1 white swing (3.6s to do 1 swing). 8-14, I lose 2 white swings. etc. I modeled on 6 slams so dropped from 16 white swings to 15.

As you can imagine, its even worse for HS/Cleave, because every SINGLE cleave/hs eliminates a white swing. As I said, there’s no point modelling this as it will make flurry look horrendous very quickly and we know that’s not accurate.

I did not bother trying to factor in execute damage because it’s nearly impossible to do - it depends how much time your target spends below 20%, how much rage you have, etc. It’s there, and its more yellow damage - probably a significant amount, but its nearly impossible to quantify. For my model I guess just assume the target never went below 20% in that 1 minute :p

So… Are there any redeeming qualities for flurry?!

There is in fact 1 very significant advantage to flurry. Because it affects white dps, it affects rage generation. Flurry produces 11% more rage from white damage than BF does.

Thats roughly like having a half-point in Endless rage.

This is mildly mitigated by the fact that you can take Tactical Mastery with BF but not with flurry. The more you stance dance, the more rage TM is “generating” (by saving 15 per switch) for you, in addition to opening instant disarm/spell reflect doors.

So what’s the bottom line?

In the absolute best case scenario for flurry (you never use slam, HS, or Cleave), it will do 2.6% more dps for you than blood frenzy would.
If you insert a few slams, and compensate for that missing talent point (consider it at worst a 1% dps increase), flurry does pretty much identical dps to blood frenzy.
If you add another source of physical dps, blood frenzy completely blows 3/5 flurry out of the water.

Flurry’s advantage is producing 11% more white-dmg rage than BF, which is like half-endless rage. It’s primary disadvantage is that you can’t take tactical mastery, which opens doors to instant-reflect/disarm/ww and saves rage when switching stances.

When all is said and done, 3/5 flurry and 2/2 blood frenzy are almost identical in terms of dps and utility. (unless you’re a noob who never uses slam, and never pvp’s with other physical dpsers, in which case flurry will give ~1.6% more dps than BF)

But at the end of the day we’re talking 1% (or less) dps difference, and you don’t see people funneling points into 2h mastery, nor re-rolling human/orc for the 1% crit. The choice really is yours.

Read the rest of this entry »

Veritas

There are things in WoW I call ancient mechanics, it’s things that are completely broken, retarded, yet still in game for some odd reason, for example one of them is the fact that deep wounds break cowering in fear effect of intimidating shout, effectively making us unable to CC someone after we crit, it’s the fact that blood craze reset whenever we get critted without giving us a tick and few other small things.

One thing that is recently floating over forums is not as minor as deep wounds/blood craze problems - it’s the Berserker Stance penalty. It was taken to public in past, before TBC, without results. We all know how berserker stance works:

An aggressive stance. Critical hit chance is increased by 3% and all damage taken is increased by 10%.

If you are not familiar with the warrior class, you can probably already smell the bullshit. 3% crit has became so minor buff since epix started to rule the world, it’s not really all that good anymore, warriors usually have extremely high crit chance, and with resilience being in game, 3% crit is not a big dps increase. Now let’s discuss this from 3 points of view: Lore, Common Sense and WoW Balance, here we go:

1. Lore

What is berserking? It’s fighting madly without worrying about self, as such berserker stance makes sense? Not really, no. Let’s take a look at similar mechanic: Barbarian’s Rage ability in mother of all rpg mechanics - D&D (3.5 version):

A barbarian can fly into a rage a certain number of times per day. In a rage, a barbarian temporarily gains a +4 bonus to Strength, a +4 bonus to Constitution, and a +2 morale bonus on Will saves, but he takes a -2 penalty to Armor Class. The increase in Constitution increases the barbarian’s hit points by 2 points per level, but these hit points go away at the end of the rage when his Constitution score drops back to normal. (…) A barbarian may prematurely end his rage. At the end of the rage, the barbarian loses the rage modifiers and restrictions and becomes fatigued (–2 penalty to Strength, –2 penalty to Dexterity, can’t charge or run) for the duration of the current encounter (unless he is a 17th-level barbarian, at which point this limitation no longer applies; see below).

And the last “ranks” of Rage:

Tireless Rage (Ex): At 17th level and higher, a barbarian no longer becomes fatigued at the end of his rage.

Mighty Rage (Ex): At 20th level, a barbarian’s bonuses to Strength and Constitution during his rage each increase to +8, and his morale bonus on Will saves increases to +4. The penalty to AC remains at -2.

Armor Class is not what Armor in WoW is, AC increase one’s avoidance and is used for describing everything like dodge chance from Dexterity, Monk’s increased avoidance from Wisdom, Duelist’s increased avoidance from Intelligence and (of course) AC from Armor protecting body and/or shield. So, in the end how does Rage affect Barbarian in D&D? Let’s make a break down:

• Increases his damage thanks to increased STR.
• Temporarily increases his HP thanks to increased CON (I died once in D&D: Online when my rage ended while I had like 8 hp left).
• Increases Will saves to help against spells that work against mind (fear, charm, paralyze etc.)
• Reduces avoidance due to lowered AC.

Barbarian becomes stronger and uncontrollable but is unaware of attacks, you don’t care about taking a hit when in Rage, but does it says anywhere that you take more damage while in Rage? No, if anything you are tougher due to temporary hit points.

Now WoW doesn’t work same way, but we can see some similarities, while in berserker stance we can use berserker rage ability that breaks fears (Will saves from morale in D&D), we have, technically, slightly higher damage done due to 3% crit (woohoo!) and take a whooping 10% more damage then everyone else in the game. Death Wish, which (despite of being moved to Arms) is a similar category of abilities also increase damage done, and again increase damage taken by 5%. It’s worth to note that in past Death Wish used to lower our Armor and Resistances by 10%. I find it funny that original developers of WoW assumed we would have some spell resistance on our gear in endgame when implementing this ability, I mean, it would make sense for a tank class to be durable against spells back in the past, but since Kalgan is taking care of WoW PvP logic is not a strong part on Blizzard. From Lore PoV it would make sence if Berserker Stance lowered our dodge/parry chance, or even lower our armor (increase damage taken against physical attacks), being Berserk means you are going offensive, you don’t defend yourself, but there should be no penalty against things (spells) you already can’t defend against.

Read the rest of this entry »

Terhix

One of the warrior buffs in 2.3 was a change to the pathing on intercept and intervene to make us less likely to have these abilities blocked by objects in the way (think of the pillars in the arenas).

A side-effect of this however has given us a bit of an exploit for warsong gulch (possibly blades edge arena as well - haven’t had a chance to try). Basically if a player is above you but technically “in range” for intervene, intercept, or charge you can now use the ability and you will be zoomed along to that point on the map.

In the video below, a healer waits on the roof, and when the warrior flag carrier gets into trouble he intervenes to the priest - sending him running at intercept speed down the tunnel, up the hill to the roof to get easily healed by the priest. Definitely a bug that will be fixed, but worth noting if you happen to find yourself in the gulch and see enemy warriors disappearing with the flag.

http://files.filefront.com/warsong+bug+ownagewmv/;9061738;/fileinfo.html

Veritas

I keep seeing people on this site, and on the official boards talking about improved execute. Now talking about it is fine - if you talk about it like improved rend and improved cleave, but these people are talking about it as a talent that is worth taking.

I’ve done piecemeal arguments against taking it, but this will be my coup de grace on the talent, and will hopefully seal the deal against it for the vast majority (if not all) of you.

For the low cost of 2 talent points, improved execute will drop the up-front rage cost of execute from 15 to 10. I’m not sure why people think this is good, but I can only assume it’s because they don’t know what execute does. Rank 7 (level 70) reads:

Attempt to finish off a wounded foe, causing 925 damage and converting each extra point of rage into 21 additional damage. Only usable on enemies that have less than 20% health.

So, with improved execute, if you had 15 rage or more improved execute has converted 5 more rage points into damage - for the insane increase of 105 damage. To put that in perspective, 2 points in improved cleave will add 56 damage (or 84 damage for 3 points), of course you can cleave whenever you want and it can hit 2 targets… but I don’t know anyone who takes imp cleave seriously, and being twice as good as crap is still crap.

The only place improved execute does anything remotely useful is when you’ve got your target below 20%, but only have between 10 and 14 rage (less than 10 and you can’t execute even with improved, more than 14 and you could have done it with non-improved). When I get into an execute-spam situation, it’s usually after each white hit that I can execute. Incoming damage is not enough for me to hit it every GCD unless I’m being focused - and if I am being focused then I should be getting 15 just as easily as I can get 10. Remember - it’s a FIVE rage margin for improved execute to help you.

If I’m relying on white swings, I can safely say my white swings generate 15 rage even versus highly armoured paladins and druids, and closer to 30 versus cloth. I tried to use the rage formula to outright prove this, but that thing is a nightmare and is tied to how much damage you do anyways. I’ve played a warrior a long time, and I often use slam after white hits coming from 0 rage, I’m confident in my estimates.

Some classes can rage starve you, making you think you should have this talent in your arsenal to tip the scales in your favour. But if you manage to get them to 20% in the first place, you’ve obviously got the capability to generate enough rage against them. Execute is good because it packs so much damage so efficiently. It’s harder getting someone from 100% to 80% with MS/WW/white damage than it is getting them from 20% to 0% with execute and white damage.

To summarize, improved execute might as well read “Increase execute damage by 105 (or 11%)”. In a game where players pack 10-15k HP in pvp, that is not exactly impressive. It’s on the same level of the tree as enrage so you never “have” to take it. If you need 2 points to reach flurry take improved slam and/or weapon mastery instead. Oh, and if you do find yourself stuck at 10-14 rage all the time after white hits, put down the [Honed Voidaxe] and go get yourself a real weapon.

Veritas

I’ve done some math earlier today regarding this issue. 33/28 (blood frenzy in arms, not second wind) have improved slam, 3/5 flurry and weapon mastery while 41/17/3 provides Improved MS and ER. With 33/28 you do damage cycles like this:

Swing, Slam, MS | Swing, Slam, WW | Swing, Slam, MS | Swing, Slam | Swing, Slam, MS (WW is ready at this point, but can’t be used due to GCD conflict) | Swing, Slam, WW…

For the sake of simplicity, let’s assume that with 3.6s weapon and 3/5 Flurry your swings are done at the speed of 3.1s, when you add Slam to that swing, with perfect timing and 0 latency (never in reality) you will swing at the speed of 3.6, which is same speed as without Flurry. Because you have to time abilities between Slams and white hits, you are not using them every time they are ready, but rather every 2 swings = 7.2s for MS and every 4 swings = 14.4s for WW. With 41/17/3 you don’t have such problems with timing, you are using MS every 5 seconds and WW every 10 seconds, no GCD conflicts either. Here is the math done on those assumptions:

2000ap
130 weapon dps
3.6 speed

White dps: 272.8

MS damage: 471 + 468 + 210 = 1149

33/28 MS dps: 1149 / 7.2 = 159.6
41/17 MS dps: (1149 / 5) * 1.05 = 241.1

WW damage: 471 + 468 = 938

33/28 WW dps: 938 / 14.4 = 65.2
41/17 WW dps: 938 / 10 = 93.8

Slam dps: 272.8 + 140 / 3.6 = 311.7

Total 41/17 dps: 272.8 + 241.1 + 93.8 = 607.7
Total 33/28 dps: 272.8 + 159.6 + 65.2 + 311.7 = 809.3

And so 33/28 can produce up to 25% more dps then 41/17. That “up to” is the key here. First, using Slam every swing is huge rage eating. How much rage can you generate per swing? 20? 30? Sure you can Slam every swing, but you need to produce rage for using MS and WW as well. 41/17 will produce much more rage then 33/28 in this scheme of things. Also, this is assuming perfect reflexes and 0 latency, Now let’s do same thing assuming being 0.5 sec late with Slam, in this case your Flurried swings go up to 4.1second:

2000ap
130 weapon dps
3.6 speed

33/28 White dps: 272.8 * 3.6 / 4.1 = 239.5

MS damage: 471 + 468 + 210 = 1149

33/28 MS dps: 1149 / 8.2 = 140.1
41/17 MS dps: (1149 / 5) * 1.05 = 241.1

WW damage: 471 + 468 = 938

33/28 WW dps: 938 / 16.4 = 57.2
41/17 WW dps: 938 / 10 = 93.8

Slam dps: 239.5 + 140 / 4.1 = 237.6

Total 41/17 dps: 272.8 + 241.1 + 93.8 = 607.7
Total 33/28 dps: 239.5 + 140.1 + 57.2 + 237.6 = 674.4

Something as simple as being 0.5s too slow with Slam every swing is lowering your ideal dps with 33/28 by ~17%, slower swings means slower Slam and slower MS/WW cycles. Now god forbid you will have any rage issues with white hit every 4.1 second, comparing to 125% rage giving swings every 3.6 with 41/17. In normal instances and heroics I’m pretty sure 41/17 can easily compete if not beat 33/28 for dps, raidbuffed however the rage issues of 33/28 can be gone (windfury totem and whatever can you get), which allows you to use all your abilities to it’s fullest. Theorycrafting is fun, but we will have to wait a bit and see how it really works out in game, people will get right conclusions after few days.

edit: Oh, yes, you should also add ~2% more dps form WM to 33/28, which I haven’t done here.

Terhix