Art of Burst - PvP Feral

Good <Time of Day> my fellow PvPers,
I want to talk to you today about an important topic to most PvPers - Burst Damage. I think I am in a fairly unique position to give some insight and details into the subject. For one, I am a voracious reader and forum troll. Because of this, I am well aware of how little information there is on this subject in existence. In addition to that, I am a multi-season 2200-2300 rated player and life-long PvPer. As such, I have a solid understanding of Arena and PvP fundamentals. Finally and most importantly, I'm something of a math nerd and have had the good fortune of having had Yawning graciously volunteer to modify Mew to assist me in my analysis.
I really really apologize for this being so long. Unfortunately, sometimes it takes a lot of words to say something worthwhile. As much as reading this might suck, trust me... writing it was pretty sucky too.
The aim of this post, as stated above, is to analyze burst damage in PvP. It's always been my philosophy to inform people as well as I can, so that they can make the best decision for themselves. Hopefully, that's what you will find here - information that helps you become more informed and better PvPers. For the most part, I think this material is directed at more advanced PvPers. Although, I'll say a few words initially for beginners.

I'd like to thank Yawning for modifying Mew, such that I was able to run some PvP-focused simulations. I'd also like to thank Ellesime, Taswind, Yawning, and Datah for reading this and giving feedback.
Finally, if you'd like to hear more from me, you can come hear myself and other knowledgeable druids on the Team Waffle Podcast:

Table of Contents
  • Section 1 :: For the Beginners
  • DPS
  • Burst Windows
  • Section 2 :: Burst Rotation "Derivations"
  • PvE Rotation & Resource Limits
  • Burst Rotation & Resource Limits
  • Burst Rotations
  • Section 3 :: Mew Rotation Results
  • Section 4 :: Applying These Ideas in Reality
  • Section 5 :: Haste as a Burst Stat?
  • Section 6 :: A Small Word about Crit
  • Section 7 :: Conclusion
  • For the Beginners :: DPS
  • PvP in this game has always been about resource management. In that regard, the game is simple: When two groups enter into a conflict, each team has a certain amount of resources at their disposal. The team which runs out of a type of resource first, loses.
    At the most basic level of play, you bash on the other team until the other team has run out of mana. This introduces the first criterion of damage in a PvP setting. You must do enough sustained DPS over the course of the fight such that:
    1) The enemy's healing is depleted.
    2) It's depleted faster than your own team's healing.
    If you can accomplish that goal, you will win the fight. At this level of play, controlling the enemy healer serves more to force them into healing less efficiently when they are free of control - thus depleting their mana faster. Likewise, controlling an enemy attacker is used more to slow down the other team's DPS, allowing your healer to perform more efficiently, putting your team ahead in the fight.
    Once you have the advantage in the fight, your team can ride that advantage to victory or leverage the advantage by having your healer contribute offensively - leading to a speedier win.
    My first advice to beginners - master this style of play.
    Feral damage is dependent on what we call: DPE or damage per energy. To steal some numbers from the Fluid Druid Blog:
  • Rip ~ 4200 DPE
  • Rake ~ 2000 DPE
  • FB ~ 1000 DPE
  • Shred ~ 600 DPE
  • Mangle ~ 400 DPE
  • The basic recipe for doing solid DPS in any fight is just to have solid Bleed Uptime (You can clearly see Rip/Rake crushing the DPE numbers) and Spend Your Energy. You regenerate roughly 10 energy every second. Every second you sit at full energy, you're wasting a minimum of 4000-6000 damage each second.
    But Clay, "Top PvPers say to pool your energy!"
    That's right they do! Firstly, that's not for you. You're a noob, so you spend your energy! Secondly, that's not really what they say. They say work with the upper 40% of your energy pool. Finally, that's still not *quite* the right thing to do anyway, and I'll get to that later.

  • For the Beginners :: Burst Windows
  • When you get better at the game, you begin to realize that health and mana aren't the only resources a team has which can be depleted. Teams also have CC-breaks and defensive cooldowns. All of which are limited. If you can cause a team to deplete either of those resources, you can win the game. Enter Burst Damage.
    By doing enough damage in a short period of time, you can put the enemy team in a position where they must use a major cooldown or die. If you can do this to the other team frequently enough, you will run them out of major cooldowns and your next round of burst damage will kill one of them.
    This brings us to the second criterion of PvP damage. Do enough damage in a short period of time such that:
    1) The other team must react defensively or die.
    2) Do it frequently enough such that they will run out of defenses.
    Key to this style of play is coordination. The more of your team's resources you can coordinate and bring into these timing windows, the stronger the effect will be. Since this is not a PvP guide, but rather a burst guide for ferals, I'm not going to even try to discuss the greater engine of coordination needed. The answers down that path are highly composition/situation dependent. Rather, I'm going to clarify for struggling intermediate players what your burst resources are.
    Ferals have four major burst resources at their disposal.
  • 1) We have Tiger's Fury, every 30 seconds
  • 2) We have a Free Ravage, from Feral Charge Cat, every 28 seconds
  • 3) We have on-use trinkets, usually available every 2 minutes
  • 4) We have Berserk available every 3 minutes
  • Since we know defensive cooldowns refresh on the order of 2-3 minute time scales, we can see immediately that Berserk and On-Use-Trinkets alone are not available frequently enough in order to deplete defensive cooldowns. Especially if they're paired together. For example, if it's your intention to use berserk to force a trinket, by the time your berserk is available again to push for a kill, the trinket will be available as a counter again.
    What this means is that your bread and butter for forcing defensive cooldowns can only be:
    Tiger's Fury and Feral Charge Cat -> Ravage!
    To put that more plainly: If you cannot force the use of a defensive cooldown by coordinating your Tiger's Fury + Free Ravage, you cannot win the game by depleting their defensive cooldowns. (assuming equal contributions from your teammates)
    My advice to intermediate players:
    Sit down with your team, learn the timings of their short offensive cooldowns (1 min or less) and plan to coordinate them in burst windows. At first, just throw the whole kitchen sink at the other team as often as possible. As you climb to higher ratings, you will find teams which manage to thwart your kitchen sink strategy by introducing clever defenses. In return, you will have to think about breaking your burst attempts into smaller ones (or more protected ones) - ideally using the minimum amount of cooldowns necessary to elicit a cooldown response from your opponent. Such that you find yourself with a few offensive cooldowns left and they have none left to defend with.
    What do you do with the TF/Ravage exactly? Well that's what the rest of this post is about. It mostly boils down to trying to FB, without sacrificing too much of your Rip uptime. Most beginners can probably get away with shred-spamming off of their TF. Pocketing the combo points generated to refresh Rip. It's more important that you are remembering to leverage your TF/Ravage against the other team frequently and coordinating that attempt with your teammates.

  • Section 2 :: Burst Rotation "Derivations"
  • I'm going to attempt to ground most of what follows in math. In order to hopefully inspire confidence in the relevance of my math, I'm going to start with something which is well codified and matches everyone's expectations and build from there. Hopefully, by seeing how the math agrees with your instincts in things which are already known, you will have faith to trust the math when it deviates from your instincts on more complex subjects. Most of it is simple algebra and I'll explain the 'meaning' of each conclusion as we go. If you trust me (and you shouldn't, because I' notoriously careless), you can skip ahead to the Rotation Section.
    Some important numbers:

    Energy Costs
    Maintaining Mangle Costs ~ 0.5 energy per second
    Maintaining Rip Costs ~ 1.36 energy per second
    Maintaining Rake Costs ~ 2.33 energy per second
    Maintaining Savage Roar ~ 0.5 energy per second
    each FB per 30 seconds ~ 0.833 -> 1.66 energy per second

    Energy Gains
    Base Energy Gains ~ 10 energy per second
    TF Energy Gains ~ 02 energy per second
    Total Energy Gains ~ 12 energy per second

    Combo Point Costs
    Maintaining Rip Costs ~ 6.8 combo points every 30 seconds
    Maintaining Savage Roar ~ 3.3 combo points every 30 seconds
    Each FB per 30 seconds ~ 5.0 combo points every 30 seconds

    Combo Point Gains
    Maintaining Mangle Gives ~ .75 combo points per 30 seconds
    Maintaining Rake Gives ~ 3.0 combo points per 30 seconds
    OOC Gives ~ 2.6 combo points per 30 seconds
  • Section 2 :: Burst Rotation "Derivations" :: PvE Rotation & Resource Limits

  • A lot can be learned by analyzing the resources that we have to spend doing damage. I'm going to start this out by investigating the energy budget of a basic PvE rotation... actually, I'm going to derive it. This will give us a good idea of how much of our energy-resource is needed to maintain a backbone of DPS. Following that, I'm going to extend that method to discover what is possible or not possible in the context of burst damage. Experienced ferals have a rough idea of what we can accomplish with our energy budget. For example, we KNOW we can't do 10 ferocious bites in a minute, while keeping bleeds rolling. It's just not possible. But what are the limits exactly? Is it 3,4,5? Let's find out!
    In a given 30 second interval, in the middle of our fight, we will have 360 energy to work with (this includes Tiger's Fury). Expressing this in terms of a regeneration rate, We have an energy budget of 12 energy per second.
    With our knowledge of the DPE of the various moves, we know our highest priority is to apply RIP with the Bleed Debuff, so we do that at the cost of 1.86 e/sec. This leaves us 10.14 e/sec left to spend.
    However, we can't apply Rip without building 5 combo points. OOCs spent on Shred will generate a healthy 2.6 combo points for free. Meanwhile, upkeep on the Mangle debuff generates 0.75 combo points every 30 seconds. Those following along will notice, we still need to generate an additional 3.45 combo points before we can apply the Rip.
    We can do this by Shredding 2.3 times at a cost of 3.06 e/sec. I know we can't actually Shred 2.3 times. Think of that as an average. Most of the time, it takes 2 shreds to generate 3 CPS, sometimes it takes us 3 Shreds.
    After subtracting the cost of generating 5 combo points, we have 7.07 e/sec left in our energy budget.
    The next most important DPS move we wish to add to our rotation is Rake. Keeping Rake rolling costs us an additional 2.33 energy per second. However, we expect it to give us 3 combo points to spend in our 30 second time-window. Tabulating our leftover resources gives us:
    4.74e/sec and 3.0 combo points
    Our next highest DPE move on the list (unfortunately not listed on the fluid druid table - but trust me, it's next!) is Savage Roar. Buying it costs very little energy... just 0.5. However, it will cost us 3.3 combo points every 30 seconds. Since keeping Rake up generates 3 combo points, it almost balances out the cost of adding Savage Roar to our rotation. Almost.
    Once again, we have to generate the combo points to make up for the deficit. It only takes .2 of a Shred in order to get .3 combo points. The energy cost for that Shred is .266 energy per second.
    Leaving us approximately 4e/sec. At this point, the rotation matches pretty well with the beginning PvE rotation... but we still have energy left to spend. So why don't we see about adding some Ferocious Bites to the rotation.
    The most energy efficient Ferocious Bite costs 1.66 energy and 5 combo points. Let's see if we have enough energy left in order to generate the 5 combo points we need. We will have to Shred 3.33 times at a cost of 4.44 e/sec, in order to get 5 combo points. That brings the total cost of Ferocious Bite up to 4.44+1.66 = 6.0 energy/second. More than is in our 30 second window budget.
    But Wait a Minute, we have an excess 4.0 e/second. Can we just wait a little longer, bank that energy, and use FB? Absolutely! If we just wait about 15 seconds, we'll have banked enough energy to Bite.
    And now we have spent our entire energy budget. Resulting in the following basic rotation:
  • We maintained: Rip/Rake/Mangle/SR
  • Shredded to generate combo points
  • Used TF on cooldown
  • And sneak a Ferocious bite in roughly once every 45 seconds.
  • Does this sound roughly like your basic PvE rotation? It should. In fact, that's how it was originally derived by Toskk and Yawning.
    Just for kicks, I've compared the analytical results to the results from running Mew:
    Over the course of a 900 second fight, this would be:
    19 Ferocious Bites
    192 Shreds
    41 Rips
    60 Rakes
    15 Mangles
    20 Savage Roars

    Compare that to a 900 second run in Mew (berserk disabled), it doesn't look too far off the mark:
    12 Ferocious Bites
    240 Shreds
    39 Rips
    62 Rakes
    16 Mangles
    28 Savage Roars

    With some basic algebra, you could have immediately jumped to the same conclusion by writing the two equations:
    Energy = 0 = (12 - 1.36 - 2.33 -.5 -.5)t - 40NS - 50NB
    CPS = 0 = (3 + .75 + 2.6 - 6.8 - 3.3)*(t/30) + 1.5NS - 5NB
    NS = number non-OOC shreds
    NB = number of Ferocious Bites
    and solving for NB(t)

    I went through the effort of building the same results, because I thought it would be more intuitive for non-math people to understand the idea behind the equations.
    There are a few things I could do to improve upon this technique (estimate combo points lost due to over-running the cap, take into account bleed clipping time)... However, I don't want you to lose the forest for the trees. The big idea is - we can get a rough estimate of what rotation we should be using by analyzing the resources available to us. For DPS, we expressed an energy budget, then spent it by choosing abilities that maximized our DPE usage. In PvP we still basically want to keep 'good' sustained damage, so we want to keep the same basic structure... just modified to make it more bursty. Maybe we can throw a few more ferocious bites in? Let's take a look at it.

  • Section 2 :: Burst Rotation "Derivations" :: Burst Rotation & Resource Limits
  • Energy
    So first, let's look at what's clearly not possible to do. We know that just maintaining 100% uptime on Rip/Rake/Mangle/Savage Roar uses up pretty much all of our available resources. Getting a Ferocious Bite in every 45 seconds is pretty weak as far as burst goes. This is before we even consider the added costs of kicking, maiming, or energy lost while in CC. Clearly, we need to free up some resources.

    Modification 1: Drop SR, Use Feral Charge Cat:
    The first thing everyone cuts from the PvE rotation, because it's the lowest DPE, move is Savage Roar. Under PvP circumstances, Savage Roar is actually lower DPE than FB, because you need 80% uptime for them to have equal DPE. Cutting SR from our rotation nets us an additional .5 e/sec as well as 3.3 combo points every 30 seconds. Also adding Feral Charge Cat into our rotation gets us about 2 combo points. Immediately, we can see something nice has happened.
    With a base DPS rotation of:
  • Maintaining Rip/Mangle/Rake
  • TF / Charge on cooldown
  • We now have: 4.5 e/sec and 5.3 combo points to spend.
    Which means we can Ferocious Bite every 30 seconds and still have.
    2.9 e/sec and 0.3 combo points to budget to other things... like kicks or occasional maims.
    Maybe 1 Ferocious Bite every thirty seconds isn't enough burst for you... Just how far are we away from a rotation with 2 ferocious bites? Every 41 seconds you will have saved enough resources for a 2nd ferocious bite. So roughly every other Burst period, at best, you can have a 2nd FB. This of course assumes you haven't "wasted" any energy on kicks or got caught in a hibernate... is it the waffles that we give each other that make us sleepy?
    With this small modification to your rotation, you should see an increase to your overall PvP-DPS. All you have to do is strive to push out a burst window every 30 seconds. And every 2nd or 3rd burst opportunity, you should have excess resources available for a 2nd finisher.

    Modification 2: Change some Shreds into Mangles:
    It's easy to see how we can do a small tweak to increase the impact of our burst windows. Our kill attempts occur in a small time windows around our use of TF. Outside of burst opportunities, mangle and shred compare 'favorably' in terms of DPE. So if you exchange the shreds you use to maintain rip with Mangles, you will gain an additional .4 e/sec... which you can put toward your burst window... or consider it energy available for interrupts. This puts us at 3.3 e/sec and .3 combo points excess every 30 seconds. Or 39 seconds needed to save up for a 2nd Ferocious Bite.

    Modification 3: Cut Rip uptime:
    It would be nice to be able to stretch that a little further, so that under ideal circumstances we would be able to push 2 50-energy bites every 30 seconds. However, this will cost us more significantly in terms of DPS. We have to cut Rip uptime in order to free up the resources. Dropping Rake uptime won't help because the limiting resource is combo-point generation. Any Rakes we drop from our rotation will have to be made up w/ Mangles.
    The next step is a little more complicated, so I just did it. I put some the equations into Mathematica to generate some data for you. I've hardcoded the number of Shreds to be 3.33 - during the burst rotation:
    {0 == (12 - 1.36*A - 2.33*B - .5)*t - 35*NA - 40*5.0/1.5 - 50*NB
    && 0 == (3*B + .75 + 2.6 - 6.8*A + 1.8)*(t/30) + 5.0 + 1.5*NA-5*NB
    && 0 <= t <= 30 && NB == 2.0 && 0 <= A <= 1.0 && B == 1.0 && NA >= 0},
    {A, B, NB, t, NA}

    A = Rip Uptime, B = Rake Uptime, NB = #FB, t = time, NA = #Mangles or Rakes
    If you wish to sacrifice a little Rip uptime in order to get more bursty kill windows, this is how much Rip uptime you will have versus Ferocious Bites per 30 seconds:

    FB Rip Uptime
    1.4 <100%
    1.6 <91%
    1.8 <74%
    2.0 <57%
    2.2 <41%
    2.4 <24%
    2.6 <7%
    2.8+ -- can't be done (on Average)

    So to quickly summarize the key conclusion of our analysis of the energetics: If we're attempting to push a burst window every 30 seconds with the TF/Ravage cooldowns, we only have two options. We can do 1 ferocious bite, every thirty seconds, with resources to spare while keeping our bleeds up. Or we can do two ferocious bites at a cost of roughly 50% of our Rip uptime.

    I'm sure some of you will be complaining at this point, because assessing resource management isn't actually enough information to tell us if we can do something. Timing is also important. Just because you save energy from the first 5 seconds of a fight, doesn't mean that you can carry that energy over to the 30 second mark. So some consideration for timing should be made.
    In a way, we can carry energy forward in the fight - by saving combo points and OOCs. Depending on how many combo points you have going into your burst phase, you'll be able to accomplish more, faster.
    Here's a quick run-down of how long you will take to Ravage + 2 Ferocious Bites, assuming you enter with X number of combo points.

    CPS Saved -------> average time to do two 50 energy FBs
    0 CPS ----> 13.2 seconds
    1 CPS ----> 11 seconds
    2 CPS ----> 8.7 seconds
    3 CPS ----> 6.5 seconds
    4 CPS ----> 5.8 seconds
    5 CPS ----> 5.2 seconds

    What if we don't use FC-Cat to Ravage?

    0 CPS ----> 17.2 seconds
    1 CPS ----> 15 seconds
    2 CPS ----> 12.7seconds
    3 CPS ----> 10.5 seconds
    4 CPS ----> 8.3 seconds
    5 CPS ----> 6.1 seconds

    What if you only do a single FB? You can always do it in Shred time: between 4 and 8 seconds.
    (Having OOC is like have 1.5 CPS banked)
    Since we know our burst rotation should begin with full energy and TF, we expect that we will generally want to use at least two Shreds before we make use of Tiger's Fury. This means if we enter our burst phase with 3,4,5 combo points we will expect to finish our second ferocious bite before the 15% damage buff from Tiger's Fury expires.

    Additionally, we also know that we want to use Rip/Rake at some time before we go into our burst phase such that both are on our target, while we are shredding/biting. Assuming we only use those 2 moves and pool energy from zero, it would actually take us roughly 15 seconds to pool energy to full. Assuming we leave our burst phase at 0 energy, with Rip soon to expire, we need to be able to build 5 combo points to spend on a rip, before the TF cooldown has 15 seconds left. Which means, we need to begin building combo points for rip at the 25 second mark.
    This tells us that our burst phase needs to end 5-6 seconds after TF in order for us to reliably be able to refresh Rip in order to be prepared for our next burst rotation... making the the 15% bonus damage from TF a good transition point.
    So, when do we need to apply Rip and Rake? Well, if we plan on ending our burst rotation 5-7 seconds after TF, we need Rip and Rake to last that long. Rip has a duration of 22 seconds, so it needs to be used after you have 16-17 seconds left on your TF cooldown. Rake on the other hand, has a shorter duration (15 seconds), so it needs to be used no sooner than the 10 second mark on the TF cooldown. Since Rake's duration fits evenly into your TF cooldown, it's easy to maintain at the same times.
    In the case of a double-bite, when we know that we will have to sacrifice Rip uptime, this leads naturally to the Rip downtime. Your Rip will fall off approximately 5-7 seconds after you use TF, you'll usually need to spend the energy you gain from the 25-15 second marks building combo points to re-apply Rip.
    In the case of a single-bite rotation, the upkeep on Rip becomes more difficult, because the 22 second duration doesn't fit cleanly into the 30 second cooldown of Tiger's Fury. Any time your TF cooldown falls within 3 seconds of the termination of your Rip, you'll have to clip your Rip to prevent it from falling off for significant time frames. This is really only troublesome if your Rip has been boosted by TF or a trinket use - in which case you'll get the 'a stronger effect is already in place' error.
    Things We've Learned
    All of the math above is all well and good, but it's more of a big picture of what's possible and isn't so helpful in terms of actually applying burst in PvP. Well, I am going to do my best to try to use this information to build some 'priority' systems or guidelines for us to use in PvP. Once I have those in hand, I'll run some simulations of those systems and we can compare and contrast how they perform.
    So let's quickly take some of the above math and distill it into things we've learned. We've learned that we shouldn't have too many problems attempting to do a Ferocious Bite every thirty seconds... and if we really push it, we can do two Ferocious Bites every thirty seconds. We know that we can make small adjustments to our rotations... like swapping mangles for shreds, or letting Rip drop for a little bit, and it *should* cascade into a surplus of energy we can use when we're ready to burst. Of all the little changes we could make to our rotation, the one that doesn't help us at all, is letting Rake drop off. We should also know, that we always would prefer to spend our energy on Ferocious Bites over Shreds. The same amount of energy does more damage when it's spent on a FB than spent on a Shred. The only exchange that really can hurt our DPS is letting Rip drop off... however, letting your rip slip off can pay dividends in terms of burst damage.
    The 'cascade of energy' that we will be using in our burst window is stored in excess Combo Points and OOC procs that you have on hand, when your Tiger's Fury becomes available. Keep in mind, our calculations were done assuming an average number of OOC procs - somewhere between 1-2 procs every 30 seconds. If you have gotten more than 2 OOC procs, you can probably expect your next burst rotation to be an extra strong one.
    You'll also be able to anticipate the strength of your upcoming burst, by the number of combo points you have when you begin pooling energy for it. If you have 0 or 1 combo point, you probably will want to just do a single Ferocious Bite or Maim. Anything more will not be 'bursty' and it will almost certainly cost you dps down the road. If you have 2-3 combo points, the Free combo points from your ravage can be used to upgrade a single FB into a double Ferocious bite, so you don't want to lose those combo points, if you can help it. If you have 4 or 5 combo points in the bank, you'll sail easily into a double Ferocious Bite finish. You can save your charge for a swap... or ignore the combo point generation in favor of getting an early Ravage crit on a high HP target.
    And in turn, you can anticipate whether or not you will have a surplus of combo points entering your next burst phase by the number of combo points that you had left-over in your previous phase.
    For example:
    If you reach the 25 second mark on your TF cooldown and you have 3 combo points left over, here are your options. If you build to 5 combo points and Ferocious Bite once more, you will now have 0 energy and 0 combo points with roughly 7-8 seconds to build 5 more combo points. Short of some lucky OOC procs, you will not have the energy to get to 5 combo points before you reach the 15 second mark. So you are either delaying your burst phase, entering it with less energy, or bursting without Rip on your target.
    Your other option is to build to five combo points, then pool your energy until the timer has less than 17 seconds remaining. Apply/refresh Rip at that time. After that, any time you reach full energy and have more than 3 seconds left on your TF cooldown, Mangle: This has a very good chance of getting you into your next burst phase with 4-5 combo points in the bank.
  • Section 2 :: Burst Rotation "Derivations" :: Burst Rotations
  • With the basic idea in hand, it's time to discuss some actual concrete plans for Bursting someone down. First... I know it's hard... we have to let go of the idea that we're ALWAYS actively trying to kill someone or do as much damage as possible. If you're going down the path of bursting someone down in a controlled window, commit to that plan. Embrace the fact that for most of the fight, doing as much damage as possible is not your highest priority. Your priority (as far as DPS goes) is simply to 'keep things rolling' so that when it's time, your pieces are in place and you can make your move.
    Single Ferocious Bite Rotations. (Can be done w/out Sacrificing Rip uptime/DPS)
  • Always Rake when your TF cooldown is between 10 and 3 seconds. Otherwise, Rake when needed.
  • Rip when the TF cooldown is between 17 and 3 seconds. Otherwise Rip when needed. If you needed to Rip before the 17 second mark, you will have to clip your next Rip before TF is available.
  • Between the 25-15 second mark, spend energy freely to generate combo points. I suggest Mangle.
  • Under the 15 second mark, only spend energy to prevent capping, to refresh Rake, or to refresh Rip.
  • At the 3 second mark, get distance for a cat-charge. Charge when you reach full energy.
  • If the target is above 80% health, Ravage immediately. (otherwise, Ravage after TF)
  • Build to 5 CPS with Shred. When you have 5 CPS + 50 or more energy, Ferocious Bite. If you ever fall below 40 energy, Tiger's Fury.
  • As long as you have the TF buff and energy remaining, Shred.
  • The 'burst phase' should end about when your TF buff wears off.
  • Double Ferocious Bite Rotations (or single bite + maim... exchanges DPS for Burst)
  • Always Rake when your TF cooldown is between 10 and 3 seconds. Otherwise, Rake when needed.
  • Rip when the TF cooldown is between 17 and 3 seconds.
  • Between the 25-15 second mark, spend energy freely to generate combo points. I suggest Mangle.
  • Under the 15 second mark, only spend energy to prevent capping, to refresh Rake, or to refresh Rip.
  • IF you have 0-2 combo points saved:
    Do a single ferocious bite rotation, with an optional cat charge.

    IF you have 3 saved:
    Option 1: do a single ferocious bite rotation, with no cat charge.
    Option 2: do a double ferocious bite rotation, using Ravage immediately

    If you have 4 or 5 CPS saved:
    Option 1: do a double FB combo, with no cat charge.
    Option 2: do a double FB combo, using Ravage between bites.

    Doing 2 Ferocious Bites will always cause your Rip to fall off. If you Rip just before TF, you will not have enough combo points saved to quickly bite twice. If you Rip well before TF, you will not have time to build CPS after your bites, before Rip falls off.

  • Section 3 :: Mew Rotation Results
  • I took the above 'rules' and wrote some Mew Scripts to simulate the DPS curves we would get if we followed their guidelines.
    First to get a baseline of comparison, I did a 'standard' mew simulation, using my character (with Berserk turned off.)
    The simulations run for a 900 seconds, 10,000 times.

    Standard Mew Run:
    15.5k DPS, 2 ferocious bites, 99% Rake uptime, 94% Rip uptime, 3800 white DPS.
    If we neglect white damage, that's 11.7k DPS.

    Single Ferocious Bite Rotations:
    13.7k DPS, 25 ferocious Bites, 93% Rake uptime, 93% Rip uptime, 1800 white DPS.
    neglecting the white damage, that's 11.9k DPS.

    Double Ferocious Bite Rotations:
    12.8k DPS, 41 ferocious bites, 96% Rake uptime, 52% Rip uptime, 1800 white DPS.
    neglecting the white damage, that's 11.0k DPS
    So aside from the fact that the default MEW script is clearly more efficient at keeping up Rake, we can see that some of our estimates from our energy and timing analysis have been borne out. It's important to consider the dps with the white-damage taken out, because limited uptime in PvP really reduces the contribution of white-damage to your DPS. We also see that increasing the number ferocious bites comes at the cost of Rip uptime... which loses us roughly 1k DPS. Unfortunately, we're not getting quite so much bang for our buck in terms of Rip uptime -> extra ferocious bite conversions as we anticipated. I'm not positive, but I think it's because I neglected to account for combo points lost on crits as well as estimating the crit rate at 50%. Anyway, those small details are the least of a PvP'ers concerns. The big picture can be seen in the following DPS vs Time graph:

    The graph is roughly a 200 second sample from the middle of the simulation runs. The blue line is the basic Mew rotation. It has the highest sustained DPS... however, as expected, the green graph has the highest burst dps. The tallest of those green spikes represents roughly 220,000 damage done over a 6 second interval. The tallest red and blue spikes represents roughly 150,000 damage over six seconds.

  • Section 4 :: Applying These Ideas in Reality
  • I think it's pretty clear at this point, that the things I've outlined above are your burst options. I don't think it should need to be said... but I'm doing it anyway: I don't expect anyone to be able to maintain any sort of rotation such as this in a PvP environment. This is something for you to strive to do. It's an ideal on the horizon. In reality, you won't be able to burst on 30 second intervals. Much of your time in PvP is spent doing things *other* than doing damage. It's my hope that by breaking down your burst options and exactly what needs to happen for you to pull them off, players can practice and become comfortable at recognizing the opportunities they are given in a match.
    The Opener
    The first real problem with our burst... is our opener. You'll notice that what we really want is to apply rake and rip BEFORE we go into TF rotations. However, in our typical opener, we tend to apply Rip and Rake after TF. Furthermore, we often apply them with trinkets making them difficult to clip. I can't emphasize enough how big of a pain in the butt this is. If you have any hope of doing a double ferocious bite on your 2nd tiger's fury, you better throw that hope away, plan on berserking off the opener, or delay your 2nd TF by about 15 seconds.

    A "Rotation" is not realistic - We Often Start From Scratch
    I've found that when playing healer/feral against healer/xx, I've had no problems keeping a 2-bite-burst rotation similar to what I've outlined. However, outside of that, keeping any sort of rotation really isn't realistic. What's probably of more general use is starting from scratch. So a question we might want to ask ourselves is, "How long will it take me to burst, if I start fresh."
    If we want to do a single bite:
  • We need 5 CPS to apply Rip. That takes roughly 3 seconds.
  • Then we have to wait 10s to get full energy .
  • Then we want to Rake.
  • Wait for full energy again (3 seconds).
  • Then Burst!
  • It will take 16 seconds to get a 'full' 1-bite-burst rotation on your target ready.

    If we want to do a double bite rotation:
    It's exactly the same, except we need to enter our burst with at least 3 combo points stored, that will take an additional 3 - 6 seconds... so, 19-22 seconds total.

    Burn Tactics for Feral/Disc
    I mentioned that healer/dps VS healer/dps is about the only time where in practice I was able to keep anything remotely resembling a 'rotation.' Although technically not a burst tactic, it's very easy to see how we can apply this knowledge to burn tactics in 2s. While you're applying a burn tactic, you'll actually just be doing the single-FB rotation. If your timing is good, your team should be able to maintain 100% uptime on your Rip, while still stunning the healer every 30 seconds for mana burns.

    Swap Tactics
    Although it hasn't been said, we've actually really narrowed down the possible ways of doing a swap, from the above information.
    Hard Swaps for Kills:
    If you're entering your burst phase with 4-5 combo points, you can do a double finisher on your current target, then cat-charge swap to the new target. There's not too much 'oomph' to the swap, but you will get the free ravage, which will probably crit.
    Alternatively, if you have 3 or more combo points, you can forgo your 2nd finisher on your current target and instead cat-charge and apply a quick FB to your swap target. This would be a pretty good hard-swap for a kill.
    Regular Swaps:
    The most basic style that everyone uses is just to swap and burn useless CPS on Savage Roar. Well, there's another way. No matter how many combo points you have going into your burst phase, you can save your Cat-Charge and assess whether or not you want to swap targets after you drop your first Ferocious Bite. If the enemy team responds with a defensive cooldown, you should be able to easily swap targets and get full bleeds running on the swap target within a few seconds.

    Minimizing Problems from Getting CC'd
    Since most of the time in PvP, you will be building to burst from scratch, we have a pretty clear idea when it really sucks to be CC'd. In general, applying Rake will be the last thing you want to do, because it has the shortest duration. Getting CC'd after you've applied Rake is the thing most likely to ruin your burst attempt. So you'll want to minimize that problem by:
  • Get full energy before applying Rip -- you're on the clock once you've Rip'd.
  • Immediately mangle to 2CPS. -- if you get CC'd you want it to happen just after this.
  • Rake when you get full energy. -- getting CC'd just after this, will ruin your rotation
  • Begin burst at full energy
  • So, the time you really want your team to CC the enemy peeler is just before you apply Rake.

    Pooling Energy
    I mentioned that there was a period where I suggested that you spend energy as you get it. This would be right after you've gone through a burst phase and are doing upkeep on your bleeds/building combo points for later on. The advantages of doing this, is that if you get CC'd you're less likely to waste energy by over-pooling it. You also shouldn't need energy in order to kick any healing, since you're now in the part of the fight where your burst is over.
    Also, you probably want to spend your energy immediately after applying Rip, when building to burst from scratch.
    However, there are real strategic reasons to keep your energy pooled, besides simply being ready to burst. Sure kicking a heal at that point in the fight may not really help you win at all, but you may need to kick CC's or damaging abilities for defensive reasons. If you keep yourself tapped on energy, you won't be able to do that. Also, positioning is a HUGE component to success in PvP. Most Ferals pool energy, in part to take advantage of positioning errors that the other team makes. Swapping to someone to attempt to burst them down, because they're out of position is an important tactic to be able to apply. If you do that, you'd simply not use Rip. Swap, Rake, pool to full - apply burst. That takes 5 seconds for a single bite or 9 seconds for a double bite.
    A Note About Armor
    Finally, under no armor circumstance are bleeds ever better burst than ferocious bite. Likewise, it is never better to shred spam, than spend that same amount of energy in a 5-point FB. Even if you're facing Paladins or Protection specs, the most burst damage you can do involves the exact same rotations. You may however, get more mileage replacing a bite with a maim.
    The Most Damage you can do in X seconds
    What probably hasn't been all that clear, is that the single-bite or double-bite rotations do almost the same damage regardless of whether it's over 5,8, or 10 seconds. It's because it's basically the same amount of energy being spent on damage, by the same moves. So what you really want to look at is the time to apply damage.
    So, if you play on a team which regularly CC's healers for 10 seconds or more, you can probably do a double-finisher with only 1 combo point saved. In which case, you may want to get into the habit of dropping a Rip, pooling, Rake, pooling and going directly into FBs. The late Rip will last the duration of your CC and with 1-2 Combo points going into your burst phase, you'll be able to get two bites out in the next 10 seconds.

    Timings While Using Berserk
    I've steered clear of talking about Berserk though most of this so far, but at the request of Ellesime, I've calculated how some of the timings change, assuming you're using berserk.
    2x Ferocious Bites, using Ravage, Berserking:

    CPS Saved -------> average time to do two 37.5 energy FBs
    0 CPS ----> 8.4 seconds
    1 CPS ----> 7.8 seconds
    2 CPS ----> 7.1 seconds
    3 CPS ----> 6.5 seconds
    4 CPS ----> 5.8 seconds
    5 CPS ----> 5.2 seconds

    3x Ferocious Bites, using Ravage, Berserking:

    CPS Saved -------> average time to do two 37.5 energy FBs
    0 CPS ----> 12.8 seconds
    1 CPS ----> 12.1 seconds
    2 CPS ----> 11.5 seconds
    3 CPS ----> 10.8 seconds
    4 CPS ----> 10.1 seconds
    5 CPS ----> 9.4 seconds

    Time to Burst from Scratch:
    It takes 6 seconds to get 5 CPS and get full energy.
    It takes 3 seconds to apply Rip and Rake and get Full Energy.
    9 Seconds of setup + 1 second for every combo point + Tables above.
    So, if I want to do a 3xFB finisher from scratch, it would take about:
    9 + 5 + 9.4 = 23.4 seconds from start to finish.
  • Section 5 :: Haste as a Burst Stat?
  • For the most part, I think the effects of Mastery and Crit in PvP are well understood, this post is already getting far longer than I had originally planned, and I'm getting pretty tired of working on it, so I will not go into detail about those Stats... (I'm even shooting for brevity on Haste.)
    However, how Haste might affect performance in PvP is really not understood well. We avoid it because haste contributes to PvE-DPS 'mostly' through white damage. And since uptime is limited in PvP, so too is white damage.
    That's not the only affect of Haste. Haste also increases (by a small amount) energy regeneration. This could have two very key effects on Burst Damage. The first is that you could theoretically get enough haste such that, in any given time interval, you generate enough energy for 1 extra move. So for example, you could shred x4 on a full energy bar instead of x3. Converting from one stat (like mastery) into haste, would cost you damage until a break-point was reached and you gained an extra attack. At which point, presumably, you'd do more damage in that same time window.
    So, once again, since I'm getting tired of this :P ... without further ado, the results:
    Energy Regeneration Rate = 10(1 + (Haste/12805))
    Haste Break needed to get 1 extra Shred, starting from full energy in this time frame:
    w/out Berserk
    In 1 Energy Bar: ~19250 haste
    In 6 Seconds: 8536 haste
    In 8 Seconds: 3201 haste
    In 11 Seconds: 3492 haste

    w/ Berserk:
    In 1 Energy Bar: 1281 haste
    In 6 Seconds: Always get 6
    In 8 Seconds: Always get 8
    In 11 Seconds: 1170 haste

    Okay, so no... haste isn't going to give us a useful breakpoint. All of the non-berserk breakpoints are basically unobtainable. While the only obtainable berserk breakpoints take place on time-scales of about 10-11 seconds. Not exactly very bursty. Although, if for some reason, shredding for 10 seconds continuously while in berserk is a part of your strategy, you *may* see a slight improvement from hitting the 1170 haste breakpoint. I say *may* because I estimate for my character for that to be roughly a 2% increase in damage. (Keep in mind, you are losing mastery or crit in exchange for this haste.)

    The second way that I anticipate haste may affect burst damage is really because our burst focuses on using Ferocious Bite. FB allows us to bypass the 'break-point' phenomenon with haste, because FB has a continuous energy to damage conversion rate. If you gain 5 extra energy because of your haste, a Ferocious bite which would normally spend 25 energy will spend 30 energy and do more damage.
    For this part, I threw character sheets with different amounts of Haste into Mew and looked at the results.
    Haste really has a three-fold effect on burst rotations involving Ferocious Bite. The first is what we were anticipating going into this test. If you're in the habit of using FB as soon as possible, excess energy from haste will go into your ferocious bite.
    Here, I ran the double bite script, only I relaxed the condition requiring that the bite only occur with more than 50 energy. So that the simulation bites as soon as possible.

    I sampled data points every 200 haste, up to 2000... then I jumped to 3000 just to see if the behavior continued. The general trend is that about 100 haste gets you about 100 damage added to each Bite. Which is actually pretty good. And no, the pattern doesn't become a 'straight line'... I just selected 'connect the dots' in excel. So you're seeing the connected dots between 2000 and 3000.
    The second effect Haste has, is that you're able bite more frequently. Or rather, you reach the condition for a 2nd bite within your burst window more frequently. As the number of bites in a 900 second fight increases with haste as well:

    If I were to estimate a trend for this, I would estimate that you get roughly 5% more bites for every 1000 haste you get.
    It's worth noting that in the above case, I was allowing the simulation to bite as soon as 5 combo points were reached. Enforcing the 50-energy bite requirement shows different results. The maximum damage, while increasing, increases at a much slower pace (30 damage / 100 haste)... and it is not made up for by an increase in the number of bites at all.
    And in both simulations for the data above, the script attempts to aggressively burst every 30 seconds. To see the final effect of Haste on burst, I utilized an older/less efficient script, which was more flexible in terms of its burst timing. In that case, we can see that the time between burst windows decreases as we increase the amount of haste we have:
    For those who are curious, both the damage and number of bites in the final script increased with haste - At approximately 50 dmg / 100 haste and 3% bites per 1000 haste.
    It really seems like haste does its best for players who aren't careful about their timings, it's a little grease for the ole claws. Which honestly, is probably needed by everyone in a real PvP environment. However, let's quickly look at how it compares to mastery in terms of contribution to overall burst damage.
    If we have a 2-bite burst window, it lasts roughly 6 seconds. Adding 2000 haste will increase your average bite damage by 2000. So you will do on average 4000 more damage in that 6 second window by increasing your haste by 2000.
    Reducing your mastery by 2000 will cause your Rip ticks to be reduced by about 2000 damage, as well. With 3 ticks in a 6 second window, you would be reducing your damage over that time by roughly 6000, just from Rip! In addition to that, you'd get 2 ticks of rake, which itself lost about 1500 damage. So you can expect to lose around 9000 bleed damage in a 6 second window and only get 4000 bite damage in return.
    If you look at max-potential damage from the bites (say both bites crit), 2000 haste roughly gives 3000 extra damage per bite. So 6000 extra damage over-all. Still less than the average return on bleeds.

  • Section 6 :: A Small Word about Crit
  • First, I didn't do a more thorough comparison of Crit/Mastery, because the fits I anticipate that I'd need to do would take me a few more hours of work. Frankly, I don't care enough to do that. With that out of the way:
    You might expect that Crit would also have a strong effect on the number of double-bite opportunities you have over the course of a fight. It does not. Changing crit rating from 0 to 3000 only increased the number of bites from 40 to 42. So haste is more helpful in terms of letting you get double-bite scenarios in simulation. However, Crit has an effect on the average bite damage. Since you're more likely to crit.
    Every 100 points of crit rating increases your average bite damage by 130. So 30% better than haste. HOWEVER, haste actually increases max bite damage, while crit only increases likelihood of max damage. So if you're purely looking at the burst potential of FB, haste is a better stat. Crit also affect bleed damage, so it's better than haste for increasing your average damage.
    The thing that makes crit a poor BURST damage stat, is that burst damage via crits is a result of compounding many stacking damages. In any given burst window, you have roughly 4 dot ticks, 6 white damage hits, 3 shreds, 1 ravage, and 2 ferocious bites. The odds of critting ALL OF THOSE THINGS is really small - roughly 1 in 100,000. The thing you really want to do, in order to make your damage more bursty, is move damage away from multi-hitting abilities, like dots toward big hitting abilities like: FB.
    In theory, Haste should be our best burst damage stat... It's really our only stat that actually increases the burstiness of our delivery - by moving damage away from bleeds to Bites. However, it sucks. It increases burst damage by LESS than our other stats increase average damage, even over relatively short time windows.
    The only case where I can see haste performing superior to Crit or Mastery, is cold burst swaps to fresh targets. If you're a player that often capitalizes on players out of position - going into burst rotations w/out Rip on your target, then you'll get the most bang for your buck from Haste... by a large margin.
    Finally, the last thing you'd expect crit to give you is a greater chance of getting critical strikes on the trifecta of FB, FB, Ravage.
    From my mew statistics, my chance of critting on FB is roughly 65% and my chance of critting on Ravage is 40%
    Odds all 3 of those crit: 17%
    Odds two of those crit: 43%
    Odds one of those crit: 32%
    Odds none of those crit: 7%
    You can see how these chances vary as we change our critical strike chances in the following chart:

    You can see from the chart, that increasing your critical strike rating from 35% to 45% will increase your odds of doing 'good burst' from about 52% to 65%, seeing an 8% gain in your odds of a triple crit and a 5% gain in your odds of a double crit.
    In fact, your total odds of 'getting good burst' increases pretty linearly with your critical strike chance (roughly 1.4 per crit chance). Although, the rate does decline slightly once you have more than 50% chance to crit (roughly 1.1 per crit chance.)
    To put that in easier to understand terms: In your average 5 minute arena match, you'll have roughly 8 burst attempts. Of those 8, roughly 4 of them will probably be double bite opportunities. If you have a 35% crit chance, you'd expect 2 of those to hit hard. If you increase your crit chance by 10%, every other match, you will have 3 hit hard instead of 2.
  • Section 7 :: Conclusion
  • As far as rotation goes, the big picture that we get from all of this is that we really want to push ourselves into double-bite scenarios in PvP. And in order to do this reliably, we need to bank 3+ combo points before we enter our burst rotation. We also have learned that the last thing we want to do before we initiate our burst is apply Rake.
    This means that the best/easiest recommendation I have for most of us will be to build 5 CPS, Pool Energy. Rip, Manglex2, Pool Energy, Rake, Pool Energy, Begin Burst. Although, I think there's plenty of information above to tailor rotations to fit your specific team.

    A: OKAY!

    Burst if you're REALLY IN A HURRY {like a hard swap}:
    1) Mangle (if not needed, skip)
    2) Pool to 100 Energy
    3) Rake
    4) Get Distance
    5) Cat Charge
    6) Ravage
    7) Shred to 5 CPS
    8) Wait for 50+ Energy
    9) Ferocious Bite
    {Tiger's Fury the first time you fall below 40 energy}
    Total Time: about 5 seconds
    Max Damage ~ 100k

    More Burst, Less Hurry {still hard-swapping}:
    1) Mangle
    2) Pool to full
    3) Rake
    4) Get Distance
    5) Charge
    6) if HP > 80% Ravage
    7) Shred to 5 CPS
    8) Wait for 50+ energy
    9 ) FB
    10) Shred to 5 CPS, if you need to TF. Ravage after TF, if you haven't used it yet.
    11) Wait for 50+ energy
    12) FB
    Total Time: about 9 seconds
    Max Damage ~ 200k

    Starting from Scratch {Maybe you were CC'd and bleeds dropped}:
    1) Mangle to 5 CPS
    2) Rip
    3) Wait for 100 energy
    4) Rake
    5) Get Distance
    7) Cat charge
    9) Ravage
    10) Shred to 5 CPS, TF when needed
    11) Wait for 50+ energy
    12) Ferocious Bite
    Total Time: about 16 seconds
    Burst Damage Roughly ~ 150k

    Starting from Scratch MORE BURST:
    1) Mangle to 5 CPS
    2) Wait for 100 energy
    3) Rip
    4) Mangle to 2 CPS
    5) Wait for 100 energy
    6) Rake
    7) Get Distance
    8) Cat Charge
    9) Ravage
    10) Shred to 5 CPS
    11) Wait for 50+ Energy
    12) FB
    13) Shred to 5 CPS, if needed TF
    14) Wait for 50+ Energy
    15) FB
    Total Time: about 21 seconds
    Burst Damage Roughly ~ 240k

    Bursting as a part of a regular Rotation (Priority System):
    Always Rake when your TF cooldown is between 10 and 3 seconds. Otherwise, Rake when needed.
    Rip when the TF cooldown is between 17 and 3 seconds.
    Between the 25-15 second mark, spend energy freely to generate combo points. I suggest Mangle.
    Under the 15 second mark, only spend energy to prevent capping, to refresh Rake, or to refresh Rip.

    IF you have 0-2 combo points saved:
    Do a single ferocious bite rotation, with an optional cat charge.

    IF you have 3 saved:
    Option 1: do a single ferocious bite rotation, with no cat charge.
    Option 2: do a double ferocious bite rotation, using Ravage immediately

    If you have 4 or 5 CPS saved:
    Option 1: do a double FB combo, with no cat charge.
    Option 2: do a double FB combo, using Ravage between bites.

    Any time you use two finisher moves, either FBx2 of Maim + FB, your Rip WILL DROP OFF. That's to be expected. It's the price of extra burst.
    finished posting,
    enjoy ... I'm going to go get waffles. No Lie.
    Great read. You deserve some serious waffles!
    Congrats on completing a great guide, I'm glad its numerically based, but just as a point of discussion (absolutely not hating on you)

    The thing that makes crit a poor BURST damage stat, is that burst damage via crits is a result of compounding many stacking damages. In any given burst window, you have roughly 4 dot ticks, 6 white damage hits, 3 shreds, 1 ravage, and 2 ferocious bites. The odds of critting ALL OF THOSE THINGS is really small - roughly 1 in 100,000. The thing you really want to do, in order to make your damage more bursty, is move damage away from multi-hitting abilities, like dots toward big hitting abilities like: FB.

    I feel this is an unfair interpretation of the numbers, although it is true that the optimal 'burst' would consist of a single hit in that period, as our numbers stand now FB is not a hugely significant part of our damage in that 6 second period. Stacking haste rating with pvp gear does not adequately shift this balance, even if haste is more effective on FB it doesnt compensate for the fact that it has no effect on the other 80% of our damage in that period. (80% was just a guess, feel free to correct).
    Sweet mother of god this is the most concise class guide ive ever seen.

    Excellent work.
    01/30/2012 05:13 AMPosted by Ventari
    I feel this is an unfair interpretation of the numbers, although it is true that the optimal 'burst' would consist of a single hit in that period, as our numbers stand now FB is not a hugely significant part of our damage in that 6 second period. Stacking haste rating with pvp gear does not adequately shift this balance, even if haste is more effective on FB it doesnt compensate for the fact that it has no effect on the other 80% of our damage in that period. (80% was just a guess, feel free to correct).

    FB + FB + Ravage is roughly 50% of our damage in a 6 second burst window. However, I pretty much agreed with you that stacking haste doesn't do enough. I think I said that just a few sentences later:

    ... [haste is] really our only stat that actually increases the burstiness of our delivery - by moving damage away from bleeds to Bites. However, it sucks. It increases burst damage by LESS than our other stats increase average damage, even over relatively short time windows.

    Although haste has the right behavior, it doesn't have a big enough effect.

    here we go again! will try to work through this version while i'm doing server work today.

    much love clay.

    Join the Conversation

    Return to Forum