Optimal Stats For All 24 Disciplines, KOTFE Edition

[mmerry]

Or of course, the perenial whipping boy, Hatred. Why on earth would a sustained melee spec with poor AoE, no burst, significant setup penalties, and literally no survivability be parsing closer to the bottom of the heap than the top?

 

this x10000.

[flem]

I would also point back to Fury, which is still enormously behind any reasonable DPS bar, and also Hatred, which is far enough behind on DPS that its glaring survivability issues aren't even remotely ignorable. No one is running either of those two specs in any content.

I assume you mean PvE? Fury is quite a legitimate PvP burst spec.

[Salakast]

In general, I agree with the principle of the thing. Marksmanship is ranged, extremely high burst (with the advent of super crit), has LITERALLY no target swapping penalties, and has good AoE and survivability. It should be the lowest parsing spec in the game. However, it doesn't take much effort looking over Bant's list to see where balance is screwy, even in this revised world.

 

.

 

Not in PvP it doesn't. It's ability to churn out high dps in a snipers limited lifespan was all it had going for it. Now it has nothing, there's no way to burst through your chosen target fast enough for a MM to bring them down before being mushed.

 

I have a couple of genuine questions, I don't Raid, Parse, wouldn't know where to start theorycrafting or owt so I don't know what counts:

 

Does having to add the corrosive dart debuff count as a target swap penalty?

 

With the new highly mobile 4.0 meta, doesn't that discount the ranged 'tax' on dps somewhat?

[TACeMossie]

Pyro (a melee sustained spec with an unavoidable 3 second channel in rotation and the worst target swapping in the game)

 

While I agree with the rest of your statement KBN, I can't help but heavily disagree with you on this one. Why?

 

Lets look at what Pyro loses by target swapping:

 

1. Scorch, should the previous target still be alive. If it died, ignore this. Reapplied from 10m.

2. Incendiary Missile, should the previous target be out of Flamethrower range. If it isn't, ignore this. Can be reapplied from 30m.

3. Plasma cell, which is usually reapplied automatically within 3 seconds anyway.

 

Now lets look at AP, as an example.

AP Loses

 

1. Retractable Blade, which requires a 4m range to reapply

2. Blood Tracker (+3% damage buff), which requires Retractable Blade to work

3. Rail Shot heat vent, which requires Retractable Blade to work

 

The loss of those 3 are significantly more in terms of damage loss when compared to what Pyro loses, and whats more is Pyro has better built-in AoE to go with it.

 

So wherever you got the idea that Pyro has the worst target swapping in the game, you're frankly wrong. I'd have put quite a few specs behind it, such as Engineering (because 4 DoTs and no dotspread), Lethality (because 2 DoTs and useless dotspread) IO (See AP's problems, though at least it has 30m range to reapply its DoTs. Then again, IO has more DoTs to deal with).

[DarthZaul]

I know he did a complicated mathsy explanation somewhere, but can someone TL;DR why madness moved all the way up from terrible to #6 without being buffed?

[Goblin_Lackey]

I know he did a complicated mathsy explanation somewhere, but can someone TL;DR why madness moved all the way up from terrible to #6 without being buffed?

 

Small changes in rotation can have large effects on DPS.

 

The short answer is that I found a more optimal rotation based on the principle of always using death field, demolish, affliction and creeping terror exactly on cooldown while simultaneously replacing most of the lighting strikes with force lightning (which even while clipped, Force lighting does more damage; but you can't clip force lightning with itself).

 

The new rotation can be found in the class notes section (post #5)

[Joefjr]

I know he did a complicated mathsy explanation somewhere, but can someone TL;DR why madness moved all the way up from terrible to #6 without being buffed?

 

Cause some above it are now below it, they got nerfed.

[KeyboardNinja]

So wherever you got the idea that Pyro has the worst target swapping in the game, you're frankly wrong. I'd have put quite a few specs behind it, such as Engineering (because 4 DoTs and no dotspread), Lethality (because 2 DoTs and useless dotspread) IO (See AP's problems, though at least it has 30m range to reapply its DoTs. Then again, IO has more DoTs to deal with).

 

I think we're defining "target swap penalty" in slightly different ways. AP's target swap penalty, in my view, is one GCD: Retractable Blade. There is a penalty there, as it must be reapplied out of sequence (imposing a heat and time cost) and it is a 4 meter ability, but otherwise it's not that bad. If you just need burst NAO, you can skip it and things still mostly work, it's just that you'll have somewhat less damage and no heat management (as I said, burst NAO).

 

If Pyro is sitting on a single target, and another target appears that it must instantly swap to (the condition I'm imposing on AP), I suppose you can mostly keep doing your rotation as long as someone else is DoTing, so perhaps the target swapping is less onerous than I believe. In a worst case though, you have to get Incendiary Round up at the very least, and Pyro's rotation is tight enough that this reshuffling is very painful.

 

IO has very limited target swapping penalties. You just keep doing the rotation. If there is no burns on the target (which can come from any spec, btw, so Watchman and Vigilance are your best friend), you do lose damage, but Plasma Cell is going to proc sooner or later, so it's really only a problem if it's right before Mag Bolt. So let's call that the same penalty as Pyro.

 

Engineering certainly has awful target swapping, though fortunately again it can mostly just keep doing its rotation and things will function normally. Lethality is pretty bad, though the flexibility of the rotation (the amount of stuff you can scoot around and/or recast is insane) makes it better than it seems on the surface. Virulence's target swapping is probably the worst in the game.

 

So I guess, now that I sit down and think about it, Pyro's really isn't the worst. But the energy constraints, the rigidity of the cooldowns, and the unforgiving nature of the debuff requirement on Rail Shot makes it one of the worse ones.

Edited December 10, 2015 by KeyboardNinja

[meisterjedi]

Honestly the penalties to target swaps for Pyro or AP depend so much on the exact encounter, that it's hard to say one is worse than the other universaly.

 

If you look at a fight like Revanite Commanders the targets die so fast and swaps are so frequent, that you can't use retractable blade at all on the adds and therefore miss out on the heat vent every 6 seconds for the majority of the fight (although it's the irrelevant part of it). Pyro on the other hand can continue it's rotation just normal (as much as that's even possible on that fight) and with great positioning (from you and the tanks) can even keep the dots alive through target swaps.

 

If you look at the droid Phase for Brontes, Pyro suddenly can't use scorch or incendary missile effectivly and you would need to flame burst before you can use Railshot when you might still be > 12m away (as other dot classes often don't put there dot's up as well). AP can handle the heat issues by using both heat cooldowns in this rather short Phase and be more effective than Pyro that might have to use suboptimal Abilities like Jet charge or Incendary missile frequently to keep it's dmg rolling. This is one of the few situations where a DPS PT really benefits from Jet charge, even if it's only Pyro. YAY

 

Looking at swapping to the kephess clones both suffer minimal losses, but I would put AP higher, as you can maintain the dot on Brontes easyly with explosive dart when leaving and a Railshot/explosive dart when Kephess jumps and would be immune anyway.

Pyro can keep scorch up on Brontes, but has a harder time doing dmg while closing in on Kephess and therefore often has to resort to jet charge to not delay Flamethrower/Flaming fist by more than 1 GCD, which is much more important than delaying Rocket punch a bit. Still both suffer only minimal losses and AP can even keep up with most ranged speccs here.

Edited December 10, 2015 by meisterjedi

[Paladrima]

Hello Bant,

 

first of all thanks for all the work you put into this, it is indeed very helpful.

 

Sadly, good news rarely come alone, so I wanted to take a specific look at the new rotation for Madness Sorcs / Balance Sages that you have come up with during the last update.

A few posts earlier you stated the following regarding a question towards Madness.

 

Small changes in rotation can have large effects on DPS.

 

The short answer is that I found a more optimal rotation based on the principle of always using death field, demolish, affliction and creeping terror exactly on cooldown while simultaneously replacing most of the lighting strikes with force lightning (which even while clipped, Force lighting does more damage; but you can't clip force lightning with itself).

 

The new rotation can be found in the class notes section (post #5)

 

So since that's a pretty strong statement, I figured I'll give it a shot and a good test, to see how it works out for us non-machines on an actual dummy. During those testings however I noticed that something didn't quite seem right, even considering the fact of timeloss due to imperfect clipping of Force Lightning.

The following is a quote from your Individual Class notes with your rotation. As you will see, I have underlined certain parts of it:

 

 

 

Sorcerer-Madness | Sage-Balance

Rotation:
[u]Demolish -> Affliction -> Death Field -> Clipped Force Lightning[/u] -> Force Leech -> Creeping Terror -> Full Force Lightning -> Full Force Lightning -> Full Force Lightning ->
Demolish -> Clipped Force Lightning -> Death Field -> Affliction -> Clipped Force Lightning -> Lightning Strike -> Clipped Force Lightning -> Creeping Terror -> Clipped Force Lightning -> Force Leech ->
[u]Demolish -> Clipped Force Lightning -> Death Field -> Clipped Force Lightning -> Lightning Strike -> Affliction -> Clipped Force Lightning[/u] -> Force Leech -> Clipped Force Lightning -> Creeping Terror ->
Demolish -> Clipped Force Lightning -> Death Field -> Full Force Lightning -> Full Force Lightning -> Full Force Lightning -> Affliction -> Force Leech -> Clipped Force Lightning ->
Demolish -> Creeping Terror -> Death Field -> Full Force Lightning -> Full Force Lightning -> Full Force Lightning -> Force Leech -> [u]Clipped Force Lightning -> Affliction ->
Demolish -> Clipped Force Lightning -> Death Field -> Creeping Terror -> Full Force Lightning[/u] -> Force Leech -> [u]Full Force Lightning -> Full Force Lightning -> Clipped Force Lightning -> [/u]
Clipped Force Lightning = clip at 1.5s

References: 

 

 

You will notice that before and after each underlined section a Force Leech cast can be found, which is the culprit I found while testing these.

The underlined sections consist of a number of GCDs, which will be somewhere between 7 and 8, specifically lower than 8.

Ignoring Alacrity we see Force Leech with a cooldown of 12 seconds (8 GCDs). However, since Force Leech is not an instant-cast, this cooldown will only start at the end of the GCD in which Force Leech is used. Thus, there need to be at least 8 GCDs between 2 consecutive uses of Force Leech.

As stated, the underlined sections fall below this 8 GCD-threshold, which means there would be downtime in the rotation.

 

I would assume that your model accidentally considers Force Leech an instant-cast (stemming from Assassins Leeching Strike maybe?), and thus results in this error, since the rotation would work fine if that was the case.

 

Figured I'd let you know about this problem, and hope that you're able to fix the mentioned problem without too much trouble

[Goblin_Lackey]

Hello Bant,

 

first of all thanks for all the work you put into this, it is indeed very helpful.

 

Sadly, good news rarely come alone, so I wanted to take a specific look at the new rotation for Madness Sorcs / Balance Sages that you have come up with during the last update.

A few posts earlier you stated the following regarding a question towards Madness.

 

 

 

So since that's a pretty strong statement, I figured I'll give it a shot and a good test, to see how it works out for us non-machines on an actual dummy. During those testings however I noticed that something didn't quite seem right, even considering the fact of timeloss due to imperfect clipping of Force Lightning.

The following is a quote from your Individual Class notes with your rotation. As you will see, I have underlined certain parts of it:

 

 

 

You will notice that before and after each underlined section a Force Leech cast can be found, which is the culprit I found while testing these.

The underlined sections consist of a number of GCDs, which will be somewhere between 7 and 8, specifically lower than 8.

Ignoring Alacrity we see Force Leech with a cooldown of 12 seconds (8 GCDs). However, since Force Leech is not an instant-cast, this cooldown will only start at the end of the GCD in which Force Leech is used. Thus, there need to be at least 8 GCDs between 2 consecutive uses of Force Leech.

As stated, the underlined sections fall below this 8 GCD-threshold, which means there would be downtime in the rotation.

 

I would assume that your model accidentally considers Force Leech an instant-cast (stemming from Assassins Leeching Strike maybe?), and thus results in this error, since the rotation would work fine if that was the case.

 

Figured I'd let you know about this problem, and hope that you're able to fix the mentioned problem without too much trouble

 

Hmm, you are quite right on this; I did accidentally think of Force Leech as an instant cast (too much time on Assassins). This is the reason why I try to post everything I can so that other people can find my errors and oversights.

 

After thinking about it further (and playing around with writing it out in Excel) I think the following changed rotation would work. It also keeps the same activation rate for abilities as my previous (impossible) rotation so there is no DPS change.

There is now always 15s of time between Force Leech usages so this should keep the previous issues lower.

 

Rotation:
Death Field -> Affliction -> Demolish -> Creeping Terror -> Force Leech -> Full Force Lighting -> Full Force Lighting -> Full Force Lighting -> Clipped Force Lightning -> 
Death Field -> Clipped Force Lightning -> Demolish -> Affliction -> Force Leech -> Creeping Terror -> Full Force Lighting -> Full Force Lighting -> Full Force Lighting -> 
Death Field -> Clipped Force Lightning -> Demolish -> Clipped Force Lightning -> Force Leech -> Affliction -> Clipped Force Lightning -> Creeping Terror -> Lighting Strike -> Clipped Force Lightning -> 
Death Field -> Clipped Force Lightning -> Demolish -> Clipped Force Lightning -> Force Leech -> Clipped Force Lightning -> Lighting Strike -> Affliction -> Clipped Force Lightning -> Creeping Terror -> 
Death Field -> Clipped Force Lightning -> Demolish -> Clipped Force Lightning -> Force Leech -> Full Force Lighting -> Full Force Lighting -> Full Force Lighting -> Affliction -> 
Death Field -> Creeping Terror -> Demolish -> Clipped Force Lightning -> Force Leech -> Full Force Lighting -> Full Force Lighting -> Full Force Lighting -> Clipped Force Lightning -> 

Clipped Force Lightning = Clip at 1.5s

 

The Basic idea for the rotation is:

Death Field -> GCD -> Demolish -> GCD -> Force Leech -> GCD -> GCD -> GCD* -> GCD -> GCD

 

With Affliction and Creeping Terror always spaced 1 GCD Apart (to allow for the 18s abilities to drift in between the 15s abilities)

 

The GCD* is the GCD that if it has either Affliction or Creeping Terror it causes a 2 GCD gap that needs to be filled. 2 GCD gaps are tricky because Force Lighting can't clip Force Lightning. That is why a Lighting Strike is used in those situations.

 

Do you think this would work?

[Paladrima]

Well, I think the general idea behind it is about as good as it's gonna get really.

For a theoretical rotation I think it'll work fine, but I don't know if any player is ever gonna be able to execute it to an extent that would make it "worth" over what's currently used, having 14 clipped Force Lightning's every 90 seconds seems like a daunting task with the way the game/engine is behaving. (I even had situations where the game just feels like it doesn't want to let me clip suddenly)

 

Besides Polarity Shift possibly spelling trouble here and there, you probably reached or got close to your goal for the theoretical maximum, but I'm not sure how close players will really get to that (looking at parsely as well).

[GrandHighAdmiral]

In general, I agree with the principle of the thing. Marksmanship is ranged, extremely high burst (with the advent of super crit), has LITERALLY no target swapping penalties, and has good AoE and survivability. It should be the lowest parsing spec in the game. However, it doesn't take much effort looking over Bant's list to see where balance is screwy, even in this revised world.

 

Marksmanship may have less target-swapping penalties than many classes, but it certainly has noticeable penalties - and no benefits like DoT Spread. First, Corrosive Dart: not re-applying it is a 5% loss, while re-applying it spends a GCD and is costly in terms of energy management. Second: Movement. Marksmanship is easily the least-mobile class in the game. Not only is almost every ability a cast or channel, but leaving cover erases our defenses, reduces our energy regeneration, and takes time to enter/leave cover. In many cases, this penalty significantly outweighs any benefits to being ranged. With the ninja-nerf to Entrench in 4.0, this is an even larger issue than in the recent past.

 

Regarding AoE, you're mostly right - but it's not without drawbacks. Suppressive Fire is powerful, but extremely costly in energy and single-target DPS. Contrast to Disciplines like Vengeance and Lightning which have powerful AoE abilities in their single-target rotation. Which method of AoE is preferable or more powerful varies on the situation, but it's unfair to simply claim that Marksmanship ought to lose potential single-target DPS based on access to a strong AoE that, in using it, already sacrifices significant single-target DPS.

 

Survivability is good on Snipers in general, but it already comes with costs: a lack of mobility and utility. Our only raid utility is Ballistic Shield - which is an excellent and under-utilized ability - but we have no mobility, no battle rez, no heals, no self-heals, no cleanse, no other role, etc. Is taking less damage in fights with unavoidable AoE is worth giving up flexibility and utility in every other fight as well? I think it's an even trade and I believe the intention is to incentivize raid leaders to bring a variety of classes. But again, it's not fair to ask us to give up our single-target DPS - our primary and only role - in return for a small incentive to bring us to a handful of fights.

 

Finally, while Marksmanship has moderate burst-damage potential, it's hardly as potent as AP or Lightning. Our offensive cooldowns are not as powerful and less frequent and our cast times are longer. In other words, we are not "more bursty" than any other burst spec. Certainly less burst than Lightning!

 

Maybe Marksmanship shouldn't be at the top of the chart, but saying that it should be the lowest parsing spec in the game is unfair - especially by as large a margin as it currently sits.

 

Update: 12/8/2015

 

Marksman || Sharpshooter : 6705 -> 6179 (-526 = -7.85%)

Speaking of that margin... how do you figure on only a 7.8% loss on Marksmanship? We lost 20% crit damage off 90% of our damage with an average crit rate of 50%. That's a 9% DPS loss without even considering the reduction on Penetrating Blasts (at least another 1%). That would more closely match the 11% loss I'm seeing in my own parses, and that others have reported as well. The top parse on Parsely (6097) since 4.0.3 is almost 13.5% below the top parse from earlier (6927), and (presumably) even with adrenals and good RNG hasn't managed to match your number yet. I recognize the small sample set there, but even the top-parsers own personal best (on Parsely) has dropped by 11.8%. Anyway, I think your calculation may be optimistic unless there is some adjustment I'm missing.

 

EDIT: Oh, just in case I came across a little belligerent - I'm not "calling yall out" or whatever, Bant and KeyboardNinja. I really appreciate yall's work.

Edited December 10, 2015 by GrandHighAdmiral

[Steefr]

Marksmanship may have less target-swapping penalties than many classes, but it certainly has noticeable penalties - and no benefits like DoT Spread. First, Corrosive Dart: not re-applying it is a 5% loss, while re-applying it spends a GCD and is costly in terms of energy management. Second: Movement. Marksmanship is easily the least-mobile class in the game. Not only is almost every ability a cast or channel, but leaving cover erases our defenses, reduces our energy regeneration, and takes time to enter/leave cover. In many cases, this penalty significantly outweighs any benefits to being ranged. With the ninja-nerf to Entrench in 4.0, this is an even larger issue than in the recent past.

 

A couple of things here, marksman has no target swapping penalties. Corrosive dart is somewhat costly i guess but on a 1,5 mil parse you only need to do 2 or 3 max rifle shots (where you skip corrosive dart) + the debuff from it actually lasts for 45 seconds so it should never fall off. Yeah ok mobility is kinda crappy in marksman tho roll is freaking awesome . Also no burst spechas dotspread.

 

 

Regarding AoE, you're mostly right - but it's not without drawbacks. Suppressive Fire is powerful, but extremely costly in energy and single-target DPS. Contrast to Disciplines like Vengeance and Lightning which have powerful AoE abilities in their single-target rotation. Which method of AoE is preferable or more powerful varies on the situation, but it's unfair to simply claim that Marksmanship ought to lose potential single-target DPS based on access to a strong AoE that, in using it, already sacrifices significant single-target DPS.

 

About the aoe, suppresive fire in marksman is still the single strongest aoe ability in the game and can be used twice in a row without seriously impacting energy. Don't forget it has a very large area of effect as well.

 

Survivability is good on Snipers in general, but it already comes with costs: a lack of mobility and utility. Our only raid utility is Ballistic Shield - which is an excellent and under-utilized ability - but we have no mobility, no battle rez, no heals, no self-heals, no cleanse, no other role, etc. Is taking less damage in fights with unavoidable AoE is worth giving up flexibility and utility in every other fight as well? I think it's an even trade and I believe the intention is to incentivize raid leaders to bring a variety of classes. But again, it's not fair to ask us to give up our single-target DPS - our primary and only role - in return for a small incentive to bring us to a handful of fights.

 

Our survivability is top notch in pve, like seriously it's excellent. Great utilities to take depending on the fight, multiple cooldowns that can be rotated very easily + a reset button. Mobilty is indeed less then other classes but it's honestly not a big of an issue as you're implying here. The no (self) cleanse is really the only part i do find annoying but oh well.

 

Finally, while Marksmanship has moderate burst-damage potential, it's hardly as potent as AP or Lightning. Our offensive cooldowns are not as powerful and less frequent and our cast times are longer. In other words, we are not "more bursty" than any other burst spec. Certainly less burst than Lightning!

 

While i can't really comment on the burstiness of lightning, Marksman still has serious burst. With laze target and sniper volley available, our burst rivals that of a pt and yeah for heavy burst phases you save your cooldowns. Imo it's not our burst that's a problem now, our sustained dmg is just too low, even for a burst spec. And i can't believe you're saying our offensive cooldowns aren't as powerful, laze target gives double super crit in the same time a pt gets 1 (even faster actually due to alacrity).

 

Maybe Marksmanship shouldn't be at the top of the chart, but saying that it should be the lowest parsing spec in the game is unfair - especially by as large a margin as it currently sits.

 

This part is true however, the nerf hits a bit too hard imo. For me personally i lost about 600-700 dps which is a tad too much imo. If they nerf woulda been by 400, then yeah np, that woulda been ok.

Speaking of that margin... how do you figure on only a 7.8% loss on Marksmanship? We lost 20% crit damage off 90% of our damage with an average crit rate of 50%. That's a 9% DPS loss without even considering the reduction on Penetrating Blasts (at least another 1%). That would more closely match the 11% loss I'm seeing in my own parses, and that others have reported as well. The top parse on Parsely (6097) since 4.0.3 is almost 13.5% below the top parse from earlier (6927), and (presumably) even with adrenals and good RNG hasn't managed to match your number yet. I recognize the small sample set there, but even the top-parsers own personal best (on Parsely) has dropped by 11.8%. Anyway, I think your calculation may be optimistic unless there is some adjustment I'm missing.

 

Well about that, i feel like i should say several things seeing as that 6097 parse atm is mine. First of all personal best in marksman pre-nerf was 6810. Secondly parsing marksman has become kinda depressing so only did a few parses, as evidenced by the low number of marksman parses on parsely atm. Sharpshooter high parse atm is 6200 or so and if i really tried i could do that too i think or even a bit higher. Lastly i'm not the best marksman player at all, much more proficient with engineering and especially virulence. But yeah 6200 average seems out of reach atm, still need 1 implant and 1 relic 224 but i figure i could maybe average 6,1 then. Maybe.

Edited December 10, 2015 by Steefr

[K_osss]

This is really good stuff. Thanks for all the hard work to put this together!

[Goblin_Lackey]

*snip*

 

Speaking of that margin... how do you figure on only a 7.8% loss on Marksmanship? We lost 20% crit damage off 90% of our damage with an average crit rate of 50%. That's a 9% DPS loss without even considering the reduction on Penetrating Blasts (at least another 1%). That would more closely match the 11% loss I'm seeing in my own parses, and that others have reported as well. The top parse on Parsely (6097) since 4.0.3 is almost 13.5% below the top parse from earlier (6927), and (presumably) even with adrenals and good RNG hasn't managed to match your number yet. I recognize the small sample set there, but even the top-parsers own personal best (on Parsely) has dropped by 11.8%. Anyway, I think your calculation may be optimistic unless there is some adjustment I'm missing.

 

EDIT: Oh, just in case I came across a little belligerent - I'm not "calling yall out" or whatever, Bant and KeyboardNinja. I really appreciate yall's work.

 

It is always good for people to check my work and make sure I am not just pulling numbers out of my ***.

 

EVERYTHING BELOW IS NOT HOW I CALCULATED THE PERCENT CHANGES!

What I actually did was record the all of the DPS numbers from 4.0.2 before I made any changes and compared them to the changed values.

 

 

Marksmanship Changes

 

Reminder: This is just a cursory look at the changes with back of the envlope math and ignores alot of smaller items.

 

Changes:

Bonus crit Damage is now 10% from 30% (effects Penetrating Blasts, Snipe, Followthrough, Ambush and Takedown)

Pentrating Blasts now does reduced damage (-5.5%)

 

Previous Percentages of Total DPS output:

Penetrating Blasts (23%)

Snipe (20.4%)

Followthrough (26.1%)

Ambush (18.8%)

Takedown (2.4%)

 

 

For ease of example; I will use 40% as the Base crit chance and 70% as the Bonus Crit

(True Optimal is at 1214 Critical (5372 Mastery): 38.15% Critical Chance and 67.4% Critical Bonus, but round numbers are nice for examples)

 

Average Damage increase from Critical chance = 1 + [Critical Bonus] * ( [Critical Chance] + [AutoCrit Percentage] )

Ignoring the Autocrit portion: Average Damage increase= 1 + [Critical Chance] * [Critical Bonus]

 

Average Damage Bonuses:

Old: Penetrating Blasts= 1 + [40% + 15% Critical Chance] * [70% + 30% Bonus Critical Damage] = 1.55

Old: Snipe = 1 + [40% + 19% Critical Chance] * [70% + 30% Bonus Critical Damage] = 1.59

Old: Followthrough = 1 + [40% + 4% Critical Chance] * [70% + 30% Bonus Critical Damage] = 1.44

Old: Ambush = 1 + [70% + 30% Bonus Critical Damage] * ( [40% Critical Chance] + [50% AutoCrit Percentage]) = 1.9

Old: Takedown = 1 + [40% + 15% Critical Chance] * [70% + 30% Bonus Critical Damage] = 1.55

 

New: Penetrating Blasts = 1 + [40% + 15% Critical Chance] * [70% + 10% Bonus Critical Damage] = 1.44

New: Snipe = 1 + [40% + 19% Critical Chance] * [70% + 10% Bonus Critical Damage] = 1.472

New: Followthrough = 1 + [40% + 4% Critical Chance] * [70% + 10% Bonus Critical Damage] = 1.352

New: Ambush = 1 + [70% + 10% Bonus Critical Damage] * ( [40% Critical Chance] + [50% AutoCrit Percentage]) = 1.72

New: Takedown = 1 + [40% + 15% Critical Chance] * [70% + 10% Bonus Critical Damage] = 1.44

 

So overall DPS changes:

Penetrating Blasts = (100% Damage - 5.5% Damage Nerf) * (1.44/1.55 Crit Change) = 87.8% of previous damage

Snipe = (1.472/1.59 Crit Change) = 92.5% of previous damage

Followthrough = (1.472/1.59 Crit Change) = 93.9% of previous damage

Ambush = (1.472/1.59 Crit Change) = 90.5% of previous damage

Takedown = (1.44/1.55 Crit Change) = 92.9% of previous damage

 

 

Taking into Account the previous percentages of total Damage:

Total Lost DPS: 23%*(1-87.8%) + 20.4%*(1-92.5%) + 26.1%*(1-93.9%) + 18.8%*(1-90.5%) + 2.4%(1-92.9%) =

7.88% DPS Loss in 4.0.3

 

/end of Back of Envelope Math

 

EVERYTHING ABOVE IS NOT HOW I CALCULATED THE PERCENT CHANGES!

What I actually did was record the all of the DPS numbers from 4.0.2 before I made any changes and compared them to the changed values.

This meant that I was using the optimal gearing for each version and it includes all the factors I have been using to calculate DPS.

[KeyboardNinja]

I suspect the differences between GrandHighAdmiral's calculations and Bant's is simply a matter of multiplicative vs comprehensive additive nerfs. A nerf from 30% to 10% of a surge increase that is, in and of itself, multiplicative would be enormous. However, other surge increases still remain that are additive with the existing 10%, and obviously the 10% itself is additive with base surge, so the whole thing is more complicated than it initially appears. And of course, this completely disregards stat changes which happen in response to the altered weights.

[GrandHighAdmiral]

Thanks for the replies.

 

I understand the concept of multiplicative vs. additive in and of itself, and your calculations all make sense to me when I read through them - but I still don't quite get why my simpler way doesn't also work. But apparently it doesn't, so I'll drop it. I am, however, still seeing a >10% drop in my own DPS in parses. Your list does not show any difference in optimal gearing, so it shouldn't be a gearing issue. It may be nothing more than my own demoralization from watching my close-to-BiS toon drop to nearly the DPS levels I used to get out of comms gear. It's hard to care about maximizing a parse when you watch such low numbers roll through. Where's my weeping rakghoul emoji, BW?

 

A couple of things here, marksman has no target swapping penalties. Corrosive dart is somewhat costly i guess but on a 1,5 mil parse you only need to do 2 or 3 max rifle shots (where you skip corrosive dart) + the debuff from it actually lasts for 45 seconds so it should never fall off..

 

A dummy parse is the epitome of a no-target-swapping situation, so the number of times you have to reapply it there is not relevant to target-swapping penalties. My point is that Marksmanship DOES have a penalty, and depending on the situation is either minor or enormous, like other classes.

 

About the aoe, suppresive fire in marksman is still the single strongest aoe ability in the game and can be used twice in a row without seriously impacting energy. Don't forget it has a very large area of effect as well.

 

You're missing the point. If AoE damage is too much, then nerf that - but Marksmanship already has to sacrifice single-target DPS in order to use Suppressive Fire. It doesn't make sense to double-tax Marksmanship. For that matter, BW has stated that they don't want to penalize any spec for having good AoE, but that every spec should have an appropriate amount of AoE. (I really like that every spec has different types of AoE. Not every type of AoE is equally useful in all situations, but each has situations in which it outperforms the others.)

 

While i can't really comment on the burstiness of lightning, Marksman still has serious burst. With laze target and sniper volley available, our burst rivals that of a pt and yeah for heavy burst phases you save your cooldowns. Imo it's not our burst that's a problem now, our sustained dmg is just too low, even for a burst spec. And i can't believe you're saying our offensive cooldowns aren't as powerful, laze target gives double super crit in the same time a pt gets 1 (even faster actually due to alacrity).

 

In short, I'm not arguing that Marksmanship doesn't have any burst damage, but that it doesn't have more burst damage than other specs. Both Lightning and AP have auto-crits in the rotation without having to use cooldowns AND have crit-boosting cooldowns. Recklessness is only a 60% boost, but considering the auto-crit Thundering Blast, that it has more charges and can be used up consecutively - it's a more potent burst-damage cooldown than Laze Target. Even Annihilation, the quintessential non-burst spec, has Annihilation (that crits for very nearly as much as our super-crit Ambush w/ half the CD once it's rolling), Berserk and Bloodthirst. In short, I think you're overestimating Marksmanship's burst relative to other classes.

 

That's really my over-arching point as well. The argument that Marksmanship should be the "lowest-parsing spec" is based off a series of assumptions about the relative strengths of Marksmanship that I believe are significantly over-estimated. I think it's more fun than any other class, and its popularity suggests I'm not unusual in that, but surely no one would claim a class should be punished for being fun? It's hardly the alpha class some folks - including BW,, based off this patch - seem to think it is.

 

Maybe when I'm not dealing with a cluster of real-life troubles I'll do another meta-comparison of relative raid performance compared to dummy parsing. Last time I did it (2.5 or so) I came up with a 3% gain for MM in raids relative to dummy performance, compared to other classes - with a 5-6% best-case gain. Then again, math clearly isn't my strongest subject so maybe someone else should do it.

 

It occurs to me that this might not be the right thread to debate a specific Discipline, so I apologize if I'm out of line. In any case, I'll drop the whole thing. I don't expect any changes based on my complaints anyway.

[KeyboardNinja]

In short, I'm not arguing that Marksmanship doesn't have any burst damage, but that it doesn't have more burst damage than other specs. Both Lightning and AP have auto-crits in the rotation without having to use cooldowns AND have crit-boosting cooldowns. Recklessness is only a 60% boost, but considering the auto-crit Thundering Blast, that it has more charges and can be used up consecutively - it's a more potent burst-damage cooldown than Laze Target. Even Annihilation, the quintessential non-burst spec, has Annihilation (that crits for very nearly as much as our super-crit Ambush w/ half the CD once it's rolling), Berserk and Bloodthirst. In short, I think you're overestimating Marksmanship's burst relative to other classes.

 

Well, I can tell you based on some recent reprogression (I guess that's a thing now?) that Marksmanship has the highest burst among IO, Arsenal and Fury, and it isn't even CLOSE. (because of some rescaling, we had to reshuffle some DPS assignments on a particular NiM boss to apply the highest burst in our group, and our Marksmanship sniper won by a mile)

 

I can also tell you that Lightning is no where near as bursty as you think it is. It used to be, but the surge nerf hit hard, and we have far fewer crit boosts and far weaker base abilities. In fact, based on openers, Lightning actually has worse burst than Madness, if you measure out to a 15 second window. Marksmanship is almost certainly much higher.

[LudhaninRolgge]

Well, I can tell you based on some recent reprogression (I guess that's a thing now?) that Marksmanship has the highest burst among IO, Arsenal and Fury, and it isn't even CLOSE. (because of some rescaling, we had to reshuffle some DPS assignments on a particular NiM boss to apply the highest burst in our group, and our Marksmanship sniper won by a mile)

 

I can also tell you that Lightning is no where near as bursty as you think it is. It used to be, but the surge nerf hit hard, and we have far fewer crit boosts and far weaker base abilities. In fact, based on openers, Lightning actually has worse burst than Madness, if you measure out to a 15 second window. Marksmanship is almost certainly much higher.

 

How do you calculate burst ? I'm very interested.

Edited December 12, 2015 by LudhaninRolgge

[nerfderp]

How do you calculate burst ? I'm very interested.

 

I guess you could look at Starparse and have a look at the damage output of each DPS within a set time frame. There are also tabs in Starparse that sort by phase so you could have a look at a particular instance where burst is required (ie, Chained Manifestation on Styrack).

[KeyboardNinja]

How do you calculate burst ? I'm very interested.

 

You can do it emperically by having someone go ham on a dummy for a set period of time, then measure how much damage they did. We did it somewhat less emperically by trying out different specs in a particular role in the fight until we found one that could beat the damage check.

[LudhaninRolgge]

I was not doing that wrong, thanks for your answers ^^

[GrandHighAdmiral]

I promise I won't rehash any arguments about Marksmanship, but I've been interested for a long time in trying to rate the "burstiness" of each spec, but I've not managed to nail down a useful and reliable way to work it out. There's too much variability in definitions, conditions, and in the parses themselves - especially given the inherently short time window.

 

Would it be possible to use the same method used to generate the expected sustained DPS numbers and optimal gear stats to generate an "expected burst damage" rating for each spec?

 

How to define/rate burst?

• A 5, 10, or 15-second window?
  
• A simulated TTK (i.e., how long to deal, say, 100k damage?)
  
• A variable time-window? (since different classes probably have differing lengths for their "burst phase")
  
• A rating (e.g.,"115% of baseline") or number ("15s @ 7000dps")?
  

 

Other complications:

• Include pre-cast or pop-out-of-stealth abilities which might not be possible to use in a boss fight?
  
• Would the optimal stats be different for burst damage vs. sustained? (That information might be quite useful for NiM, come to think of it!)
  
• Speaking of NiM, should consistency be considered or rated? (E.g., Auto-crits) Perhaps like tanks' "spikiness"?
  
• While I presume raid buffs should be excluded, they are still worth mentioning, since classes with a raid buff are guaranteed to have that ability available to them. They can also, perhaps, time the buff better to benefit their own damage.
  

 

Anyway, I don't know how difficult this would be, because I don't know what your tools look like, but it might be useful information - interesting, at least - if it's feasible.

[Simmerr]

Regarding burst, another aspect to consider is a spec's set up time (ie target swapping when an add that needs to die quick spawns) (think TFB kephess white orbs or Underlurker adds)

 

Sure if you can burst for 100k in 3 gcd's its nice, but if it takes another 3 gcd's to set up for that, its really a 6 gcd burst.

 

(This would have to include dots that proc things and buffs required from rotational abilities)

 

I have no idea how we could test for this though