Optimal Stats For All 24 Disciplines, KOTFE Edition

[NepomukLP]

Tank gearing question:

 

I see there are b mods in 220,

b enhancements (random drops from OPs) and c enhancements (glowing crystals) in 220.

 

Actually I consider in max out the hp, so it would be intresting to see what theory crafting says to b mods and c enhancement. Could someone post it?

[GrandLordMenace]

Tank gearing question:

 

I see there are b mods in 220,

b enhancements (random drops from OPs) and c enhancements (glowing crystals) in 220.

 

Actually I consider in max out the hp, so it would be intresting to see what theory crafting says to b mods and c enhancement. Could someone post it?

 

I don't think it was ever theorycrafting, more of practicality juxtaposed with theory.

 

see: http://dulfy.net/2014/12/27/swtor-3-0-shield-specialist-vanguard-tanking-guide-by-milas/#Gearing_and_Stats_Priority

 

Basically Zorz had a very different mindset from the tanks at the time that was fairly revolutionary in comparison to the mean mit appoach. Neither at the time were inherently WRONG, and you COULD say Zorz's approach would be wrong by some standards if you're looking to reduce max damage, but Zorz from what I took, didn't really care about miniscule damage reduction in relation to gear, and more about damage reduction in relation to DCDs, and giving healers enough of a buffer so that the tanks could survive burst damage, all while ensuring only a relatively minimal loss in mitigation.

 

KBN could probably give a better explanation, but that's what I've gleaned talking to Justin, KBN, etc over the past 2 years.

[dipstik]

Tank gearing question:

 

I see there are b mods in 220,

b enhancements (random drops from OPs) and c enhancements (glowing crystals) in 220.

 

Actually I consider in max out the hp, so it would be intresting to see what theory crafting says to b mods and c enhancement. Could someone post it?

 

there is some discussion about that in my tank stats thread around post 23

 

 

_____quote:

 

lets compare the 138 and 168 endurance enhancements to the 104 endurance enhancements.

168v104

168-104: 64 end

44-116: -72 def

159-168: -9 shield/absorb

 

now 138v104:

138-104: 34 end

77-116: -39 def

168-168=0 shield /absorb

 

looking at adding 10 defense, shield or absorb to m=0.5 and m=0.8 profiles gives for damage taken:

m=0.5 base +10def +10shd +10abs ddef dshd dabs

sin 1492.7 1492.3 1491.1 1491.1 0.4 1.6 1.6

jugg 1303.6 1303.1 1302.6 1302.6 0.5 1 1

pt 1316.3 1315.9 1315 1315 0.4 1.3 1.3

 

m=.8 ddef dshd dabs

sin 1169.8 1169.2 1168.6 1168.6 0.6 1.2 1.2

jugg 1044.5 1043.8 1043.6 1043.6 0.7 0.9 0.9

pt 1122.9 1122.3 1121.8 1121.8 0.6 1.1 1.1

 

assuming 6k boss dps.

 

for sins in m=.5, there is a 4:1 ratio between the worth of defense to shild/absorb, 2:1 for juggs and a little less than 4:1 for pts. at m=.8 we get 2:1 for sin, almost 1:1 for jugg and almost 2:1 for pts.

 

lets take some averages and take the worth of shield/absorb in relation to defense as 3:1 for low melee/ranged fights. so losing 72 points of defense is like losing 24 points of shield/absorb for low m/r fights. so the total for 168v104 endurance enhancements would be 9+27=33 points of shield/absorb loss for 64 endurance. thats a ratio of 2:1 of endurance gained from mitigation lost. for the 138v104 you lose an effective 13 points of mitigation for 34 endurance, which is 2.6:1, much better.

 

for the m=0.8 fights we get shield/absorb to defense has a 1.5 ish ratio. 168x104 gives (72/1.5)+9=35 mitigation loss for 64 endurance, so a 1.8:1 trade. 138v104 we get 39/1.5=26 mitigation for 34 endurance, giving a 1.3:1 trade, much worse.

 

the bmods give 34 endurance for 29 more defense which equates to 10 to 15 shield/absorb, so the trade comes out to 2.3:1 to 3:1. so it looks like trading out 104 endurance enhancements for 138 endurance enhancements before mods might be a thing for low melee/ranged fights (since the 2.6:1 is better than the 2.3:1).

 

tl/dr: the 168 endurance enhancements look better for high melee/ranged fights, and the 138 endurance enhancements look better for low melee/ranged fights.

Edited March 3, 2016 by dipstik

[Ryuku-sama]

Annihilate > Force Rend (Deadly Saber) > Battering Assault > Dual Saber Throw > Rupture > Filler
Annihilate > Ravage > Ravage* > Force Rend (Deadly Saber) > Rupture > Filler
Annihilate > Force Scream > Battering Assault > Dual Saber Throw > Rupture > Force Rend (Deadly Saber)
Annihilate > Ravage > Ravage* > Filler > Rupture > Filler
Filler = Vicious Throw > Vicious Slash > Assault

 

Why are you using Force Scream? Force Scream damage is just slightly better than Vicious Slash. Not using it would remove any need to use Assault. You use Force Scream once every 24 GCD and Assault once every 30 GCD. Replacing Force Scream by Vicious Slash would remove the need for Assault and even make it so you have more rage than you can spend.

 

And it is more damage.

FS/24 + Assault/30 = 5542/24 + 1945/30 = 295.75 damage/GCD

VS/24 + VS/30 = 5245/24 + 5245/30 = 393.38 damage/GCD

 

EDIT : Damn.. Took me some time to understand how you worked with Deadly Saber..

Edited March 9, 2016 by Ryuku-sama

[Goblin_Lackey]

Why are you using Force Scream? Force Scream damage is just slightly better than Vicious Slash. Not using it would remove any need to use Assault. You use Force Scream once every 24 GCD and Assault once every 30 GCD. Replacing Force Scream by Vicious Slash would remove the need for Assault and even make it so you have more rage than you can spend.

 

And it is more damage.

FS/24 + Assault/30 = 5542/24 + 1945/30 = 295.75 damage/GCD

VS/24 + VS/30 = 5245/24 + 5245/30 = 393.38 damage/GCD

 

EDIT : Damn.. Took me some time to understand how you worked with Deadly Saber..

 

It was mainly because I was looking around on parsely to try and better match up my theoretical with what can actually be done. Because annihilation/watchman was doing much better than I was projecting I looked into the reasons. Once I eliminated the luck aspect I focused on the rotation. I was surprised to find force Scream used as much as it wasl; but further examination showed that it benefited the rotation and was present in most of the top parses.

 

I browse through parsely regularly for this post because it provides a example window for rotations and where I should focus my attention. The actual dps numbers are irrelevant as are the critical hit rates; those are both far too RNG based to be useful. It's the player controlled rotations and ability usage rates that are the gold mine for useful information.

[Ryuku-sama]

It was mainly because I was looking around on parsely to try and better match up my theoretical with what can actually be done. Because annihilation/watchman was doing much better than I was projecting I looked into the reasons. Once I eliminated the luck aspect I focused on the rotation. I was surprised to find force Scream used as much as it wasl; but further examination showed that it benefited the rotation and was present in most of the top parses.

 

I browse through parsely regularly for this post because it provides a example window for rotations and where I should focus my attention. The actual dps numbers are irrelevant as are the critical hit rates; those are both far too RNG based to be useful. It's the player controlled rotations and ability usage rates that are the gold mine for useful information.

 

Well... When I get some simulated data ready to be shared, I'll bring the subject back on the floor Until then.. I am skeptical.

[Goblin_Lackey]

Well... When I get some simulated data ready to be shared, I'll bring the subject back on the floor Until then.. I am skeptical.

 

Completely fair, I assume that I'm likely wrong (in everything) as well.

But currently, I don't have the available time to ready drill down and do the problem justice.

As always, I welcome other people's inputs and work because it is always a chance to improve.

Edited March 9, 2016 by Goblin_Lackey

[Trasid]

I have a question regarding Sorc/Sage healing - I have followed the below stat distribution and have the exact amount of power, crit and alacrity but my mastery and endurance stats are less than what is stated here, at 5025 mastery and 6105 endurance. I am using Advanced polybiotic versatile stim.

 

Sorcerer - Corruption || Sage - Seer

11431 HPS @ 47.2 APM | 6127 Endurance | 5179 Mastery (Stim) | 2788 Power

1432 Critical (7xE, 2xA, 2xC) | 1356 Alacrity (3xE, 12xA) | 0 Accuracy

Edited March 12, 2016 by Trasid

[LudhaninRolgge]

I have a question regarding Sorc/Sage healing - I have followed the below stat distribution and have the exact amount of power, crit and alacrity but my mastery and endurance stats are less than what is stated here, at 5025 mastery and 6105 endurance. I am using Advanced polybiotic versatile stim.

 

Sorcerer - Corruption || Sage - Seer

11431 HPS @ 47.2 APM | 6127 Endurance | 5179 Mastery (Stim) | 2788 Power

1432 Critical (7xE, 2xA, 2xC) | 1356 Alacrity (3xE, 12xA) | 0 Accuracy

 

Go hunt for datacrons

[Trasid]

Go hunt for datacrons

 

Of course, thanks!! Can't believe I overlooked that.

[AjentWashington]

I believe your virulence calculations may be a bit off, at least in terms of DPS. The rotation you use is

 

 

Corrosive Dart > Corrosive Grenade > Weakening Blast > Lethal Shot > Cull > Cull*

Lethal Shot > Filler > Filler > Takedown > Cull > Cull*

Series of Shots > Series of Shots* > Weakening Blast > Lethal Shot > Cull > Cull*

Corrosive Dart > Corrosive Grenade > Filler > Takedown > Cull > Cull*

Series of Shots > Series of Shots* > Weakening Blast > Lethal Shot > Cull > Cull*

Lethal Shot > Filler > Filler > Takedown > Cull > Cull*

Filler = Lethal Shot > Takedown > Corrosive Mine > Rifle Shot

 

 

However, the correct rotation is

Corrosive Dart > Corrosive Grenade > Takedown > Weakening Blast > Cull > Cull*

Lethal Shot > Filler > Filler > Filler > Cull > Cull*

Series of Shots > Series of Shots > Takedown > Weakening Blast > Cull > Cull*

Corrosive Dart > Corrosive Grenade > Filler > filler > Cull > Cull*

Series of Shots > Series of Shots > Takedown > Weakening Blast > Cull > Cull*

Lethal Shot > Filler > Filler > Filler > Cull > Cull*

 

(All changes are highlighted in red)

 

The reason for weakening blast being moved directly in front of cull is because with high enough alacrity (800 is enough) you can fit the first tick of the 2nd cull after a weakening blast into that 10 second window, and so you get 3 ticks of weakening blast in (1 from cull, 1 from CD, one from CG) rather than one from the lethal shot. The damage loss from using a takedown instead of a lethal shot inside WB is very minor before execute phase (you can also hard cast lethal shot before a weakening blast and the damage actually ticks after you activate weakening blast (I will often do this in my parses, especially sub 30%), however it is not easily doable in HM ops and I urge you to use TD > WB rather than LS > WB in your calculations).

 

Also, overload shot should be in the fillers as well right before rifle shot. Overload shot is free with the utility calculated pursuit, and does slightly more damage than rifle shot, thus the fillers priority should be

Filler = Lethal Shot > Takedown > Corrosive Mine > Overload Shot > Rifle Shot

 

The additional fillers instead of Takedown are because it is better to use a lethal shot under WB than a Takedown, so lethal shot should be considered before using takedown

Edited March 13, 2016 by AjentWashington

[Wattyzehotty]

love this post keep up teh good work

[Trasid]

After going on my datacron hunt, I have come to 5177 mastery, but the guide says 5179. I have checked to make sure I collected all mastery datacrons, and I have - I checked like 10 times!! Anyone know why I am out by 2?

[Ryuku-sama]

Marauder-Fury | Sentinel-Concentration

Rotation:
Raging Burst > Ravage > Ravage* > Furious Strike > Force Scream > Battering Assault > Obliterate
Raging Burst > Force Crush > Filler > Furious Strike > Filler > Obliterate
Filler = Vicious Throw > Vicious Slash

References: Oofalong

 

Something is wrong here. Your spreadsheet code says both your RB are under the 4p. For that to be possible, you have to pop Berserk just before the second cycle's RB and even then, it's tight without alacrity. That means your FC > VS/VT following it will never generate any Fury. That also means Your FS may not generate any if you miss your timing by half a second. That means your rotation will generate, depending on how many times Ravage hits, 23 and 26 Fury. You're off by 4 to 7 Fury.

That's unless RB and Sc are actually increasing your Fury despite costing nothing.

 

EDIT :

With enough alacrity (around 8%) you could use Berserk before your first RB and it would cover the second. that mean the 6 Berserk stacks would last for 4 GCD's. That means your FS wouldn't generate Fury. This means you'd generate 27-30 Fury. That's a little too much luck reliant IMO. You could, by using Berserk immediately after Ob, make it so your FS generate Fury. But that would put your second RB out of the 4p...

 

EDIT :

After looking closely at Artorias video I've seen RB generating Fury stacks. So I guess this works

 

EDIT :

So your rotation is viable.. And your results are good as long as you don't go under 3-4% alacrity. Under that and your second RB will be out of your 4p.

Edited March 15, 2016 by Ryuku-sama

[Goblin_Lackey]

I believe your virulence calculations may be a bit off, at least in terms of DPS. The rotation you use is

 

 

 

However, the correct rotation is

Corrosive Dart > Corrosive Grenade > Takedown > Weakening Blast > Cull > Cull*

Lethal Shot > Filler > Filler > Filler > Cull > Cull*

Series of Shots > Series of Shots > Takedown > Weakening Blast > Cull > Cull*

Corrosive Dart > Corrosive Grenade > Filler > filler > Cull > Cull*

Series of Shots > Series of Shots > Takedown > Weakening Blast > Cull > Cull*

Lethal Shot > Filler > Filler > Filler > Cull > Cull*

 

(All changes are highlighted in red)

 

The reason for weakening blast being moved directly in front of cull is because with high enough alacrity (800 is enough) you can fit the first tick of the 2nd cull after a weakening blast into that 10 second window, and so you get 3 ticks of weakening blast in (1 from cull, 1 from CD, one from CG) rather than one from the lethal shot. The damage loss from using a takedown instead of a lethal shot inside WB is very minor before execute phase (you can also hard cast lethal shot before a weakening blast and the damage actually ticks after you activate weakening blast (I will often do this in my parses, especially sub 30%), however it is not easily doable in HM ops and I urge you to use TD > WB rather than LS > WB in your calculations).

 

Also, overload shot should be in the fillers as well right before rifle shot. Overload shot is free with the utility calculated pursuit, and does slightly more damage than rifle shot, thus the fillers priority should be

Filler = Lethal Shot > Takedown > Corrosive Mine > Overload Shot > Rifle Shot

 

The additional fillers instead of Takedown are because it is better to use a lethal shot under WB than a Takedown, so lethal shot should be considered before using takedown

 

 

I am actually going to refer you back to a previous post in this same thread:

Post #48, on Page 5

http://www.swtor.com/community/showthread.php?t=847112&page=5

 

 

 

That was where I got the rotation I used and explained how it changed from the previous rotation I had. The Rotation you posted actually has the same ability activation rates (because you just swapped things around without changing the total number) so in the eyes of the optimizer it looks identical.

 

In any case, I estimated the number of hits for Weakening Blast at ~24.8. This is based on the time averaged tick rate of triggering abilities per second, then multiplying the result by 10 and adding 1 for initial hit. This actually does makes the tick rate match the number of ticks on parsely.io.

 

 

Regarding Overload Shot: I didn't include it for any of the snipers, mainly because I missed it. In Virulence's case, I have the ability activation rate for Rifle shot to be only used once every 81 seconds on average (due to filler priority and avalible energy). This means Overload Shot replacing it would not add much to the total.

[Goblin_Lackey]

Something is wrong here. Your spreadsheet code says both your RB are under the 4p. For that to be possible, you have to pop Berserk just before the second cycle's RB and even then, it's tight without alacrity. That means your FC > VS/VT following it will never generate any Fury. That also means Your FS may not generate any if you miss your timing by half a second. That means your rotation will generate, depending on how many times Ravage hits, 23 and 26 Fury. You're off by 4 to 7 Fury.

That's unless RB and Sc are actually increasing your Fury despite costing nothing.

 

EDIT :

With enough alacrity (around 8%) you could use Berserk before your first RB and it would cover the second. that mean the 6 Berserk stacks would last for 4 GCD's. That means your FS wouldn't generate Fury. This means you'd generate 27-30 Fury. That's a little too much luck reliant IMO. You could, by using Berserk immediately after Ob, make it so your FS generate Fury. But that would put your second RB out of the 4p...

 

EDIT :

After looking closely at Artorias video I've seen RB generating Fury stacks. So I guess this works

 

EDIT :

So your rotation is viable.. And your results are good as long as you don't go under 3-4% alacrity. Under that and your second RB will be out of your 4p.

 

 

First of all, Thanks for taking the time to audit the spreadsheet. It badly needs it due to the sheer number of assumptions made (and likely not evenly applied).

 

 

With Fury | Concentration, My starting point was the steady state rotation put forth by Oofalong.

 

Originally, I time averaging every buff, like the 4 piece, across the entire rotation. This was adequate for my purpose for a while, but then when I was working on Rage | Focus Juggernauts I realized that because the rotation was set in stone, certain buffs would always be present for certain abilities. I went back later and realized that since berserk is used in the rotation at the same place every time, its buffs always affected the same abilities. I then redetermined the buffs that applied to each of the moves based entirely on the rotation.

 

I gave the benefit of the doubt to any ability that could be affected by the buffs due to the affects of Alacrity and the fact that Rage and Fury always seemed to be undeservedly at the bottom.

[ProcessHeals]

APM for corruption sorc/seer sage appears to have been miscalculated. I understand that it's not supposed to be representative of what could be achieved in an actual fight, but the math did not add up for me.

 

Using the cited rotation by Orderken, APM should be closer to 40. 2 healing trances @ about 2.12~ sec + 8 abilities that will take at least 1.32~ sec each (potentially more depending on filler used and the amount of force surge / resplendence stacks available for salv/revi)

 

So assuming only gcd-length fillers used (ie anything but unprocced dark infusion/deliverance) and perfect crit rng for healing trance/innervate that leaves us with:

 

2.12s x2

+

1.32s x8

_______

14.8s per rotation of 10 abilities

 

You can complete this 10 ability rotation 4.05~ times per minutes which totals 40~ APM.

 

After studying the math, I can reasonably conclude that your calculator does not reflect the following:

 

-Consuming darkness/vindicate are always ON the gcd (unlike pre-3.3 when the old regen abilities could be activated off gcd when used with stacks of force surge / resplendence).

 

The difference of those 1.32 (x2) seconds per rotation adds up to your listed APM values

 

 

I wanted to bring this to your attention, because it should have a noticeable effect on hps and may even change the stat priority.

 

 

On a lighter note, I would like to sincerely thank you for all the work you've put into this thread. I hope that my contributions help improve the accuracy of your model.

Edited March 16, 2016 by Tshain

[Ryuku-sama]

First of all, Thanks for taking the time to audit the spreadsheet. It badly needs it due to the sheer number of assumptions made (and likely not evenly applied).

 

 

With Fury | Concentration, My starting point was the steady state rotation put forth by Oofalong.

 

Originally, I time averaging every buff, like the 4 piece, across the entire rotation. This was adequate for my purpose for a while, but then when I was working on Rage | Focus Juggernauts I realized that because the rotation was set in stone, certain buffs would always be present for certain abilities. I went back later and realized that since berserk is used in the rotation at the same place every time, its buffs always affected the same abilities. I then redetermined the buffs that applied to each of the moves based entirely on the rotation.

 

I gave the benefit of the doubt to any ability that could be affected by the buffs due to the affects of Alacrity and the fact that Rage and Fury always seemed to be undeservedly at the bottom.

 

No worry.. I'm currently building rotations for simulation purposes.. And I use your spreadsheet to time average everything It's as good as it can get without simulating So.. Well.. You will likely see me pop out and comment most spec's rotation if I get something else than what you got

[Goblin_Lackey]

APM for corruption sorc/seer sage appears to have been miscalculated. I understand that it's not supposed to be representative of what could be achieved in an actual fight, but the math did not add up for me.

 

Using the cited rotation by Orderken, APM should be closer to 40. 2 healing trances @ about 2.12~ sec + 8 abilities that will take at least 1.32~ sec each (potentially more depending on filler used and the amount of force surge / resplendence stacks available for salv/revi)

 

So assuming only gcd-length fillers used (ie anything but unprocced dark infusion/deliverance) and perfect crit rng for healing trance/innervate that leaves us with:

 

2.12s x2

+

1.32s x8

_______

14.8s per rotation of 10 abilities

 

You can complete this 10 ability rotation 4.05~ times per minutes which totals 40~ APM.

 

After studying the math, I can reasonably conclude that your calculator does not reflect the following:

 

-Consuming darkness/vindicate are always ON the gcd (unlike pre-3.3 when the old regen abilities could be activated off gcd when used with stacks of force surge / resplendence).

 

The difference of those 1.32 (x2) seconds per rotation adds up to your listed APM values

 

 

I wanted to bring this to your attention, because it should have a noticeable effect on hps and may even change the stat priority.

 

 

On a lighter note, I would like to sincerely thank you for all the work you've put into this thread. I hope that my contributions help improve the accuracy of your model.

 

I took a look into it and the reason for the higher apm is actually due to a minor error on my part that won't affect anything.

 

The apm gain above the rotation was because I include a catch-all ability I call activated abilities; this is just a place holder ability that just functions as apm counter for all of the non gcd abilities so that I can include an accurate apm output.

 

In corruptions case, I include polarity shift and recklessness in this category so that they will be included in the apm count. I also include activating unnatural preservation on cooldown as well.

 

The error I made was that I had a pointer that should have been aimed at the usage rate of unnatural preservation, but was instead pointing at static barrier (column F instead of column E, just a misclick)

 

This all adds up to no change in hps, just apm changes due to input error.

 

224: 44.2 apm

220: 44.1 apm

216: 44 apm

Edited March 16, 2016 by Goblin_Lackey

[Ryuku-sama]

I've found a rotation for Fury that does a whopping 5 DPS over your rotation Pretty much nothing. But I messed around this due to PvP habits where I will normally be at range for RB and close in with Ob or Leap for FS. It would have been easy if I hadn't tried to get the most out of Berserk and Cascading Power. Finally I settled with the following change :

 

RB + Berserk > Rav > Rav* > Ob > FS > VS > BA >
RB > Sc/VT > Ob > FS > VS/Sc > FC

 

So this means....

• Every ticks of FC except the last will be out of both Cascading Power and Berserk. In fact it is likely the third tick will be inside Cascading Power.
  
• One RB will be out of Berzerk, the other will be inside at more than 2% alacrity.
  
• Both Ob will be inside Cascading Power.
  
• At more than 8% alacrity, Scream will also be inside Cascading Power during the Execute.
  

 

So, this is a DPS increase as long as you have more than 8% alacrity. I didn't test it for lower amount. Seems too much of a bother considering optimal amount for every spec is above that.

Edited March 16, 2016 by Ryuku-sama

[AjentWashington]

I am actually going to refer you back to a previous post in this same thread:

Post #48, on Page 5

http://www.swtor.com/community/showthread.php?t=847112&page=5

 

 

 

That was where I got the rotation I used and explained how it changed from the previous rotation I had. The Rotation you posted actually has the same ability activation rates (because you just swapped things around without changing the total number) so in the eyes of the optimizer it looks identical.

 

In any case, I estimated the number of hits for Weakening Blast at ~24.8. This is based on the time averaged tick rate of triggering abilities per second, then multiplying the result by 10 and adding 1 for initial hit. This actually does makes the tick rate match the number of ticks on parsely.io.

 

 

Regarding Overload Shot: I didn't include it for any of the snipers, mainly because I missed it. In Virulence's case, I have the ability activation rate for Rifle shot to be only used once every 81 seconds on average (due to filler priority and avalible energy). This means Overload Shot replacing it would not add much to the total.

 

I don't think that your value for Weakening blast is correct. In both practice and theory I am getting around 27.4 ticks per activation of weakening blast (I isually have 13 activations of weakening blast and 355-360 hits). Without counting dots caused by ticks of cull, weakening blast is guaranteed to trigger 7 times based off of out for ticks with ~10% alacrity. There is the possibility of an eight tick to be guaranteed with 0 ability lag, but I don't count that because it is impossible to actually do. Normally dots would only trigger it 6 times for sure if they were about to run out, but they can be reapplied 1 GCD late beforehand to keep it at 7 ticks. 5 ticks per dot happen inside WB because of cull, so there are 17 dot ticks withen the window. Multiply this by 10/9 to account for the chance for dots to tick twice and you get 18.888... . Add in the first WB tick and the 5 ticks of cull and you get 24.888... Ticks without lethal shot added in. I find that I get about 2.5 lethal shots per weakening blast activation (total lethal shot hits/(2xweakening blast activations)), so the average number of hits should be about 27.388.... You can't take the average activation rate of lethal shot x 10 because it should only be used under weakening blast.

[Goblin_Lackey]

I don't think that your value for Weakening blast is correct. In both practice and theory I am getting around 27.4 ticks per activation of weakening blast (I isually have 13 activations of weakening blast and 355-360 hits). Without counting dots caused by ticks of cull, weakening blast is guaranteed to trigger 7 times based off of out for ticks with ~10% alacrity. There is the possibility of an eight tick to be guaranteed with 0 ability lag, but I don't count that because it is impossible to actually do. Normally dots would only trigger it 6 times for sure if they were about to run out, but they can be reapplied 1 GCD late beforehand to keep it at 7 ticks. 5 ticks per dot happen inside WB because of cull, so there are 17 dot ticks withen the window. Multiply this by 10/9 to account for the chance for dots to tick twice and you get 18.888... . Add in the first WB tick and the 5 ticks of cull and you get 24.888... Ticks without lethal shot added in. I find that I get about 2.5 lethal shots per weakening blast activation (total lethal shot hits/(2xweakening blast activations)), so the average number of hits should be about 27.388.... You can't take the average activation rate of lethal shot x 10 because it should only be used under weakening blast.

 

I would like to clarify why I am using this number for Weakening Blast Hits:

 

Total Weakening Blast hits = 10s * [# of Weakening Blast trigging hits per second]

 

Triggering Hits

Corrosive Grenade: 1.1*(9 + 3*4)/ 27s = 0.856 Hits/second

Cull: 4 / 9s = 0.444 Hits/second

Vital Shot: 1.1*(9 + 3*4)/ 27s = 0.856 Hits/second

Lethal Shot: 1 / ~8.5s = 0.117 Hits/second

Corrosive Mine (Roll): 3 / 27s = 0.111 Hits/second

 

Weakening Blast: +1 initial Hit

 

Total # of Weakening Blast Hits = 10s * (0.856 + 0.444 + 0.856 + 0.117 + 0.111) + 1 Hit

Total # of Weakening Blast Hits = 24.8 Hits

 

 

The biggest issue here of course is that I am only using a flat 10s with no alacrity boost to determine the number of hits.

With 10% alacrity for example, the window should effectively be 11s in relation to the rotation, with an estimated # of hits of 27.2.

 

One of the weaknesses of the spreadsheet is that I have to calculate all ability uses before I know how much alacrity a discipline has from gear. This means that I can't assume that I would have a minimum Alacrity value at any given moment. Removing the not knowing gear restriction would be a ton of work and involve completely remaking the entire document from the ground up. I originally built the spreadsheet to handle early 3.0 stats (with lower alacrity values) and have just iterated on top of that since; the decisions I made then still affect and sometimes limit the rest of the structure.

 

 

TL/DR

I handle buff durations with Alacrity gear poorly.

[Ryuku-sama]

Forget that. I got even better...

 

And by using a more modified version.. I got to 6363-6380 DPS on your model depending on if I put the second FS inside Berserk or not. A high alacrity build (8%) can likely do it.. Here is the rotation :

RB > Ob + Berserk > FS > VS > Rav (clip after the GCD) > BA >
RB > Ob > FS > VS/VT > Sc > FC

 

Removing Ravage last tick actually speeds up the rotation enough that the higher use of pretty much everything is enough to overcome the loss. This also free up space inside everything so i can put all the heavy hitters inside Cascading Power. Note that this will only ever work with at least 4% alacrity. Else you risk your last attack falling out of Cascading Power which is a rather steep blow to your damage. Use VT only during the second block because travel time would put it out of Cascading Power in the first cycle. Use Berserk as soon as you get 30 Fury after Ob.

 

EDIT : Everything should be working

Edited March 16, 2016 by Ryuku-sama

[Ryuku-sama]

New update

So you got 6375 DPS on Fury after a fix I did on your math explained below....

 

After making sure everything is fine buff/debuff wise, under the assumption that everyone got at least 8% alacrity, and by using the rotation here :

RB > Ob + Berserk > FS > VS > Rav (clip after the GCD) > BA >
RB > Ob > FS > VS/VT > Sc > FC

 

I got 6359. So clipping Ravage doesn't actually increase DPS... Not once I made sure your work was under the same alacrity assumption as mine.

 

On another note. Your way to handle 2p could easily be reworked for Fury and any other spec with rotation that never change. For Fury, this means your rotation DPS is actually a little higher than what you got.

Without the execute section, which is useless for Fury since their timeUsage on BA never change, you use :

 

duration/(roundup(CD/timeUsage)*timeUsage)

 

I would use :

 

1/roundup(CD/timeUsage)

 

You don't need to time average.. With Fury, you can just determine what is in it and apply the set bonus or not. That's about 7 DPS you're missing here.

 

Another thing I found strange in Fury is how you calculate APM.. Fury abilities that aren't on the GCD are Berserk (once every 19.2s in your math), Frenzy and Bloodthirst. Activating Frenzy outside of bloodthirst is useless for Fury.. It would just mess up the rotation without any real gain.

 

One last thing... Why does Fury have a negative execute damage bonus of -5%/2??? That's almost 1 DPS here

 

I lied... Another thing.. you do know, since you determine DPS before execute time, your way to handle execute variation in buff/debuff uptime isn't perfect? It's a decent approximation, but to get a good one you would have to determine execute time averaged Damage Bonus separately. I couldn,t change that before having to rework your spreadsheet from the ground up.. So I worked under the assumption that for calculating buff and debuff uptime, the execute last 30% of the time.

 

Around the same thing... Your VS and Force Lash bonus damage is badly handled during the execute. During the execute you don't have half of your VS in Cascading Power. You have 0 of them.. So I just changed it too. That's 9 DPS you add that just doesn't exist.

EDIT : That's in fact false.. If you consider 8% alacrity or more, all your VS could be in Cascading Power. This actually raise your DPS by almost 15 DPS.

Edited March 17, 2016 by Ryuku-sama

[Goblin_Lackey]

So you got 6375 DPS on Fury after a fix I did on your math explained below....

 

After making sure everything is fine buff/debuff wise, under the assumption that everyone got at least 8% alacrity, and by using the rotation here :

RB > Ob + Berserk > FS > VS > Rav (clip after the GCD) > BA >
RB > Ob > FS > VS/VT > Sc > FC

 

I got 6359. So clipping Ravage doesn't actually increase DPS... Not once I made sure your work was under the same alacrity assumption as mine.

 

On another note. Your way to handle 2p could easily be reworked for Fury and any other spec with rotation that never change. For Fury, this means your rotation DPS is actually a little higher than what you got.

Without the execute section, which is useless for Fury since their timeUsage on BA never change, you use :

 

duration/(roundup(CD/timeUsage)*timeUsage)

 

I would use :

 

1/roundup(CD/timeUsage)

 

You don't need to time average.. With Fury, you can just determine what is in it and apply the set bonus or not. That's about 7 DPS you're missing here.

 

Another thing I found strange in Fury is how you calculate APM.. Fury abilities that aren't on the GCD are Berserk (once every 19.2s in your math), Frenzy and Bloodthirst. Activating Frenzy outside of bloodthirst is useless for Fury.. It would just mess up the rotation without any real gain.

 

One last thing... Why does Fury have a negative execute damage bonus of -5%/2??? That's almost 1 DPS here

 

I lied... Another thing.. you do know, since you determine DPS before execute time, your way to handle execute variation in buff/debuff uptime isn't perfect? It's a decent approximation, but to get a good one you would have to determine execute time averaged Damage Bonus separately. I couldn,t change that before having to rework your spreadsheet from the ground up.. So I worked under the assumption that for calculating buff and debuff uptime, the execute last 30% of the time.

 

Around the same thing... Your VS and Force Lash bonus damage is badly handled during the execute. During the execute you don't have half of your VS in Cascading Power. You have 0 of them.. So I just changed it too. That's 9 DPS you add that just doesn't exist.

EDIT : That's in fact false.. If you consider 8% alacrity or more, all your VS could be in Cascading Power. This actually raise your DPS by almost 15 DPS.

 

Answering you can be tricky, You have the same bad habit I do of editing in new information and corrections as the thoughts occur.

 

Skipping the post with Ravage Clip before this one, I dislike rotations that can't be sustained indefinatly without outside influnce (excluding tank rotations). Plus that reminds me of the 3.0 Annhilation rotation where there wasn't enough avalible GCDs or Rage to do anything but clip Ravage.

 

 

Point by Point

1) 8% Alacrity Assumtion

The abilities page has no idea what gear is worn. This makes and assumptions of higher alacrity values when determining buffs to be inaccuracte and potentially bias the optimizier/stat cycler toward a lower alacrity value (And thus makeing false outputs). This is really a consequence of the way I organized the spreadsheet. The components of the spreadsheet are meant to bemodular and are also meant to be flexible enough for easy updates. If I had to start fresh again, I would further breakdown each of the abilities so that each buff can be determined how it is applied after alacrity and rotation are calculated.

 

2) two piece Set bonus

I used the duration/(roundup(CD/timeUsage)*timeUsage) becuase that allows me to see the true activation rate for the two set rather than just the ICD and ability cooldown: it would get activated every 38.4s (every other cycle). I also did this becuase it was the same methodology as I used in some of the other Disciplines that couldn't activate the 2 set exactly on cooldown. (most of the time I just did the lazy flat 50% uptime that was the compermise between alacrity and delayed abilities). Again this is the partially the fault of not knowing the alacrity level (do you sense a pattern here yet?)

 

3) APM

You are right in that I just arbitarily said that I was using Berserk on Cooldown for no reason at all. I just tweaked it to set it to 5 minutes (but due to the alacrity issue this will be reduced by alacrity even though Bloodthirst will always be 5 realtime minutes), that worth a reduction of a whole 1/4 of an action. (135s to 300s, meh).

On an a very related note, I charged the apm to use Berserk and Bloodthist, but I never gave the Damage Bonus for it. In fact you will find that I skipped all of the 5 minute raid buffs. The reason for that is actually Sorcerers | Sages. I could never figure out a good way to have the mainstat buff be used without knowning the gear. That and the fact that the buffs are short duration, long cooldown and ignore alacrity meant that rather than try to integrate them, I just skipped all of them.

 

4) Negative Execute Damage

This was my solution for the fact that Vicious Throw replaces Vicious Slash (and Force Lash) in the Raging Burst Window during the execute phase. Rather than my normal method of multiplying the buff by (1- execute %), I put a negative number in the execute bonus damge. Since the Exectute bonus damage is additive with the normal bonus damages when the target is below 30%, it fully cancels out the bonus that I gave it before. In addition this alows me to use the better execute time percentage rather than the 30% estimate (29.5% versus 30%, so not much of one). (hazzah, something that isn't alacrity's fault)

 

5) Execute vs non-execute

None of the ways I determine buff uptimes is truely optimal (dang unknown alacrity). For the most part I always weight the differences by multipliplying the non-exectute by (1-execute%) and adding it to the execute part. I have that set in B2 so that I can change it on the fly (or set it to 1 or 0 to look at the different parts of the rotation). For the outputing of the DPS, I weight the DPS of each section: DPS:1 / (0.7/[non-execute DPS]+0.3/[execute DPS]) but I can't use that number to go back and redetermine buffs (circular references)

 

6) 8% or More Alacrity

Look, another issue with the way I handle alacrity and buff duration! How unexpected.

 

 

For the record, even though I kept saying that it was Alacrity's fault, it is still my favorite Stat by far. Everything discussed just shows that I am actually still undervalueing it in the calculations.

 

 

Edit: This is not an edit.