Jump to content

Optimal Stats For All 24 Disciplines, KOTFE Edition


Goblin_Lackey

Recommended Posts

Man, I really wish trying to play Watchman didn't want to make me gouge my eyes out. Good to see that Sharpshooter got an indirect buff though. Maybe I'll try out Tactics.

 

And that Scoundrel healing discrepancy. Damn.

Edited by Bugattiboy
Link to comment
Share on other sites

  • Replies 896
  • Created
  • Last Reply

Top Posters In This Topic

I can't for the life of me reach 5K DPS on the dummy with my Telekinetics Sage in 216 gear, let alone 6k. I like to think that I execute the rotation just fine since it's easy and I've been playing that class since 2011, but this thread mentions 6292 as average DPS for that class with 216 gear. I even have a bunch of 220 pieces on top of all the other 216 ones, but the best I ever get on the dummy on 4-5 minute parses is 4700 DPS.

 

I have Critical in all my augments except for 3 Alacrity ones. My Accuracy is at 110%.

 

Am I doing something terribly wrong or do I just not understand what average DPS means in the OP's calculations? I suspect it's the latter. If OP's average numbers have nothing to do with dummy parsing, can anyone tell me if 4700 DPS is alright for that class in that gear?

Edited by Birtram
Link to comment
Share on other sites

I can't for the life of me reach 5K DPS on the dummy with my Telekinetics Sage in 216 gear, let alone 6k. I like to think that I execute the rotation just fine since it's easy and I've been playing that class since 2011, but this thread mentions 6292 as average DPS for that class with 216 gear. I even have a bunch of 220 pieces on top of all the other 216 ones, but the best I ever get on the dummy on 4-5 minute parses is 4700 DPS.

 

I have Critical in all my augments except for 3 Alacrity ones. My Accuracy is at 110%.

 

Am I doing something terribly wrong or do I just not understand what average DPS means in the OP's calculations? I suspect it's the latter. If OP's average numbers have nothing to do with dummy parsing, can anyone tell me if 4700 DPS is alright for that class in that gear?

 

Perchance you have com gear? Its low power trash, I would not be surprised if the tertiary stats don't do anything on that as there is nothing to scale from. The stat guide and dps estimations are for token gear.

Link to comment
Share on other sites

Perchance you have com gear? Its low power trash, I would not be surprised if the tertiary stats don't do anything on that as there is nothing to scale from. The stat guide and dps estimations are for token gear.

 

Ah, that's it of course. I forgot about that, I do have mostly comm gear. I just haven't had many chances to raid with my Sage for now since I often need to log in with my tank for guild runs. I am aware that comm gear is always trash compared to token gear, but I wasn't expecting such a huge difference. Now that I look into it carefully I see that the difference in Power is indeed huge.

 

I gotta say that I was still hoping to hit 5k DPS without a full bonus set, however. Oh well, looks like Sages aren't too hot right now.

Edited by Birtram
Link to comment
Share on other sites

I can't for the life of me reach 5K DPS on the dummy with my Telekinetics Sage in 216 gear, let alone 6k. I like to think that I execute the rotation just fine since it's easy and I've been playing that class since 2011, but this thread mentions 6292 as average DPS for that class with 216 gear. I even have a bunch of 220 pieces on top of all the other 216 ones, but the best I ever get on the dummy on 4-5 minute parses is 4700 DPS.

 

I have Critical in all my augments except for 3 Alacrity ones. My Accuracy is at 110%.

 

Am I doing something terribly wrong or do I just not understand what average DPS means in the OP's calculations? I suspect it's the latter. If OP's average numbers have nothing to do with dummy parsing, can anyone tell me if 4700 DPS is alright for that class in that gear?

 

Yea mate you token gear, the 216/220 comm is incredibly low on power.

 

Put it this way its so low on power that 192 token has more power. Not saying its better just that comm gear is very unoptimized.

Edited by benficakungfu
Link to comment
Share on other sites

What implants and hears should be used with these stats, am I missing something?

 

For the purposes of this analysis, the implants and earpieces are effectively Enhancements. They share the same stat pool for tertiary stats as Enhancements; thus, you can think about 10 different Enhancement slots. (The other stats on Implants & Earpieces - Mastery, Power, Endurance - are the same regardless of which one you select within the same tier.)

Link to comment
Share on other sites

Any chance you could run some numbers at the 224 level with 5 accuracy augs + 2 enhancements enforced? Even if it isn't optimal, I prefer 110% accuracy.

 

Appreciate the time you spent on this.

 

Me too, actually. Especially on dual wielding classes I've always rather surpassed 100% accuracy than undercut it, since OH misses are a problem anyway and contrary to common belief accuracy above 100% is not totally wasted.

 

So I'd be very much interested how builds with slightly more than 100% accuracy compare mathematically (especially for dual wielding specs), since it's what I will probably be doing "in the field" anyway :-)

 

Apart from that: Thanks a ton for your efforts, very helpful!!

Edited by Ardarell_Solo
Link to comment
Share on other sites

Me too, actually. Especially on dual wielding classes I've always rather surpassed 100% accuracy than undercut it, since OH misses are a problem anyway and contrary to common belief accuracy above 100% is not totally wasted.

 

So I'd be very much interested how builds with slightly more than 100% accuracy compare mathematically (especially for dual wielding specs), since it's what I will probably be doing "in the field" anyway :-)

 

Apart from that: Thanks a ton for your efforts, very helpful!!

 

See post #46 where I put the full datset for it inside the spoiler tag. I just did the stats, (which round the dps output) and didnt post the rankings.

(I over use that tag but is is useful for isolating specific data.)

 

-sent via the Bantphone

Edited by Goblin_Lackey
Link to comment
Share on other sites

Hey folks! Kamikaze here, thought I'd chime in since this topic is getting a lot of interest on my server and sparking discussions in guild. You perhaps may have run into me as a fellow tank theorycrafter over at http://www.kamikazetank.com (yes I'm still working on the 4.0 updates sorry), or for commando assault spec work done elsewhere. Anyway, I applaud the work and effort that has gone in to this so far, many may not appreciate the investment required when trying to do some of this work!

 

Now, to play devil's advocate I think the crit over power concept for 4.0 is being way oversold for what it really is. There is a fundamental kink in the idea here which I have also been playing with in light of the new itemization we have. Mainly, it's that crit rate will ultimately be subject to the dice roll, and an offline system that computes it such that we're assuming all your abilities will actually crit at your fixed crit rate is not being realistic. Moreover, the greater your investment into a set that relies on the randomness of "dice rolls" the higher your actual variance will be in short burst moments.

 

After all that said, in the end here we're ultimately comparing a power vs crit set where the actual difference in output between the two is around 1% (6474.14 vs 6404.20) in a your well controlled system, so this isn't exactly game changing even if the system was realistic such that we'd always find us parsing our abilities at our given crit rate, not more or less.

 

This brings me to a little experiment I conducted for my guildies, a real world test between the two sets to illustrate it's a wash in the long run. To start, my base gear is a partially optimized set in mostly 220s, with two 198 relics and a few 216s. All four class buffs are applied along with an Advanced Anodyne Versatile Stim. I then change out the augments/crystals to create two versions, one favoring power and one crit.

 

Power Set (14 Versatile Augments, 2 Hawkeye Crystals)

R/T Damage: 1887.4/2654.9 (2276 power, 3178 tech power)

R/T Critical: 39.81/34.81 (769 crit rating)

Surge: 62.83 (769 crit rating)

Accuracy: 111.00 (777 accuracy rating)

Alacrity: 0.00 (0 alacrity rating)

 

Crit Set (14 Critical Augments, 2 Eviscerating Crystals)

R/T Damage: 1642.3/2409.8 (2194 power, 3178 tech power)

R/T Critical: 47.80/42.80 (1873 crit rating)

Surge: 72.15 (1873 crit rating)

Accuracy: 111.00 (777 accuracy rating)

Alacrity: 0.00 (0 alacrity rating)

 

For each set, I parsed against an ops training dummy (with armor reduction applied but not health to support very long parses) for 200 passes on a 10 GCD basic commando assault spec rotation. The rotation is executed by a macro so there is no human error and identical every time. I use this rotation merely for comparative analysis purposes, it is not representative of max potential for a gear set at all (obviously), as that would require a human executed rotation with openers, resource dump phases, sub 30% execution actions, and various interweave tweaks with utilities and cooldowns. All that said, it's perfectly good tool for comparing two sets in the real world like we are here. The only uncontrolled variables would be your actual RNG on crit, relic procs, and the variable min/max damage by attack. I try to compensate for the RNG by doing a fairly long parse which is nearly an hour long (so just before the class buffs fall off), and doing this all many times to get good data and averages.

 

Parse #1: Power Set = 4838 DPS

http://imgur.com/6xoGAZV

 

Parse #2: Crit Set = 4843 DPS

http://imgur.com/p3j67q7

 

The average output between these two sets yielded pretty much identical numbers. When looking at all the power parses versus crit parses, the power ones had a much more narrow spectrum of difference (not unexpected). If you want to rely on consistent output, you may be better off with the power, since you're reducing the reliance on randomness to get more crits of less strength. In the end, neither set choice is going to make or break anyone's raiding, you'll have a slightly better chance at getting a lucky parse with the crit set, on average you'll probably be the same, and you have a higher risk at getting a low burst moment when the crit rolls just fail you without the power to back it up.

 

If you think I missed something let me know, I'm merely lending additional data and perspective to help us all gear best. This experiment attempted to apply this crit vs power theory in a realistic manner that pulls the situation into perspective and what may actually be better in the field, versus what a fixed mathematical system may suggest would net out.

 

Cheers and keep up the good work!

Edited by KamikazeKommando
Link to comment
Share on other sites

Power Set (14 Versatile Augments, 2 Hawkeye Crystals)

R/T Damage: 1887.4/2654.9 (2276 power, 3178 tech power)

R/T Critical: 39.81/34.81 (769 crit rating)

Surge: 62.83 (769 crit rating)

Accuracy: 111.00 (777 accuracy rating)

Alacrity: 0.00 (0 alacrity rating)

 

Crit Set (14 Critical Augments, 2 Eviscerating Crystals)

R/T Damage: 1642.3/2409.8 (2194 power, 3178 tech power)

R/T Critical: 47.80/42.80 (1873 crit rating)

Surge: 72.15 (1873 crit rating)

Accuracy: 111.00 (777 accuracy rating)

Alacrity: 0.00 (0 alacrity rating)

 

Your test is skewed in favor of the Power set because the Crit rating you're using is far beyond the point where Crit DR makes Power or even Mastery a better stat. It's not just a choice of Crit vs. Power or Mastery, it's choosing tertiary stats (both Crit AND Alacrity in addition to Accuracy) over Power or Mastery. Not that the percentages we're talking about are big enough to be seen in a parse or make a real difference in a raid, but there is still a difference. I would be interested in your test run with ~1100 Crit and ~700 Alacrity vs. your Power set.

Link to comment
Share on other sites

Your test is skewed in favor of the Power set because the Crit rating you're using is far beyond the point where Crit DR makes Power or even Mastery a better stat. It's not just a choice of Crit vs. Power or Mastery, it's choosing tertiary stats (both Crit AND Alacrity in addition to Accuracy) over Power or Mastery. Not that the percentages we're talking about are big enough to be seen in a parse or make a real difference in a raid, but there is still a difference. I would be interested in your test run with ~1100 Crit and ~700 Alacrity vs. your Power set.

 

You talk about the crit DR. So is the 1100 crit and 700 Alacrity the DR for both ? Or just Crit

Link to comment
Share on other sites

Your test is skewed in favor of the Power set because the Crit rating you're using is far beyond the point where Crit DR makes Power or even Mastery a better stat. It's not just a choice of Crit vs. Power or Mastery, it's choosing tertiary stats (both Crit AND Alacrity in addition to Accuracy) over Power or Mastery. Not that the percentages we're talking about are big enough to be seen in a parse or make a real difference in a raid, but there is still a difference. I would be interested in your test run with ~1100 Crit and ~700 Alacrity vs. your Power set.

 

Ok but the problem with alacrity on an assault spec commando specifically is the reduction in GCDs will misalign the synergy of a [ionic Accelerator] procs since the 7.5s ICD of it will cause your abilities to be ready and be triggered when there is still a slight CD on it, meaning, you won't be able to get two IA procs within the 15s base rotation. That's one reason they removed alacrity from the spec since it was really messing with us, and its also the reason many have a hard time getting top assault parses (key is those IA proc synergies at the mid and end point of the rotation). I can't speak to other classes, but this is one spec that doesn't want to mess much with it, so when you don't have another choice after accuracy then the crit vs power debate does hit into the DR curves quite a bit.

 

I could steal some alacrity enhancements from my heal set and run a parse tomorrow with ~1100/700 just for giggles. Yet, the macro is still going to execute the rotation assuming the GCDs are 1.5s, so that will need to be tweaked accordingly. :)

Edited by KamikazeKommando
Link to comment
Share on other sites

You talk about the crit DR. So is the 1100 crit and 700 Alacrity the DR for both ? Or just Crit

 

DR starts from a crit rating of 1 for Alacrity and Crit. I'm not sure of the exact point where the DR from Crit rating results in Power or Mastery giving you more DPS, but I'm pretty confident 1870 is too high by just looking at what numbers you have seen in the OP of this thread. The numbers I chose were what I assume is optimal given the stat budget that existed in the test based on what Bant has put forward as optimal in his simulations. It was just a guess to try and get some better test results since my numbers are definitely closer to the optimal stat value than 0 Alacrity and 1870 Crit.

Link to comment
Share on other sites

Ok but the problem with alacrity on an assault spec commando specifically is the reduction in GCDs will misalign the synergy of a [ionic Accelerator] procs since the 7.5s ICD of it will cause your abilities to be ready and be triggered when there is still a slight CD on it, meaning, you won't be able to get two IA procs within the 15s base rotation. That's one reason they removed alacrity from the spec since it was really messing with us, and its also the reason many have a hard time getting top assault parses (key is those IA proc synergies at the mid and end point of the rotation). I can't speak to other classes, but this is one spec that doesn't want to mess much with it, so when you don't have another choice after accuracy then the crit vs power debate does hit into the DR curves quite a bit.

 

No. Alacrity also reduces ICD. So your IA ICD will always finish 5 GCDs after being procced.

Link to comment
Share on other sites

Thank you for all the great info.

 

Noob question: Until I can get the new set-bonus gear, I am wondering what is best to use in the mean time? I have a mix of 192/198 set-bonus armorings from 3.0, should I use those armorings with new 216 mods/enhancements until I can get 216 set bonus armorings? Or should I just use the non-set 216 armorings from the comm vendor or random fp/ops drops? If someone has already done the analysis on this, it would save me a lot of time and money. :) Thanks

Link to comment
Share on other sites

Thank you for all the great info.

 

Noob question: Until I can get the new set-bonus gear, I am wondering what is best to use in the mean time? I have a mix of 192/198 set-bonus armorings from 3.0, should I use those armorings with new 216 mods/enhancements until I can get 216 set bonus armorings? Or should I just use the non-set 216 armorings from the comm vendor or random fp/ops drops? If someone has already done the analysis on this, it would save me a lot of time and money. :) Thanks

 

Definately keep the setbonus, alternatively you can use pvp armorings if you have comms, those do stack with the 216 ones.

 

But yeah the setbonus adds alot of dps, as an example on a sniper, the 6th bonus basically gives you an extra autocrit which are amazing now. So yeah keep setbonus i'd say. Others will probably back me up

Link to comment
Share on other sites

Hey folks! Kamikaze here, thought I'd chime in since this topic is getting a lot of interest on my server and sparking discussions in guild. You perhaps may have run into me as a fellow tank theorycrafter over at http://www.kamikazetank.com (yes I'm still working on the 4.0 updates sorry), or for commando assault spec work done elsewhere. Anyway, I applaud the work and effort that has gone in to this so far, many may not appreciate the investment required when trying to do some of this work!

 

Always good to have more people looking into this double checking people's work and assumuptions. I have blind spots as much as anyone else.

 

Now, to play devil's advocate I think the crit over power concept for 4.0 is being way oversold for what it really is. There is a fundamental kink in the idea here which I have also been playing with in light of the new itemization we have. Mainly, it's that crit rate will ultimately be subject to the dice roll, and an offline system that computes it such that we're assuming all your abilities will actually crit at your fixed crit rate is not being realistic. Moreover, the greater your investment into a set that relies on the randomness of "dice rolls" the higher your actual variance will be in short burst moments.

 

I agree with your statement that critical is being overemphasized, but not for the reason that you stated. The "Dice roll" of critical still is a major factor in the damage output from power. Increasing the probability of a critical hit actually has an effect of Decreasing the damage variance because of the higher chance of occurrence. Over a long enough period of time, the number of critical hits will trend toward the mathematical mean.

 

But the key word from your earlier statement is "short burst moments" which I agree that luck plays a far out-sized role and it would be foolish to rely just on luck to accomplish a goal.

 

More on the reason why I think critical is being overemphasized below:

 

After all that said, in the end here we're ultimately comparing a power vs crit set where the actual difference in output between the two is around 1% (6474.14 vs 6404.20) in a your well controlled system, so this isn't exactly game changing even if the system was realistic such that we'd always find us parsing our abilities at our given crit rate, not more or less.

 

This brings me to a little experiment I conducted for my guildies, a real world test between the two sets to illustrate it's a wash in the long run. To start, my base gear is a partially optimized set in mostly 220s, with two 198 relics and a few 216s. All four class buffs are applied along with an Advanced Anodyne Versatile Stim. I then change out the augments/crystals to create two versions, one favoring power and one crit.

 

Power Set (14 Versatile Augments, 2 Hawkeye Crystals)

R/T Damage: 1887.4/2654.9 (2276 power, 3178 tech power)

R/T Critical: 39.81/34.81 (769 crit rating)

Surge: 62.83 (769 crit rating)

Accuracy: 111.00 (777 accuracy rating)

Alacrity: 0.00 (0 alacrity rating)

 

Crit Set (14 Critical Augments, 2 Eviscerating Crystals)

R/T Damage: 1642.3/2409.8 (2194 power, 3178 tech power)

R/T Critical: 47.80/42.80 (1873 crit rating)

Surge: 72.15 (1873 crit rating)

Accuracy: 111.00 (777 accuracy rating)

Alacrity: 0.00 (0 alacrity rating)

 

For each set, I parsed against an ops training dummy (with armor reduction applied but not health to support very long parses) for 200 passes on a 10 GCD basic commando assault spec rotation. The rotation is executed by a macro so there is no human error and identical every time. I use this rotation merely for comparative analysis purposes, it is not representative of max potential for a gear set at all (obviously), as that would require a human executed rotation with openers, resource dump phases, sub 30% execution actions, and various interweave tweaks with utilities and cooldowns. All that said, it's perfectly good tool for comparing two sets in the real world like we are here. The only uncontrolled variables would be your actual RNG on crit, relic procs, and the variable min/max damage by attack. I try to compensate for the RNG by doing a fairly long parse which is nearly an hour long (so just before the class buffs fall off), and doing this all many times to get good data and averages.

 

Parse #1: Power Set = 4838 DPS

http://imgur.com/6xoGAZV

 

Parse #2: Crit Set = 4843 DPS

http://imgur.com/p3j67q7

 

The average output between these two sets yielded pretty much identical numbers. When looking at all the power parses versus crit parses, the power ones had a much more narrow spectrum of difference (not unexpected). If you want to rely on consistent output, you may be better off with the power, since you're reducing the reliance on randomness to get more crits of less strength. In the end, neither set choice is going to make or break anyone's raiding, you'll have a slightly better chance at getting a lucky parse with the crit set, on average you'll probably be the same, and you have a higher risk at getting a low burst moment when the crit rolls just fail you without the power to back it up.

 

If you think I missed something let me know, I'm merely lending additional data and perspective to help us all gear best. This experiment attempted to apply this crit vs power theory in a realistic manner that pulls the situation into perspective and what may actually be better in the field, versus what a fixed mathematical system may suggest would net out.

 

Cheers and keep up the good work!

 

The biggest problem is this statement: "power vs crit"

It is actually: "power vs crit vs alacrity"

 

Your setup is actually very similar to Jinre_the_jedi's (Kre'a): http://www.swtor.com/community/showthread.php?t=846803

And you both have the same blind spot: Alacrity.

 

 

If you look at the analysis I did here (Post #1 & Post #41): http://www.swtor.com/community/showthread.php?t=844210

In this I started with a base stat set and from there I looked at what the effects would be from adding a different augment type one at a time. I utilized my dps calculator so that I could get a simple analysis of what the effect would be each time I added an augment.

I considered four augments each time: Mastery, Power, Critical and Alacrity and choose the one that provided the largest boost.

 

If you look through the changes from each time, you will find that every time I put in an augment into Critical or Alacrity it actually reduced the effectiveness of adding additional points to that stat. As the stat point totals get higher and higher, the step gain from using Mastery and Power could be larger than the step point gains from additional Critical and Alacrity. If we had more than 14 augments and 2 crystals to optimize with then Mastery or Power augments would be the best choice. But with our current statpools that point is reachable, IF the stats are balanced.

 

Sidenote: The diminishing returns curve is present from the very first point for all of the tertiary stats (and the crit bonus from Mastery), there is no magic point where it starts to take effect. It is always a balancing act where putting more points in is not as efficient as putting points into a different stat.

 

This brings me to the blindspot: Alacrity. The problem with your current statbudget is that it does not follow the balancing act that is present in the tertiary stats. All 3 of the dps tertiary stats provide multiplictive bonuses. Accuracy affects the hit chance roll. Critical affects the critical chance roll. Alacrity affects total number of Abilities used. By ignoring Alacrity, this causes my statement of not using Power or Mastery augments to become incorrect because you are starting from a much higher base of critical compared to your Power and Mastery Budgets. This is what causes Power (and Mastery) augments to provide larger results.

 

Alacrity is the most consistent stat. It provides a constant and identical boost to effective APM by reducing the GCD, ability cooldowns, cast times, Internal cooldowns, dot duration, dot tick rate and boosts energy regen to exactly match all of the other changes. This is not effected by any rolls or luck, it is a constant boost to APM which directly translates to DPS. Having 1% Alacrity allows your rotation to be completed 1% faster. Also, if your goal is to reduce damage variance, Alacrity is the stat for you.

 

The problem with Alacrity is not the Alacrity from gear. It is the short duration alacrity boosts (I like to call them variable alacrity) that are harmful to rotations (ie Vent Heat). This is the only way (besides errors and misses) that ICD would no longer line up. This fully applies to Assault Spec Commandos, I have known quite a few that ran 700+ Alacrity during 3.0 with no harmful effects (Besides timing Vent Heat to avoid potential issues)

 

 

The reason I think Critical is being overemphasized is because Alacrity is just as important and also provides a similar multiplicative boost.

Link to comment
Share on other sites

EDIT: I WAS WRONG. Disregard my comments about the ability coefficients for certain Saboteur abilities. Whoops!

 

Hey Bant - great work as always. :) I'm working through my own Gunslinger model as we speak, but wanted to chime in to correct a couple of the Formula variables you're using. Like you, I got my AMP, Base Min, Base Max, and Bonus Coefficient values, but have tested those against in-game tooltips and found a couple of them to be inaccurate.

 

Here are the correct values:

 

https://docs.google.com/spreadsheets/d/1eD2JI0xJLNYqro3NcbXMlnRBf289XC5JZvykLGOntnU/edit?usp=sharing

 

You can validate these in game - just write out the formula and insert your Tech Damage Bonus for the Tech Abilities, and insert your Weapon Damage and Ranged Damage Bonus for Flurry of Bolts. Compare that to the Tooltip value and it should match.

 

Hope that helps!

 

EDIT: Hey Bant, if you read this, can you tell me what's going into Mastery to arrive at 5,372 Mastery? When I look at all the gear + Stims I see 4,218 Mastery from gear plus 710 base Mastery = 4,918 Mastery * 1.05% from Force Valor = 5,174. What am I missing here?

 

https://docs.google.com/spreadsheets/d/1pNrRgEfC-eZbouKusLddkWXgzDrk5aKomcj75TsbR5c/edit?usp=sharing

Edited by Frettz
Link to comment
Share on other sites

Alright, either I'm stupid or unable to comprehend this because it's a bunch of numbers and numbers confuse the heck out of me but...

 

I don't see any information here about what enhancement combinations you used with these augments to generate these results? I don't really feel like going through 7 enhancements and getting one each of crit and alacrity to figure it out - expensive and time consuming.

 

Would it be possible to get a FULL breakdown instead of just augments, perhaps written in clear English for those of us who are numerically-challenged? I'm an author, not a mathematician.

Link to comment
Share on other sites

×
×
  • Create New...