[Patch 1.0-1.1 ARCHIVE] Mercenary | Commando DPS Compendium

[Patch 1.0-1.1 ARCHIVE] Mercenary | Commando DPS Compendium - Printable Version

+- SWTOR Mechanics Forums (http://mmo-mechanics.com/swtor/forums)
+-- Forum: Class Discussion (/forum-7.html)
+--- Forum: Bounty Hunter and Trooper (/forum-10.html)
+--- Thread: [Patch 1.0-1.1 ARCHIVE] Mercenary | Commando DPS Compendium (/thread-1003.html)

Pages: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38

RE: Arsenal Mercenary | Gunnery Commando DPS Compendium - z3roshot - 01-26-2012 07:52 AM

(01-25-2012 07:07 PM)Granchas Wrote: If you're already high enough on crit and surge (35/85) , does it make more sense to go for >110% accuracy instead of more crit/surge ?

I'm gonna have to agree with Stab, at this point I think power will give you a better return. As far as I'm aware, power doesn't have a cap at all, accuracy has a soft cap where more isn't useful.

RE: Arsenal Mercenary | Gunnery Commando DPS Compendium - danoobsminion - 01-26-2012 08:19 AM

Is that crit soft cap (30%) unbuffed?

How did we arrive at that being our soft cap?

RE: Arsenal Mercenary | Gunnery Commando DPS Compendium - Granchas - 01-26-2012 08:53 AM

Yes, I get that power scales linearly and is clear winner at those kind of situations, but I was wondering about extra accuracy compared to extra crit and extra surge in this scenario. According to the stat description, if you have more accuracy than required to hit a target, it lowers the target's defense. I have no idea how that defense reduction scales though. Also, the fact that Q58 enhancements that give power also give a lot of accuracy makes gearing even more complicated.

As for the crit, I'm assuming we're talking unbuffed since the IA buff actually gives you percentage, not rating. You could factor in the slight crit increase you get from the aim stim I guess. It's not exactly a soft cap, it's just that crit percentage gains get very low relative to crit rating gains around this point so your gear's bonus points are better spent elsewhere .

RE: Arsenal Mercenary | Gunnery Commando DPS Compendium - Gorlough - 01-26-2012 09:28 PM

Actually there are NO soft caps on crit and surge - and especially not on power (as it has a linear progression). I'm kind of guilty for using a wrong term here in the past. The correct term is StatOptimals. Those optimals vary with the skills used and are (iterated) around 38% critchance and 95% critmodifier with Q58 mods and enhancements. After that (and putting Accuracy to 100% for normal attacks), power gets the stat to go for.
Additionally, if augment slots on your gear are present, they're best filled with +aim augments, as the overall benefit is better than putting power into those slots.

RE: Arsenal Mercenary | Gunnery Commando DPS Compendium - Stab - 01-27-2012 03:48 AM

Primary stat for augments 100% of the time. We have no way to gauge additional benefits from higher levels of accuracy. If we could determine Boss XYZ has a 10% dodge rate, 10% parry rate, and whatever possible "defenses" we might be able to put a quality judgement on accuracy past hit cap. A way to test "soft caps" would be throw on some crit gear, see result, throw on some more, see result, throw on even more, see result. If it continues to grant X% crit/surge to Y amount of crit in a linear fashion, then you have not achieved "soft cap" Otherwise, optimal stats come into play where you find the appropriate percentages to maximize dps. Which we can obviously do with the amazing combat log and parsing capabilities at hand...

RE: Arsenal Mercenary | Gunnery Commando DPS Compendium - Tumri - 01-27-2012 07:15 AM

(01-25-2012 12:52 PM)GWARRR Wrote:
(01-25-2012 11:36 AM)Esoteric Wrote: So is that accuracy rating for tech power for for ranged? I am at the 110% tech power but I was curious if I should be looking to get more accuracy.

It's for your special attacks, which have base 100% Accuracy, so yes, you can certainly refer to your Tech Accuracy and aim for 110% for that.

As far as I've heard, even 108% may be optimal as the base defense chance of bosses might actually be 8%, meaning 108% is enough, so you can start looking at other stats to stack. Not confirmed though, EZ and Style are supposed to be checking/confirming soon, so I'd wait for them but for sure at max 110% Accuracy is what you need.

I'll clear up the wording in the original post soon. It should really read "Special Attacks Accuracy" since Unload and Rail Shot should also have base 100% Accuracy as they're classified by the "special attack" flag in the code, so hitting 108% Tech Accuracy will by default also cover Unload and Rail Shot.

Any confirmation yet for the 108% cap? It seems unusual that they would go from 110% to 108%. If it really is an 8% cap is this also the same cap that other classes have(Marauders for example)?

RE: Arsenal Mercenary | Gunnery Commando DPS Compendium - Qed - 01-28-2012 08:50 AM

(Sorry for using Trooper names in a mostly BH thread, I'm used to the former).

I have been thinking about whether it's better to use High Impact Bolt on cooldown, (because it's more efficient than Grav Shot), or on max charges (because then it's even MORE efficient than Grav Shot). Clearly, this is a mathy question, so lets give a mathy answer:

Let's assume we stay in the 0.6 ammo/second regime.
Let's assume that we average over a long time.
Let's assume no long cooldowns (reload, etc).
Let's use simple numbers for damage rather than plugging in the specifics right away:

I propose the following metric: extra Damage per Ammo.

The extra Damage per Ammo of an ability is the amount of damage it would do, minus the amount of damage hammer should would do in the same time. Thus, Full Auto will do, say 5 damage, but takes the place of two hammer shots, which would each do 1 damage. It does 3 "extra" damage for a cost of 2 ammo, or 1.5 eDPA.

Now, we construct a rotation by using up all of our ammo/sec with abilities that have a high eDPA, and then fill in the rest with hammer shot. If we only had independent abilities, this would determine our rotation: Use the highest eDPA ability and then proceed down the line until you are use all of your ammo regen, then spam hammer shot. Unfortunatly for our calculations, Grav Shot alters the rate at which we use Full Auto (ie: our abilities are not independant), and High Impact Bolt forces us to choose between many different versions of the abilitiy with different ammo use and eDPA. The question of on-cooldown vs on-5x-charged-barrel HI shots is now a question about filling up more of our ammo/sec with a lower eDPA ability, vs filling up less of it with a higher eDPA ability.

To the maths:

If we just use HI shot every cooldown, we'll play with the grav shot/sec number until we get about the right ammo usage and then figure out the rest. Our Grav/Sec number is going to change the effective cooldown of Full Auto via Unload. With a 6 second internal cooldown on unload, our Full Auto cooldown will be:

6*.3 + (6+Sec/Grav*.7*.3+6+2*Sec/Grav*.7*.7*.3+...+ (chance of no proc before 15 seconds)*15).

For large Sec/Grav, this number tends to just be 6*.3+(6+Sec/Grav)*.7*.3+.52*15.

As our crit rate goes up, the chance of Full Auto refunding an ammo goes up, which is why this is not a nice round number. We include the fractional refund per use.

Not wanting to solve the system of equations to get the right number for Sec/Grav, we just play with this by hand until our ammo usage is about right.

For the first set, we assume that HI shot is used whenever it is cooled down, regardless of charged barrel

Code:

Ammo/sec    0.6                

Crit Chance    0.3                

Sec/Grav    7                

                Damage  Ammo    Duration        Cooldown        eDPA

Hammer Shot     1       0       1.5             1.5    

HI Shot         3       1       1.5             15              2

Grav Shot       3       2       1.5             1.5             1

Demolition      4       2       1.5             15              1.5

Full Auto       5       1.343   3               11.94           2.233804914

Every HI:                    

Full Auto Ammo Use      0.112479                

Demolition AU           0.133333                

Grav AU                 0.285714                

HI Ammo Use             0.066666                

Ammo Use/Sec            0.598193                

DPS =                   1.540041

(I've included the charged barrel bonus in the ammo use->dps calulation, so it's hidden by this dump.... sorry about that, but it is there)

For our second set, we assume that HI shot is only used when charged barrel is at 5. Again, we tweak the Sec/Grav by hand until the ammo usage comes out right.

Code:

Ammo/sec    0.6                

Crit Chance    0.3                

Sec/Grav    6.266                

                Damage  Ammo    Duration        Cooldown        eDPA

Hammer Shot     1       0       1.5             1.5    

HI Shot         3       1       1.5             15              2

Grav Shot       3       2       1.5             1.5             1

Demolition      4       2       1.5             15              1.5

Full Auto       5       1.343   3               11.80788        2.233804914

Every HI:                    

Full Auto Ammo Use      0.113738                

Demolition AU           0.133333                

Grav AU                 0.319183                

HI Ammo Use             0.031918                

Ammo Use/Sec            0.598172                

DPS =                   1.521391

Looking at total DPS, the answer is....

It really doesn't matter, the difference is in the few percent regime, certainly bigger than the assumptions we've made.

On the other hand, one notes that, on average, we do grav shot surprisingly rarely given how 'spammy' it can feel. With that in mind, a min-max'd rotation might benefit from filling in with grav shot specifically when the internal cooldown on unload is off. This means something like:
Unload
(cooldown starts)
Full Auto
Hi Impact Bolt
Hammer Shots
(cooldown ends)
Grav Shot
Grav Shot
Hammer Shots
Grav Shot

Unfortunately, even if we solved the system of equations and fixed our assumptions above, it would be very hard to model this internal cooldown dependant rotation. I think a truly optimized rotation would need to be tested in simulation rather than spreadsheet.

RE: Arsenal Mercenary | Gunnery Commando DPS Compendium - bural - 01-30-2012 11:23 AM

Hello everyone!
I'm currently working on a Mercenary-Arsenal dps spreadsheet. It is still far from complete and because I will have to focus on getting my master thesis launched during the coming weeks, it will be a while before I release it. In it's current form it's still missing a proper front end, toggle-able raid buffs, pve set bonuses, EP calculations and a calculations for ability sequencing and boss armor calculations just to name a few. I intensionally use "sequencing" and not priority as I'm referring to the chronological order skills are best used in.
With all these admittedly gaping caveats in mind, the sheet does offer some calculations of optimum steady-state DPS and ability prioritation or combination if you will i.e. a rotation that is manageable without extraordinary action such as TSO or Vent Heat. The shortcomings of the sheet will obviously skew the results somewhat and the DPS output only makes sense for comparative comments and not absolute ones i.e. it won't match the output of your character - not yet anyway.

As I don't wish to release the sheet unfinished, I'll describe what the sheet does and ultimately what it suggests for uptimum dps. If I've made a brain-fart somewhere or are working based on a false assuption, I'm hoping you will point it out and correct me so the outcome will be that much better.

The sheet calculates ability damage after stat, armor (any info on boss armor mitigation?) and talent modifiers and uses crit rate to calculate the effective ability heat cost thus factoring in Terminal Velocity. The image below doesn't properly reflect this as it's merely the base damage of an ability (before modifiers) being shown. The sheet does not factor in the internal CD of Terminal Velocity yet (suggestions for implementation are welcome), but either way it is something that would be mitigated by proper ability sequencing. For instance: A TV proc followed by 2x Tracer Missile (back-loaded) followed by another Tracer Missile and then Heatseeking Missile (front loaded) essentially robs you of Terminal Velocity proc chance. Additionally, the lack of Terminal Velocity internal cooldown would inherently improve Tracer Missile spam, something the sheet isn't trying to promote. Alternating between abilities seems to be key.

It should be noted it specifically calculates Rail Shot buff size and hence damage, as ability prioritation does affect this significantly.

Working under the assumption that we're resource limited, one hypothesis can quickly be confirmed: Rail Shot should always be used on cooldown. It offers significantly higher DPHeat than any other skill outside of Barrage Unload. Thus, all rotation-permutations assumes 1 Rail Shot every 15s.

Similarly, all rotations assume Heatseeker should be used on cooldown and atleast 1 Unload per 15 seconds. Rotations without these characteristics has high DPS but are unsustainable due to Heat buildup. This last effect is explicitly calculated below.

Finally, the sheet currently uses a dummy character to calculate stat effects. The dummy has 32% crit, 80% crit mod and doesn't miss (I've intentionally omitted this part due to the observed inconsistencies around this aspect and the need for a proper frontend).

I will spare you the details of the various rotation permutations but simply list them and their characteristics:

1RS/1HS/1Unload/Tracer Spam: 1242 DPS, highest DPS rotation at a glance, gets full size Rail Shots but builds up heat at a rate of 2,34HPS or 210 Heat over 90 seconds - significantly more than a TSO or even Vent Heat can handle. Thus it's not sustainable unless augmented with a Rapid Shots-rotation, causing a loss in dps proportional to the uptime of each rotation.

1RS/1HS/2Unload/Tracer Spam: 1168DPS, only shoots Rail Shots buffed to 4, but gains a substantial amount of energy through Barrage Unloads. Builds up 102 Heat over 90 seconds, which can be managed (bar 3 Heat) through TSO+VH or 1 Rapid Shot per 15 secs.

1RS/1HS/Unload Spam/Tracer Filler: DPS loss compared to the previous permutation due to small size Rail Shots and absence of Barrage procs. Admittedly, it's only a permutation you'd be able to hit following a streak of luck.

1/RS/1HS/1/Unload/1TM/5Rapid Shot: 729 DPS, effectively vents 2.9 HPS making it capable of mitigating the Heat buildup from the permutation with an initial 1242 DPS. One would have to alternate between the two proportionately, giving the combined rotation an average dps of (2.34*1242+2.97*729)/(2.34+2.97)=955DPS.
Consequently we must assume that any cycle that builds up significant heat is incompetitive, nomather its initial DPS.
Perhaps unsurprisingly: when you're resource restricted (even when you have a resource free damage ability), resource efficiency (DPH) is king.

Naturally I'm working on an ever growing TODO list, but outside the previously mentioned caveats I'm happy to take suggestions. Given the nature of gear in swtor compared to "That Other Game"™ I'll be especially happy to receive comments regarding a gear frontend and how to best implement it. To the extend possible, I'd like to stay away from macros so as to keep the sheet assessible to everyone.

[Image: Y45SLh.png]

edit: what's the tag for thumbnails around here? the images size seem to mess with the width of the frame, eating the side of my wall of text. (cheers)
edit2: oi, definite bug in the 5Rapid Shot rotation. Doesn't significantly change the output of the combined, heat-neutral rotation however.

RE: Arsenal Mercenary | Gunnery Commando DPS Compendium - Fearius - 02-08-2012 01:46 AM

Has anyone done an analysis of the DPS gain for the various relics?

For example, is http://www.torhead.com/item/dDStgBn/dark-energy-surge
a dps gain over the matrix cube?

RE: Arsenal Mercenary | Gunnery Commando DPS Compendium - bural - 02-08-2012 03:02 AM

Well, you'd need to compare the benefit of the proc versus +58 Aim and +24 Critical Rating. Some napkin math from the spreadsheet: We're currently hitting on average ones every 1.38 secs. Thus, we can expect a proc every 4.5secs+(1/30%)*1.38=9.13secs. The proc hits for 168 and must be assumed to be affected by the +5% damage buff and the +5% crit buff. Given that the proc deals internal damage, it seems most likely it'd use ranged crit. Additionally, given that it's an "on hit" proc, I doubt damage calculations would need to factor in the hittable a second time. These two assumptions may be false, but without a link to the proc's details on torhead, it's hard to confirm. Consequently, we must assume the proc to hit for 1.05*(168*(1-.346)+168*.346*(1+0.84) )= 227 (at 34.6% crit and 84% surge). This gives you an average of 24.9dps for Dark Energy Surge whereas +58Aim and +24Crit should net you something like 33dps according to the spreadsheet. Secondly it's worth noticing passive stat bonuses anecdotically tends to perform better than procs of equal estimated value - atleast in "that other game"™.
Considering the way the game handles certain other buffs (or fails to handle) like TSO, it's possible you could get back-to-back procs of a back-loaded and front-loaded ability combination or even multible procs from DfA impacts. Testing would be much appreciated.

I can see how the trinket would be good for some more frequently hitting classes, but given that the proc is percentage-based, it's simply not very appealing to hardcaster type damagedealer that is Arsenal/Gunnery. It's possible it'd fair better if specced in Pyrotech, but that's another story.

I will include support for trinkets at some point, but have postponed it thus far due to the lack of effect details for relic procs. If you can supply the this info, support for trinkets could be included that much earlier.