Arsenal DPS spreadsheet

[Sarkastro]

While it doesn't cover all BH/Trooper issues, http://www.swtor.com/community/showthread.php?t=190674 is the best resource I've found for a quick list of current bugs/issues.

[GWARRR]

This looks great. I've finally stumbled on some time to do stuff so I can tinker with the table and see what happens. I may link to this thread/spreadsheet in the Compendium, if you don't mind at all. This will definitely allow for complete optimization. Guess I should start looking at that BiS list I've been meaning to compile...

[Rakunvar]

Not sure if you've gotten back online yet or not but figured I should let ya know ahead of time on this.

PTS Note for new build recently. Surge rating has been re-balanced. It now reaches diminishing returns the same way as other damage ratings, and its per point damage contribution has been reduced by approximately 10%.

[bural]

(02-11-2012 03:58 AM)Rakunvar Wrote:  Not sure if you've gotten back online yet or not but figured I should let ya know ahead of time on this.

PTS Note for new build recently. Surge rating has been re-balanced. It now reaches diminishing returns the same way as other damage ratings, and its per point damage contribution has been reduced by approximately 10%.

I have taken notes of your comments, they're most appreciated and I will adress them as soon as I find the time. I've just had a lot on my plate these past few days and will unfortunately have so in weekend as well. Furthermore it's damn frustrating not being able to test bugs ingame myself...
I've seen the patch notes. I'll make an option to toggle between the current and future surge calculations. Naturally it'll reduce the value of surge and crit. This should make the advice mentioned in the DPS compendium about focusing on accuracy before crit and surge pretty much universally correct.

(02-10-2012 03:44 AM)Rakunvar Wrote:  I'm assuming I am understanding the EP+1 correctly as they are just weights basically, but I wanted to confirm. If this is so then my questions below still follow.

Is there something wrong with the Diminishing returns on crit. now? Seems it has a very high value that barely moves and between 360-370 it drops from 1.00 weight to .56ish.

Your understanding of EP+1 is correct. There'll be a DPS/stat in the version as well. I could make a cross table to show how much of each stat can be traded for 1 of another, but frankly it wouldn't be very helpfull. Do notice the complications discussed in this thread on stat weights.
The sudden drop in the value of crit rating is noteworthy. However, I don't believe there's anything wrong with the calculations per se. The implementation of Terminal Velocity (TV) is still a work in progress and the drop you see relates to this implementation. As previously mentioned, I calculate the average time between crits from abilities capable of proccing TV. The number of procs over a given time period is used as the number of TV procs. However, the calculated average has a minimum value of 3 to reflect the internal cooldown. The word "softcap" comes to mind, but do notice it's one that is mainly introduced by the sheet. It exist ingame as well, but the dropoff isn't as abrupt as currently depicted by the sheet. It appears it kicks in around 356 crit rating depending on your level of alacrity and aim. See cell CalcsV2!B216.
In sum, there's nothing wrong per se and it doesn't relate to the diminishing returns, but the TV model could be improved. I find it surprising we're hitting "the TV wall" this early on atleast. I'm all ears and open to suggestions if anyone have a good suggestion as to how to incorporate ability sequencing and its effect on the internal cooldown of TV.

(02-10-2012 03:44 AM)Rakunvar Wrote:  Also is alacrity modeled right and such? It doesnt seem to affect "calculated DPS" when its changed for me for some reason. Also shows kinda strange as #.##E-## could be intentional as well.. #=w/e numbers. I "Thought" the more you messed with the 1.5 on tracer etc with alacrity the more you just hurt yourself with heat and such is also why I ask.

I believe the alacrity model is pretty much spot on. The cells uses the "scientific" formatting of numbers. E-05 means *10^(-5) so 3.65E-05 = 0.0000365. It's to save space on a frankly speaking horrible dps stat. I've outlined some if the issues with alacrity in a previous post, but in addition in worth noticing some levels of alacrity may be decrimental to TV heat venting. This isn't modeled by the sheet however.

(02-10-2012 03:44 AM)Rakunvar Wrote:  Is Critical Reaction from Bodyguard supposed to only show a 2.5 increase when the talent has a 50% chance to proc but a 5% boost to alacrity @ 1 point and also @ 6.50 for 2 points. These numbers are in the K-38/39.

I believe it's implemented as 2.5% per point, which admittedly is incorrect when only speccing 1 point in it. However, I actually have the calculations for number crits per second and by inverting this I should be able to calculate uptime of Critical Reaction and use it as a proper implementation of the talent. You can manually edit it to 5% for 1 talent point if you want. For sustained dps purposes there appears to be little reason for speccing more than 1 point in it. Atleast, you see far larger returns from the first talent point than from the second point.

(02-11-2012 02:51 AM)GWARRR Wrote:  This looks great. I've finally stumbled on some time to do stuff so I can tinker with the table and see what happens. I may link to this thread/spreadsheet in the Compendium, if you don't mind at all. This will definitely allow for complete optimization. Guess I should start looking at that BiS list I've been meaning to compile...

Please, be my guest Naturally, I don't mind at all, the more the merrier etc. Do link to the thread though. With the issues still haunting the sheet, I feel it's paramount people read the currently known issues before using it - a list I'll make sure to keep as comprehensive as possible. While I'm fairly satisfied with the way the sheet models our way of doing damage, there are definitely still bugs in there that skews the results to a significant degree.

I found a exhaustive list of names for mirror abilities and talents earlier today and will use as a basis for a Trooper port. The previously discussed issues in this regard will still need to be handled somehow. It's a side project as I want to improve and verify the validity of the BH model first.

I appreciate the link to the issues with Trooper mechanics, but once again it would be a great help if a list of known BH and Trooper issues could be compiled and maintained. Given BWs trackrecord this past month, discrepancies between the model and ingame behavior may as much be due to ingame bugs as bugs in the sheet.
Once the major bugs have been sorted out, I will be cleaning up the ability tables and then move to implement various trinkets. Another feature that could be interesting to have in the HUD is modifications suggestions based on current EP weights. Not sure if the surge change makes this feature useless though.

Almost universally, every mark 25 enhancement should have 37 power and 51 to a secondary and every mod should have 37 power 48 primary, or else you're probably doing it wrong.

[Crythe]

About that "high resource regen area" being 33% for Trooper vs 40% for BH...I'm not so sure anymore if it's correct, could just be the UI rounding numbers weirdly. The ammo bar shows 0,6 regen at 8 bars and 0,36 at 7, if it really is 40% then transition point should be at 7,2 bars. Someone in official forums mentioned there being a dev post saying that both heat and ammo work identically, just the UI shows them differently. No luck finding the source though. That would also mean that some of the trooper abilities that appear to cost 0,67 or 0,33 more when converted to heat are just rounding errors or taken into account in some other way.

I've been trying to fingure out how exactly does ammo work by Frapsing and counting usage and regen ticks but haven't found anything conclusive. Ticks aren't even showing every second, sometimes there are 2 or 3 sec caps where there isn't enough accumulated to be shown as +1 ammo. Example:
75% to 83% to 91% to 100%
----2 sec----2 sec----1 sec----
3 cells generated in 5 sec making the overall rate being 5% or 0,6 cells per sec.

Does anyone have an idea how to test the if
1) highest regen area lasts for 40% not 33% as UI suggests
2) 2 ammo cost abilites actually cost 1,92
3) there is some other mechanic in place making Trooper & BH resource systems function identically

Just something I thought we should figure out before bural starts working on trooper version of spreadsheet.

[Rakunvar]

Thanks for all the in depth info and help on those questions! Keep up the awesome work on all of this.

Sorry again if the questions seemed a bit stupid. Like I was saying, haven't played with a lot of this stuff in quite a long time. Either way, thanks again!

[bural]

Regarding Unload:
I've fixed some of the issues, but I simply can't see why Unload is doing lower damage than it does ingame. I have a feeling something really basic is escaping me, so I'm hoping one of you guys would take a look at the Ability DMG of Unload in the CalcsV2 tab. For instance, does it use the correct weapon damage or have I forgot to factor in something stupidly simple? It would make sense the bug related to weapon damage given that it's apparently affecting both Rapid Shots and Unload (and presumably Rail Shot).

Regarding Trooper regen:
Nice work. Would be soooo much easier if the regen rates and intervals turned out to be the same after all. I noticed your version of TV also gives a indicates a slightly different rate.

As 1) and 2) keeps you from testing specific ammo regen intervals, I'd suggest emptying your bar completely and see how long it takes to regen. If the systems works identically and it's UI rounding issues, it should take equally long time to regenerate a full bar. So, use that as a working hypothesis, fill up your bar and see how long it takes to recharge. Make sure to use a backloaded ability as the last ability. If it works the same way as Bounty Hunters, you'll have 3 intervals of 20% and one of 40% where regen is 2,3,4,5% per second respectively, giving you a total regen time of 29.67 seconds. If the 5% interval goes to 33% and not 40%, it should take 29.92 seconds to fully recharge. You're gonna want at least 30fps for the test and the more observations the better although there randomness shouldn't play too large a factor. If the regen rate is higher, it should be easy enough to see.
This should cover the test of 1).

Let's cover 2) and 3) once 1) sorted.

edit: forget what I wrote about trooper regen, it's a brainfart.
Focus instead on http://ge.tt/8HmsRaD and these tests:
Hypothesis1: Trooper regen is different to BH and ability cost is the same 16%
Support if regen time after 10 abilities is 22,6secs
Hypothesis2: Trooper regen is different to BH and ability cost is 2ammo
Support if regen time after 10 abilities is 27,16 secs
Hypothesis3: Trooper regen is identical to BH and ability cost is the same 16%
Support if regen time after 10 abilities is 22,42secs
Hypothesis4: Trooper regen is identical to BH and ability cost is 2ammo
Support if regen time after 10 abilities is 25,42 secs

[bural]

With the help of Crythe, I believe I've located the problem causing various abilities to do incorrect damage. There are certain discrepancies between Trooper and Bounty Hunter abilities that aren't warranted by dualwielding or higher weapon damage.

First, when comparing Demolition Round and Heatseeker Missiles we have the same effect details for effect 1.2 (TravelSpeed is for some reason different but nvm):

Code:

SpellDamage: SpellType=>Tech, StandardHealthPercentMin=>0.17, Slot=>None, Coefficient=>2, StandardHealthPercentMax=>0.23, DamageType=>Kinetic
SpellDamage: SpellType=>Tech, StandardHealthPercentMin=>0.17, Slot=>None, Coefficient=>2, StandardHealthPercentMax=>0.23, DamageType=>Kinetic

Now compare this to Unload / Full auto

Code:

WeaponDamage: Slots=>[ SecondaryRanged ], FlurryBlowsMax=>1, AmountModifierPercent=>-0.3, FlurryBlowsMin=>1, IsSpecialAbility=>1
WeaponDamage: StandardHealthPercentMin=>0.0263, Slots=>[ PrimaryRanged ], StandardHealthPercentMax=>0.0263, Coefficient=>0.263, FlurryBlowsMax=>1, AmountModifierPercent=>-0.82, IgnoreDualWieldModifier=>1, FlurryBlowsMin=>1, IsSpecialAbility=>1

WeaponDamage: StandardHealthPercentMin=>0.105, Slots=>[ PrimaryRanged ], StandardHealthPercentMax=>0.105, Coefficient=>1.05, FlurryBlowsMax=>1, AmountModifierPercent=>-0.3, IgnoreDualWieldModifier=>1, FlurryBlowsMin=>1, IsSpecialAbility=>1

The difference in AmountModifierPercent can be explained by dualwielding, however, it makes little sense it's the mainhand that hits the weakest. Sweeping Blasters appears to only use the offhand as well, leading me to believe Torhead mixes up mainhand and offhand effects.
I can't explain the differences in StandardHealthMax/Min and frankly I believe they're bugged.

Using the effect data for Full Auto, I get some numbers which seem to fit much better with anecdotical observations i.e. Barrage hits for 1700 damage and crits for 3700.

The good news is that it's fairly easy to fix. The bad news is it'll be hard to tell when Trooper or Bounty Hunter data is correct. Additionally, the Primary/Secondary weapon slot inconsistencies makes hittable implementations somewhat...shady.

[bural]

New version uploaded.

Changes:
-Fixed Unload and Barrage damage.
-Fusion Missile dot not affected by accuracy
-Fixed a bug causing DfA to do too little damage
-Implemented Patch 1.1.3 Surge calculations
-Setting HUD!B7 to more than 1 now allows Heatseeker Missiles to benefit from multible stacks of Heat Signature
-Implemented DPS per stat point feature
-Critical Reaction and its uptime partially modeled
-Initial calculations for the proc of Dark Energy Surge shown in HUD. Will be included in total DPS once a better relic menu has been implemented.
-Changed Ability Info to show hit and crit rather than average damage. The displayed ability damage numbers should be identical to what can be observed ingame. If they aren't, please bug report.

Initial calculations for Dark Energy Surge shown in the HUD. I will make a tap for relics at some point and include potential damage in total dps - currently it isn't included. You will notice the proc is worth some 73 Power. You should be able to choose gear accordingly. Additionally, it shouldn't be too hard to change the calculations to fit alternative relics.

I haven't included this feature in the sheet this time around, but using HSM with TSO+DfA is a dps increase of 1.77% assuming cooldowns stack up.

I'm undecided if the HSM calculation is still haunted by a bug, but it does not appear worthwhile to cast HSM if only 1 stack of Heat Signatures is present. When multible stacks are present however, it hits like a truck.

Fusion Missile isn't worthwhile untill 3 targets are in range.
DfA surpasses the efficiency of Barrage when hitting 3 targets, in which case it's superior to AoE with DfA and use TSO with Fusion Missile+HSM if it isn't possible to line up the cooldowns.

Admittedly, this version isn't as polished as I hoped for, but I wanned to push out a version with the Unload and Surge fixes as quickly as possible.

[Crythe]

Few comments based on last version of spreadsheet:

- minor bug: Dfa hit & crit cells are switched in HUD.
- Fusion Missile DOT ticks 6 times(SetTickRate: Interval=>1000), you have it set to 3 in CalcsV2 row 145.
- using HSM with only 1 stack of Heat Signature is now 0,6% DPS loss. Kinda sad if it's true, with only 1 player in group providing those debuffs it's only useful while moving or if target is about to put up shield/lose extra damage buff in next 1,5 sec.
- Dfa overall damage is about the same damage as non-barraged Unload(full Dfa is ahead with just 224 damage for me). Wouldn't it be better to use TSO with Unload as it can also vent some extra heat and use Dfa only in multi target situations? That's assuming you can't get free HSM off at the end of Dfa(never managed to do that as trooper).