Superstat/Mods/Spec testing with Sniper Rifle

selphea · July 2014

Here are the results of a week of spamming Sniper Rifle on the Private Test Server with various superstat setups, and over 50 combinations of Superstat, Form, Passive and Specializations:

Warning: Spreadsheet with wall of numbers

The Primary gear I used was Justice Agility with Gambler/Impact, Justice Precision with 2x Gambler and Justice Speed with 2x Impact. Stat Mods and Secondaries are described in the spreadsheet. The objective was to see what could be done with various superstat combinations with Superstat Con and maintaining a base of ~8000HP. Don't forget to look at the Options tab for tweakable settings and some explanations.

Special thanks to Championshewolf for generously providing the Hospitable Eyepiece and Barricade Bracers.

Full wall of text with observations that I made are on my blog: Click Here.

monsterdaddy · July 2014

Haven't at numbers yet but I'm pretty sure the damage resist debuffs follow the DR formula. i.e. a 10% debuff = 9.1% added damage.

aiqa · July 2014

Very interesting I'll look over it when I'm home.

One thing I have a question about is effective HP.
I dont understand how you do that one.

I'd do it like this

mitigation from resistance = 1-1/(1+resistance)
effective health form resistance = health / 1-mitigation from resistance = health/(1/(1+resistance)) = health * 1+resistance

mitigation from dodge = dodge * avoidance
effective health from dodge = health / 1-mitigation from dodge = health / (1 - dodge * avoidance)

putting that together would be
EHP = HP * (1+resistance) / (1 - dodge * avoidance)

So for the first row with a HP=10240, resistance=58%, dodge=22.7%, avoidance=43.8%, I'd get.
EHP = 10240 * (1+0.58) / (1-0.227*0.438) = 17965,4

quasimojo1 · July 2014

It's interesting and I'm sure it was a lot of work. Thanks for sharing.

selphea · July 2014

I calculated it based on Blue Laser damage. There's some rounding issues with Defense if I do a simulation based on what the character screen says.

So suppose the first entry has 0% Dodge, it would calculate EHP as:

HP x (Base blue laser damage / actual undodged blue laser damage)

At 10,240HP divided 949 actual damage, it can take 10.79 blue lasers. Multiply that by 1500 and it's 16,185

And then assuming it has 100% dodge, we'd have

HP x (Base blue laser damage / actual dodged blue laser damage)

So again at 10,240HP divided 533 dodged damage, it'd take 19.2 blue lasers. Multiply by 1500 and that's 28,818.

And then adjust for actual dodge chance: 77.3% of 16,185 and 22.7% of 28,818.

And factor them together to get

(0.227 * 28818) + (0.773 * 16185) = 19052

flowcyto · July 2014

Very nice work! Interestingly, if you don't assume things like DUC, adding just 16% resist from the regen dummies makes the Int PSS options comparable to EGO/Int/Con and DEX/Int/Con. I guess the take-home message is INT is better if the player isn't using things like DUC, applying to most content. If they are, then there's prob better options.

Nice note about Pres PSS too- I guess for builds using MS its a viable dps route (does come w/ giving up Force of Will, though- which kinda sucks). If End PSS is behind in dps, it does seem to have a pretty nice effect on EHP w/ Hardened.

I had honestly suspected for a while if soft-caping the Str/Ego additive bonuses at 70 was not really worth it just for that alone. Appears that for committed dps builds they usually shouldn't do it and just focus more on SS's to boost those dmg modifiers and the passive (unless they want some extra knock resist from Str, or w/e). Pres soft-capping can still be of value, but for completely diff reasons.

So good to know- will alter building advice I give in the future.
(and yes, Juggernaut is silly; too bad this wasn't a melee test to compare it w/ Overpower)

selphea · July 2014

flowcyto wrote: »

Very nice work! Interestingly, if you don't assume things like DUC, adding just 16% resist from the regen dummies makes the Int PSS options comparable to EGO/Int/Con and DEX/Int/Con. I guess the take-home message is INT is better if the player isn't using things like DUC, applying to most content. If they are, then there's prob better options.

Nice note about Pres PSS too- I guess for builds using MS its a viable dps route (does come w/ giving up Force of Will, though- which kinda sucks). If End PSS is behind in dps, it does seem to have a pretty nice effect on EHP w/ Hardened.

I had honestly suspected for a while if soft-caping the Str/Ego additive bonuses at 70 was not really worth it just for that alone. Appears that for committed dps builds they usually shouldn't do it and just focus more on SS's to boost those dmg modifiers and the passive (unless they want some extra knock resist from Str, or w/e). Pres soft-capping can still be of value, but for completely diff reasons.

So good to know- will alter building advice I give in the future.
(and yes, Juggernaut is silly; too bad this wasn't a melee test to compare it w/ Overpower)

I honestly hadn't thought about 1 mod for softcap until you mentioned it in the other thread

so I had to try it for SCIENCE! I think it would be a gain if the test case had 10 ego rather than 25. 25 is half-way to softcap as it were. I'd estimate the gain would be very close to 1%, although the CDR loss from less Int would be an issue.

And yes, Pri. Int is really good at low resist levels. Strangely, if you crank resist up to F&I's 200% it kinda falls back down again although it's still not too far behind.

flowcyto · July 2014

Well, even if its a gain from 10 instead of 25 base Ego, a ~1% diff is barely anything tbh. There's usually other gains from upping the SSs vs. Ego when its not an SS (and the defense/energy portions of the passives at least) that would prob make those more worthwhile (unless you just really hate holds).

selphea · July 2014

aiqa wrote: »

Very interesting I'll look over it when I'm home.

One thing I have a question about is effective HP.
I dont understand how you do that one.

I'd do it like this

mitigation from resistance = 1-1/(1+resistance)
effective health form resistance = health / 1-mitigation from resistance = health/(1/(1+resistance)) = health * 1+resistance

mitigation from dodge = dodge * avoidance
effective health from dodge = health / 1-mitigation from dodge = health / (1 - dodge * avoidance)

putting that together would be
EHP = HP * (1+resistance) / (1 - dodge * avoidance)

So for the first row with a HP=10240, resistance=58%, dodge=22.7%, avoidance=43.8%, I'd get.
EHP = 10240 * (1+0.58) / (1-0.227*0.438) = 17965,4

I see where the formula borked now. Updated

quasimojo1 · July 2014

I made a little decision matrix based upon your work. First I created new normalized summary columns, that for each build are the percentage of the potential best for Attack, Defense, Energy utilization, and Cooldown. Then I created a weight value for each of those categories to enter relative importance. Then I added a Weighted Average which calculates a final score based upon the Attack, Defense, etc. scores, factoring in the weighted importance of each.

So you just enter the relative weightings, and then sort the Weighted Average column to find the best build for the criteria based upon the data. If you have certain minimum requirements for some attribute (e.g. CD less than 40), you can apply that as a column filter.

Here's a link to the Google Doc:

https://docs.google.com/spreadsheets/d/1enwRenqe0BtOBYFB3v5NdaSUBHyGh87QWG6Un_YXxwA/edit#gid=938423748

Feel free to copy it in order to make your own edits.

Thanks again, Selphea.

selphea · July 2014

quasimojo1 wrote: »

I made a little decision matrix based upon your work. First I created new normalized summary columns, that for each build are the percentage of the potential best for Attack, Defense, Energy utilization, and Cooldown. Then I created a weight value for each of those categories to enter relative importance. Then I added a Weighted Average which calculates a final score based upon the Attack, Defense, etc. scores, factoring in the weighted importance of each.

So you just enter the relative weightings, and then sort the Weighted Average column to find the best build for the criteria based upon the data. If you have certain minimum requirements for some attribute (e.g. CD less than 40), you can apply that as a column filter.

Here's a link to the Google Doc:

https://docs.google.com/spreadsheets/d/1enwRenqe0BtOBYFB3v5NdaSUBHyGh87QWG6Un_YXxwA/edit#gid=938423748

Feel free to copy it in order to make your own edits.

Thanks again, Selphea.

Wow that's pretty cool :O

it seems to be based on the version before I fixed EHP though, my bad there.

I gotta do a disclaimer here - these numbers were done with Sniper rifle, which is a Physical attack, so Quarry will naturally appear higher. Second thing is Energy doesn't account for returns from things like MSA, Concentration etc, so incoming energy is also a factor. There's other complications, like whether there's additional debuffs taking debuffs to cap, and as mentioned in the blog, Active Offenses throwing numbers off, so use this as a guideline for general trends rather than absolute gospel.

But I'm planning a restat of a couple of my characters based on this so yes it does have a use and thanks for the matrix

aiqa · July 2014

selphea wrote: »

I calculated it based on Blue Laser damage *snip*

Ops sorry, I was to hasty and somehow lookeed at the "average attack" value in stead of "HP".

flowcyto · July 2014

Silly/simple question: but this is all running in Hybrid, even for the offensive passives, right? Cause the PSS vs. SSS passive dmg weighting is different for dps roles. May not really matter for dps FF's high into the dmg DR, though.

selphea · July 2014

Only AoPM was done in Hybrid.

Superstat/Mods/Spec testing with Sniper Rifle

Comments