PvE DPS Balance Initiative: Scaling and Midrange

Discussion in 'War Room (Powers, Artifacts, & Builds)' started by Sore, Aug 31, 2015.

  1. Sore Steadfast Player


    Following GU50, the developers the developers defined a spec for baseline testing in the Testing Tips, Do's and Don'ts thread. In the GU50, Tunso Testing, and Player Testing thread, players decided to meet the devs more than half way and conduct range and melee balance tests in the baseline Tunso Test spec. It also included player feedback expressing concern about how that spec scales to live performance and the need for acknowledgement of the midrange playstyle. The devs formally acknowledged a DPS Balance Initiative with three phases.
    Phase 1: Baseline Balance
    The first phase was based on making rapid adjustments to powersets using the Tunso Test spec. There was continued player feedback about the use of this spec but players agreed to suspend disbelief long enough to see this phase concluded. The goal of this phase was to get all powersets performing within a +/- 5% tolerance of specific targets.
    • Melee Target: 20,644 DPS
    • Range Target: 14,120 DPS
    Adjustments were made in rapid succession by tweaking the numbers on powersets. These were not revamps. They were not aimed to change the design of the powersets. We are now going to conclude this phase of the initiative. The targets were met with five exceptions: Earth, Sorcery, Quantum, Celestial and Gadgets. Read on for details on them.

    Phase 2: Scaling
    At the heart of scaling is agreeing upon a scaled-up spec for the purposes of determining how different powerets scale up between the baseline "Tunso Test" spec and the scaled "Live Test" spec. This may include changing distribution of stats, tweaking how design components work, and/or additional adjustments to base numbers to make the results perform within +/- 5% tolerance of scaled up targets.

    Live Test Targets
    Melee Target: (To be announced)
    Range Target: (To be announced)

    Live Test Spec
    The Live Test spec is best done on PC Test to control variables. In general most players aren't willing to downgrade their modding on Live and might not have the plans. Some might not be CR126 anymore either. So I'd rather just keep it to PC Test.

    I buy iconic and rare styles to get my SP in that range. That does take a good amount of time to setup the toon. Tunso Test toons are quick/easy to create comparatively. You can mod differently. I have 5 PC Test Toons. One for trying content that I try to keep fully maximized. One for Tunso Tests. One for Precision Live Tests. One for Might Live Tests. One for Prec/Might Live Tests. Feat unlocks really help making additional toons. Each toon has 4 bases named: NatureElec, MunFireIce, QuanMenGad, RageCelLight. Each base has armories setup for the powersets the base names. I stock up on testing helpers to get access to power change tokens. I use those for switching powers of for field respecs.

    Skillpoints are generally,
      • Middle of every tree
      • Any desired non-WM combos
      • All T1 Crit Innates
      • WM for any desired WM combos
      • WM Prec/Might Innates for any WM you unlock
      • Unlock additional WM based on the T2 Innates provided
      • Might build goes for Might T2 innates first
      • Precision build goes for Precision T2 innates first
    CR 110 Gear
    10423 Might
    3380 Precision
    Mainframe Mods: Core Strength, Escalating Might, Max Damage
    RD Mods: Might VII, Might/Health VII, Might/Powers VII
    Generator Mods: All Precision 6.2
    League Proficiencies: Damage Out, Critical Attack, Might & Precision
    160 SP getting All T1 Crit Innates, T2 Might Innates, and any available for WM Damage Innates

    CR 110 Gear
    8874 Might
    4873 Precision
    Mainframe Mods: Relentless Precision or Escalating Might, Core Strength, Max Damage
    RD Mods: Prec VII, Prec & Health VII, Prec & Power VII
    Generator Mods: All Precision 6.2
    League Proficiencies: Damage Out, Critical Attack, Might & Precision
    160 SP getting All T1 Crit Innates, T2 Precision Innates, and any available for WM Damage Innates

    Might/Precision Based
    CR 110 Gear
    10049 Might
    3928 Precision
    Mainframe Mods: Core Strength, Escalating Might, Max Damage
    RD Mods: Might/Prec VII, Might/Health VII, Might/Powers VII
    Generator Mods: All Precision 6.2
    League Proficiencies: Damage Out, Critical Attack, Might & Precision
    160 SP getting All T1 Crit Innates, T2 Might Innates, and any available for WM Damage Innates

    Testing results should look like this:

    Scaling Concerns
    • Pets:Pet damage does not scale up between Baseline Test and Live Test
      • For this reason, Earth and Sorcery were balanced directly to the Live Test
    • Mainframe Mods:Quantum takes advantage of Core Strength / Penetrating Strikes twice, once while loading up a Time Bomb and then again when the Time Bomb pops.
      • For this reason, Quantum will be balanced directly to the Live Test
    • Precision Damage:Weapon DPS does not scale up between Baseline Test and Live Test
      • It is believed this benefits powersets like Ice, Fire, Munitions, Gadgets, Mental, Quantum over powersets with a precision component like Hard Light, Rage, Celestial, Nature and Earth

    Phase 3: Midrange
    Currently we are awaiting a decisions from the developers on how to address mid-range. The primary choices for them are to:
    1. Remove Midrange from all Powersets
    2. Add Midrange to all Powersets
    There are variations to how those options are executed. Fundamentally these adjustments will mean design changes to every powerset.

    Melee and Range
    Regardless if any content supports it, powersets have been designed to enable melee damage at 0m for max damage potential and range damage at up to 25m for acceptable damage potential. Players have always been able to mix melee and range techniques for a damage potential higher than acceptable while approaching but less than the max potential.

    Midrange History
    This discussion identifies midrange as taking a ranged positioning within 12.5m of your target and leverage abilities that work within that position. This is of interest because of the history of AMs. Gadgets was the grandfather of the 12.5m move with EMP. Player envy led to future AMs of other powersets to request melee damage ranges to extend to 12.5m for other powersets or to include a 12.5m damage scalar added. These tweaks were done in the mentality of hitting additional adds while engaging in 0m.

    Gadgets and Celestial
    Gadgets has historically broken a number of rules that are worthy of adopting for player preference sake or addressing. It is for this reason that Gadgets will not be balanced in phase 1 but rather at some point during phase 2 or 3 to ensure it can be looked at in a single pass. Similarly, Celestial doesn't differentiate melee and range damage at all and will be balanced in a single pass also.

    Nature and Electricity
    One key consideration will need to be how midrange is addressed for DoT powers. There's a difference between needing to stay at 0m or needing to stay within 12.5m for the entire duration of a fight compared to only needing to jump into melee or midrange to cast a single power and then safely retreat to full range. This is specifically a consideration for Nature and Electricity.
    • Like x 16
  2. Sore Steadfast Player

    Other Design Considerations
    It is worth noting the other aspects of balance that are not actively a part of this initiative. It does not mean these things won't get attention at some future point. It doesn't mean they will. For now it is important to acknowledge them:
    • Setup Time: Burst powersets can unload damage faster, providing better scoreboard results against trash mobs
    • Complexity: Some powersets have AMs as easy as 1-2-3-4 while others require complex input sequences and timings.
    • Interrupts: Some powersets are highly vulnerable to being countered while executing their AMs while others are not.
    • Targets: The balance targets were set against 8-targets. This initiative does not address performance against a single target.
    • PvP: The entire topic of PvP is worthy of its own breakdown as the design attributes that favor PvE are not the same that are important to PvP.
    Charts and Graphs
    I still have some live test results from the old thread to add in or that we still need to collect. I will update the images as we get those numbers.





    Live Tests









    • Like x 8
  3. Sore Steadfast Player

    I want to thank everyone that helped contribute to this. I've played a role helping organize this and keep its momentum going but this isn't my effort. This is the effort of many contributors. This is countless hours of testing and great feedback from many people. And the work isn't over. I'm humbled by the greatness of this community. You guys deserve all the credit. Keep up the good work because there's still more to be done.
    • Like x 6
  4. Sore Steadfast Player

  5. Sore Steadfast Player

  6. Sore Steadfast Player

    Mental Stuff


  7. Sore Steadfast Player

    Munitions Stuff


  8. Sore Steadfast Player

    Quantum Stuff



  9. Sore Steadfast Player

    That's about it for the fun graphs. More can be made in the future. It's just all I have in the spreadsheet.

    Here's a couple tables of when I was trying to see where the maximum damage was and comparing it across the AM types.

    I did similar tables but summarizing up to averages in case certain AM types should have different baseline targets

    I know the Tunso Test spec isn't ideal. I think we could have done the entire initiative with just the Live Test spec. But the Live Test spec is kind of grueling to put together. There's convenience in setting up just a Tunso Test toon. I'm wondering if the scalars I've found can be applied to each Tunso Test result to predict live test results. In that sense, maybe we don't think of things as having a single scalar. That would let Tunso Test results be more predictive and lower the barrier of entry into testing.
  10. Remander Steadfast Player

    I'll also say that having this done should make PvP easier to address. Rather than little band-aid fixes to PvP damage, this should allow more standardized scaling.
  11. Remander Steadfast Player

    I think this was also constructive to see what's under the hood, so to speak. Without TT vs LT, the scaling inconsistencies would remain. Now, we can hopefully address these for the long-term.
  12. L T Devoted Player

    Are the Celestial Live tests run with Escalating Might or Relentless Precision? Since you're not actually using weapon attacks, does Relentless Precision even offer any benefit?
  13. Sage-Rapha Steadfast Player

    Celestial cannot activate Relentless Precision.
    It's useless to Celestial, so it's likely to assume Escalating Might was used
  14. Sore Steadfast Player

    Remander did them. I updated the post to indicate the Precision Live Test spec can choose accordingly. That's a concern of scaling I also have. Relentless Precision is harder to activate and maintain. It doesn't benefit a precision-based class like Escalating Might benefits a might-based class. The same kinda goes for mixed classes. Nature is part might and part precision. It'll take Escalating Might but it won't benefit from it as much as Ice benefits from it.

    That also reminds me about my concern regarding PC Test not having Precision/Might mods available. I really think they should be. I think they should be options for mixed classes. Minimally Electricity and Nature should be using them because they would be on live.
  15. Remander Steadfast Player

    Yeah, twas Escalating Might, for whatever that's worth, lol!
  16. Sore Steadfast Player

    One thing I've learned is that I really am not keen on assessing scaling when a different person conducted the Tunso Test from the Live Test. I've retested a few of those personally and saw notably more aligned results.
    • Like x 3
  17. Remander Steadfast Player

    Exactly why I ran the Earth tests both ways. I wanted to be sure I was close to Penryn and Karasawa, before running the live tests.
  18. Sore Steadfast Player

    Do you mind repeating those as Tunso Tests? I'd like to see the scaling yourself against yourself. Right now I'm scaling from Karasawa (TT) to you (LT).
  19. Sore Steadfast Player


    We still need the following:
    • Tunso Test and Live Test of Rage Melee from same player
    • Tunso Test and Live Test of Rage Range from same player
  20. Sore Steadfast Player

    For reference, this is what an ideal test submission should look like.

    Quantum Ranged + Robot Sidekick

    Tunso Test

    Range: 10,201 - 21,571
    Mean: 15,952

    [Summary] Damage [10.0s] 11380/s - 113800 total - 140 hits (4480 max) - 31 (22.1%) crits - 8 targets
    [Summary] Damage [10.0s] 20486/s - 204862 total - 119 hits (7754 max) - 29 (24.4%) crits - 8 targets
    [Summary] Damage [10.0s] 19112/s - 191122 total - 151 hits (7754 max) - 38 (25.2%) crits - 8 targets
    [Summary] Damage [10.0s] 10201/s - 102006 total - 115 hits (4065 max) - 30 (26.1%) crits - 8 targets
    [Summary] Damage [10.0s] 12674/s - 126739 total - 133 hits (4992 max) - 30 (22.6%) crits - 8 targets
    [Summary] Damage [10.0s] 14293/s - 142930 total - 125 hits (4281 max) - 29 (23.2%) crits - 8 targets
    [Summary] Damage [10.0s] 13188/s - 131875 total - 139 hits (5353 max) - 36 (25.9%) crits - 8 targets
    [Summary] Damage [10.0s] 18313/s - 183130 total - 125 hits (7754 max) - 26 (20.8%) crits - 8 targets
    [Summary] Damage [10.0s] 21516/s - 215160 total - 154 hits (7754 max) - 29 (18.8%) crits - 8 targets
    [Summary] Damage [10.0s] 11601/s - 116013 total - 140 hits (4057 max) - 30 (21.4%) crits - 8 targets
    [Summary] Damage [10.0s] 21571/s - 215709 total - 125 hits (7754 max) - 26 (20.8%) crits - 8 targets
    [Summary] Damage [10.0s] 12453/s - 124534 total - 113 hits (5442 max) - 28 (24.8%) crits - 8 targets
    [Summary] Damage [10.0s] 19029/s - 190290 total - 153 hits (7754 max) - 32 (20.9%) crits - 8 targets
    [Summary] Damage [10.0s] 17523/s - 175226 total - 137 hits (7754 max) - 32 (23.4%) crits - 8 targets

    Live Test

    Range: 16,615 - 31,717
    Mean: 23,862

    [Summary] Damage [10.0s] 17638/s - 176376 total - 120 hits (6822 max) - 33 (27.5%) crits - 8 targets
    [Summary] Damage [10.0s] 25884/s - 258840 total - 111 hits (11039 max) - 27 (24.3%) crits - 8 targets
    [Summary] Damage [10.0s] 19125/s - 191246 total - 126 hits (7673 max) - 37 (29.4%) crits - 8 targets
    [Summary] Damage [10.0s] 23321/s - 233211 total - 113 hits (7782 max) - 32 (28.3%) crits - 8 targets
    [Summary] Damage [10.0s] 22591/s - 225912 total - 104 hits (11343 max) - 27 (26.0%) crits - 8 targets
    [Summary] Damage [10.0s] 22632/s - 226324 total - 131 hits (11343 max) - 27 (20.6%) crits - 8 targets
    [Summary] Damage [10.0s] 29035/s - 290351 total - 114 hits (11343 max) - 30 (26.3%) crits - 8 targets
    [Summary] Damage [10.0s] 19787/s - 197868 total - 131 hits (7761 max) - 34 (26.0%) crits - 8 targets
    [Summary] Damage [10.0s] 28299/s - 282989 total - 106 hits (11343 max) - 32 (30.2%) crits - 8 targets
    [Summary] Damage [10.0s] 22250/s - 222500 total - 121 hits (11343 max) - 27 (22.3%) crits - 8 targets
    [Summary] Damage [10.0s] 31717/s - 317171 total - 121 hits (11343 max) - 32 (26.4%) crits - 8 targets
    [Summary] Damage [10.0s] 16615/s - 166146 total - 123 hits (6049 max) - 33 (26.8%) crits - 8 targets
    [Summary] Damage [10.0s] 28220/s - 282197 total - 116 hits (11343 max) - 33 (28.4%) crits - 8 targets
    [Summary] Damage [10.0s] 23913/s - 239133 total - 128 hits (11343 max) - 43 (33.6%) crits - 8 targets
    [Summary] Damage [10.0s] 27355/s - 273551 total - 111 hits (11343 max) - 30 (27.0%) crits - 8 targets
    [Summary] Damage [10.0s] 22496/s - 224963 total - 106 hits (11343 max) - 35 (33.0%) crits - 8 targets
    [Summary] Damage [10.0s] 28736/s - 287360 total - 127 hits (11343 max) - 37 (29.1%) crits - 8 targets
    [Summary] Damage [10.0s] 19903/s - 199033 total - 125 hits (7680 max) - 42 (33.6%) crits - 8 targets

    Scaling: x1.50