Fractal Softworks Forum

Please login or register.

Login with username, password and session length
Advanced search  

News:

Starsector 0.95.1a is out! (12/10/21); Blog post: The Pilgrim's Path (07/19/22)

Author Topic: Misleading Danger (Auto-Resolve Strength) Assessment WRT D-Mods  (Read 299 times)

Dark.Revenant

  • Admiral
  • *****
  • Posts: 2740
    • View Profile
    • Sc2Mafia
    • Email
Misleading Danger (Auto-Resolve Strength) Assessment WRT D-Mods
« on: December 28, 2021, 04:48:44 PM »

Fleet danger assessment is impacted far too much by the presence of d-mods.  This is most noticeable with the Remnant ghost, since it's comprised of powerful, officered ships that are covered in d-mods%u2014especially carriers, which are not as negatively affected by d-mods as other types of ships.  At one point it showed me a 2-star assessment, which I was skeptical of... so I saved and tried to fight it.  It was a one-sided slaughter; I had absolutely no hope of victory.

Internally, 0 d-mods is a 1.5x multiplier within the strength assessor in Misc, and each d-mod reduces that multiplier by 0.2, to a minimum of 0.5x after five d-mods.  Thus a ship with 5 d-mods has a 0.5x strength multiplier, i.e. 1/3 the strength of an otherwise identical but fully restored ship.  This is frankly way too much with the current d-mod effects.


Generally speaking, a ship with half the stats in a few major categories is roughly half the performance, if we ignore the finer points about armor cells and weapon/engine mounts.  I'd say there are about four stats that, if you were to halve all of them, would result in a half-strength ship: flux capacity, flux dissipation, hull, and armor.  Other major stats that would have similar impact include speed, maneuverability, range, shield/phase efficiency, and CR.  For carriers, the impact is split between d-mods that affect the mothership, and d-mods that affect the fighters.  This generally means that carriers are affected less by d-mods, since they get the same severity of d-mods as regular combat ships, but comparatively less of their performance is affected by them.

Specifically:
Compromised Armor: -20% to one stat.
Compromised Hull: -30% to one stat.
Damaged Flight Deck: -30% to one particularly major carrier stat.
Damaged Weapon Mounts: -30% to one relatively minor stat, -25% to one relatively minor stat.
Defective Manufactory: -25% to two carrier stats.
Degraded Engines: -15% to two stats.
Degraded Life Support: -7.5% to one particularly major stat.
Degraded Shields: -10% to one particularly major stat.
Erratic Fuel Injector: -20% to one relatively minor stat.
Faulty Power Grid: -15% to two stats.
Glitched Sensor Array: -10% to one particularly major stat.
Increased Maintenance: -7.5% to one particularly major stat.
Malfunctioning Comms: -40% to one carrier stat.
Phase Coil Instability: -33% to one stat and -30% to one relatively minor stat.
Structural Damage: -20% to two stats.
Unreliable Subsystems: -30% to one relatively minor stat.

Mathematically, on average for a non-carrier ship, each d-mod is equivalent to a 6% penalty to the ship's performance, going by the 4-stat theory.  This is consistent with Derelict Operations reducing the deployment cost by 6% per d-mod.  If a carrier's performance is evenly split between fighters and direct combat strength, then it's more like a 3.5% penalty for those carriers.  For shieldless/phaseless ships, flux capacity is barely a minor stat, so the overall pool of meaningful stats is lower and we can approximate about an 8% penalty per d-mod.

That 6% figure is far smaller than the roughly 13.5% penalty that the strength assessment algorithm currently assumes, and for carriers it's almost twice as bad.


If it were instead a 1.2x multiplier with 0 d-mods and -0.08x per d-mod, it would still conservatively overestimate the effects of d-mods, but that's fine because compound penalties tend to be slightly exponential when applied to a single ship.  The 5 d-mod remnant ships would be treated as 2/3 their original strength, rather than 1/3; in other words, it would show four or five stars of danger instead of two, which feels way more accurate.  This could be further broken down to react differently with regards to carriers, perhaps by decreasing the penalty to 0.065x for combat carriers and 0.05x for pure carriers.  I'd suggest a similar thing for Derelict Operations as well; the 6% DP reduction really ought to be 3% or 4% for carriers.

This has trickle-down effects to ship quality from the inflater parameters and, by extension, the faction doctrine system.  Right now, ship quality is rather important for auto-resolve strength, but ship quantity does way more for actually winning battles in-engine.  As an aside: officers are also slightly overestimated (+20% ship strength per level); it's probably closer to +15%, but that's close enough for a rough estimate because stacking bonuses tend to be slightly exponential when applied to a single ship.
« Last Edit: December 29, 2021, 11:38:07 AM by Dark.Revenant »
Logged

Alex

  • Administrator
  • Admiral
  • *****
  • Posts: 20575
    • View Profile
Re: Misleading Danger (Auto-Resolve Strength) Assessment WRT D-Mods
« Reply #1 on: December 28, 2021, 04:55:54 PM »

(Moved this to suggestions since I don't think it belongs in bug reports. That's for more cut-and-dry things.)

Right. Thank you for the write-up! Made a note to have a closer look later; as you noted this has repercussions for the value of the various doctrine parameters.
Logged