Second test run is now complete.
I ran 8 series of tests: 5.1, 5.2, 6.1, 6.2, 7.1, 7.2, 8.1, 8.2. Details further below.
If have simplified the time distribution charts: one chart rather than two, no more dashed lines.
Tested shipsVigilance is now being ignored.
Lasher, Centurion, Wolf, Brawler, Tempest, Shrike, Medusa and Hyperion got new builds.
Scarab was not tested again, I re-used previous results for this particular ship.
Spoiler
custom Lasher: 9 capacitors, 10 vents, reinforcedhull, 2 lightag, 2 sabot, 2 vulcan, 1 lightdualac
custom Centurion: 10 capacitors, 10 vents, reinforcedhull, 1 sabot, 2 lightac, 2 lightag, 2 vulcan
custom Wolf: 10 capacitors, 10 vents, reinforcedhull, 1 ionpulser, 1 amblaster, 2 sabot_single, 2 pdlaser
custom Brawler: 7 capacitors, 10 vents, reinforcedhull, 2 hveldriver, 2 sabot_single
custom Tempest: 10 capacitors, 10 vents, reinforcedhull, 1 amblaster, 1 sabot, 1 ionpulser
custom Scarab: 8 capacitors, 10 vents, reinforcedhull, fluxdistributor, fluxcoil, 2 sabot_single, 1 amblaster, 2 irpulse, 2 pdlaser
custom Shrike: 15 capacitors, 18 vents, reinforcedhull, 1 ionpulser, 1 sabot, 1 amblaster, 2 irpulse, 2 pdlaser
custom Medusa: 20 capacitors, 14 vents, reinforcedhull, 2 lightmg, 1 ionpulser, 1 amblaster, 5 pdlaser
custom Hyperion: 4 capacitors, 10 vents, reinforcedhull, extendedshieldemitter, hardenedshieldemitter, stabilizedshieldemitter, advancedshieldemitter, adaptiveshields, unstable_injector, 1 heavyblaster, 1 heavymg
Tested ships were either used as single ship fleets (5.1 to 6.2), or used as part of 16 DP (7.1 and 7.2) and 25 DP fleets (8.1 and 8.2).
Enemy ships(unchanged)
Test series with performance ranking and data visualisation(5.1) Single ship test fleet vs
1 Kite (A): Tempest > Scarab > Hyperion > Shrike=Lasher > Brawler > the rest
(5.2) Single ship test fleet vs
2 Kite (A): Tempest > Scarab > Hyperion > Shrike=Medusa > the rest
(6.1) Single ship test fleet vs
1 Cerberus: Tempest=Shrike=Scarab=Medusa > Hyperion=Lasher > the rest
(6.2) Single ship test fleet vs
2 Cerberus: Hyperion=Scarab > Tempest > Shrike > Medusa > the rest
(7.1) 16 DP test fleet vs
1 Kite (A) + 1 Lasher + 1 Condor: Tempest > Scarab > the rest
(7.2) 16 DP test fleet vs
1 Falcon: Tempest > Lasher > Scarab > Shrike > Centurion=Wolf > the rest
(8.1) 25 DP test fleet vs
1 Kite (A) + 1 Lasher + 1 Condor + 1 Enforcer: Tempest > Scarab > the rest
(8.2) 25 DP test fleet vs
1 Dominator: Tempest > Scarab > Lasher > Shrike > Wolf > Medusa=Centurion > the rest
CommentsI am not going to dive into performance comparison between run 1 to run 2, because it could then get complicated. Suffice to say the performance of ships increased or decreased depending on situation.
One thing worth mentioning that I initially did under-estimate is, some of those matchups do really test AI’s ability to deal with enemy’s missile/fighter pressure. Having effective defence against those threats proved helpful, that is for ships/builds that had such defence. Both Lasher and Centurion get honourable mentions in this regard.
If we focus on the big picture, what do we see here… again? Overall Tempest is above everything else. Then comes Scarab. Yes, in 0.95, 16-DP worth of Tempest is better than 16-DP worth of anything else tested here. Same with 24-DP. Which most certainly means in 0.95 a Tempest's effective worth is more than it's 8-DP cost.
If anyone reading this still do wonder about Hyperion not having better performance/rank: this is skill-less non-SO, and AI-controlled Hyperion happens to almost never use its ship system with such builds. The build used here has omni shield conversion which tricks AI into lower shield more often in order to do some passive venting, and one empty medium hard-point in order save OP and have easier flux management, but it’s not enough to have the ship perform well in harder matchups. That said, it doesn't mean Hyperion is worth less than 15-DP. Quite the contrary IMO, given how it performs in player hands and/or with SO or elite Helmanship skill, ... but this was not in the test scope.
Please note all test results shown in this thread will be obsolete once 0.95.1 is out.