Well, there's some weirdness with the shield power stuff as well, which could account for the variance. I started to run another test at shield power, before my connection dropped me. Got the following:
Cruiser - 119 power = (119-50)/500 = 13.8% expected.
So these numbers don't quite match up and varied a bit for some reason. Not quite sure what is up with that.
Additionally, a test at 125 power (119 + 33, actually), gave the following:
344 (411) = 83.6
I lost other numbers with that power level, due to lost connection, so I guess that one will have to do. =/
Greater than the 15% expected I'd expect with (power - 50) / 500. Assuming the (power - 50) part is correct and my carrier's ~50 power tests seem to confirm that, it would mean the divisor is not 500. Given that, 75 / X = 16.4%. So 75 = 0.164X. So 457, but that seems a bit strange number for a designer to pick, so going to round it up to 460.
Plugging that back into the 119 power test, (119-50)/460 = 15% expected. Closer.
Small sample size on my part and using the rounded numbers from the basic combat chat window, so I'd recommend further testing, but here's the potential new formula:
(power_level - 50) / 460.
I'm seeing with net 85 Shield Power, RSF II at +49% DR and EPtS II with +32% DR. I'm seeing 66.6% damage reduction.
but using your (SP - 50) / 500 formula for Shield Power DR it does come close at .6775
(85 - 50) / 460 = 7.6%. That would give 0.924 * 0.51 * 0.68 = 32.04% damage remaining, a 67.96% reduction. Slightly further away.
Well, it all needs further testing. RSF is unfortunately not ideal to drop into the mix, since its exact percentage is going to be rounded in the UI. Would have to figure out your exact skill bonus to see what the decimal parts being rounded are, assuming that a total of 100 skill gets you 2x the base.
For that matter, shield power is actually fractional as well, as far as I can tell, and rounded in the UI.
Some thoughts on what type of character would be best used to test just shield power, nothing else, to generate a nice clean line.
It'd need to avoid warp core, efficiency, and performance skills, including from traits and BOffs. Probably an engineer for power transfer to boost shield power up over 125 (possibly with console help) from 100. In a non-cruiser, to avoid the cruiser shield power bonus. And that should let one test from 25 all the way to 125 in 25 point increments.
And to add to the weirdness, I tested out my Lt Cmdr/ Klingon Captain in B'Rel BoP. The Shield DR based on Shield Power performed exactly like the devs said it would. 10% Shield DR at 50%, 6.8% at 34 power, and 20% DR at 100 power. Adding in other Shield buffs give calvulated DR value close to observed, but off by several percent much more then rounding errors can account for. Definately something other thea we expect is going on under the covers.
Guess I might have to take my B'rel out for a spin tomorrow and see what sort of numbers turn up at 51. Be interesting to see if it matches my expected results or what you saw. Wondering if ship or player tier might factor in =/
Only need 1 ship to fight, really, as my usual testing method is to fly up to one, let it shoot at me a bit, and then run away. But, yes, the ability to quickly get into a fight with a single ship is handy, so the biggest factors for me are a short travel distance to find the fight, coupled with enough room to run away from it.
I've been using Sortie Alpha's Maiewwhatever system mission Klingon side, for example.