But its not 0-100.It is actual damage reduction (at least that is how I understand this). And in effect resistance you actually see these 8 points in variance when comparing padded armor and steel armor.
You quoted me, but I don't understand your point. It IS actual damage reduction. Yes.
It IS 0-100. (limitations apply) It is 1 AR = 1% damage reduction, each piece added on top of one another. The tooltips/journal reflect that and I've tested it at points to confirm.
The possible values are hard limited to 90 AR = 90% reduction (based on a dev comment earlier (propably Gazz, maybe faatal)) and I think the current gear max adds up to around 75 total max (which is why I used that in my calcs)
The range is further reduced by the decision to make Padded Armor useful (anything below 20% would literally be a rounding error in this implementation with small integers in the damage event side). I don't disagree that Padded should be useful, but I'm proposing Changing The Maths from "armor value" to "damage reduction", to enable a functional scaling / variance system, by just having more numbers to play with.
Alternatively I'd propose switching to floating point math for hit points, or at least banking math (counting pennies instead of pounds) to allow for less-than-one distinctions to be made. But that alone would look terrible when deciding between armor pieces at the top end of the spectrum.. you'd be deciding if a loss of 0.15 armor is actually worth the "other thing x".