User:Zoboso has updated the Armor page with information based off your most recent guide. I'm bringing this to your attention because you commented on your change to Ironclad Charge saying that you were wrong about some part of your formula at one point.
Can you give it a look over and verify that his update to the formula is consistent with the final results of your testing?
There is an issue in that Steel Fiber is being described as a base armour modifier. This is untrue as per definition of "base".
Mods that grant you a percentual bonus require a specific absolute value upon which the percentage is applied. The base value is the value from which the bonus is calculated. All mods and abilities that incur a percentual bonus for a parameter draw from their distinct base value of the parameter, and then apply it to the net value, which is the resulting value, or actually to the base value itself.
To change base value inherently means that all the percentual bonuses that draw from it will yield greater results. Therefore,if Steel Fiber was to apply itself to base value, all abilities that influence your armour would have a larger effect. This is untrue, as Hallowed Ground in particular stacks with Steel Fiber merely additively. Therefore, Steel Fiber is NOT a mod that applies itself to base armour value, but to the net value.
The reason it stacks multiplicatively with IC is that IC applies to base, increasing the base for SF's percentual bonus. Through commutativity and associativity, it can be explained as SF empowering IC, even though it's actually IC empowering SF.
Other than that, the formula appears to be correct.
I have no qualms with the changes either of you guys have made. The previous information about these various ability augments are based off assumptions; they are there because in the absence of testing, we think it's reasonable to expect these mechanics to work consistently with earlier examples of modifiers. In the face of Pelia's recorded, citable testing, there is no way we could argue to keep what was previously written.
I brought this up because I wanted to make sure that Zoboso's changes reflected Pelia's tested formula, and I had concern about that because Pelia at one point updated his blog post and the Ironclad Charge page in order to update that formula.
As for the "base" semantic... I agree with you. People bandy about with that word all the time without defining it properly, as they do with "additively" and "multiplicatively". I, too, became soft in that usage, especially because lots of the text on the Armor page I had written was the regurgitation of text written on their respective ability pages. Let's give a tighter scrutiny to "base" from now: if it increases the modifier, it increases net armor additively, but if it increases the base, it increases base armor multiplicatively.
Actually, multiplicative or additive relation is a derived parameter, and can only describe relation between two mods, therefore, I can't say I entirely agree with your proposal of terms, as they imply this relation is describing a single mod, without having a comparative peer.
The reason I had to point this out in particular is that when you take two "net" mods, they stack additively. Then take a "net" and a "base" mod, and stack is suddenly multiplicative, even though a net mod "increases net armor additively". While mechanically, the base mod stacks up the net mod, due to commutativity and asocciativity, you are to be able to describe it the other way around. Your formulation excludes this description and that would make the formula void, as basic attributes of binary operations would be conflicting this definition.
Additive stacking is a relation describing combining two net mods' bonuses that amplify the same attribute. It may be described by this general formula:
New net value = Base × (1 + Mod A + Mod B)
Multiplicative stacking is a relation describing combining a net mod and base mod's bonuses that amplify the same attribute, or a combination of two base mods of the aforementioned nature. It may be described by this general formula: