Armor as DR (sicko version)

So, the simple damage-reducing armor system I posted last time is the one that I would actually use, the one that I'm probably going to include in that B/X hack I've been messing with, on and off.

However. I'm an admitted simulation sicko, and I've also got another version of that mechanic in mind. It might be too annoying to use at the table due to the small but persistent cognitive load it would add to the already-more-hassle-than-AC armor-as-DR system I proposed previously.

The idea is simple: If the attacker rolls high enough, they bypass the attacker's armor. So, effectively, every armored creature would have two "armor classes": a defense score that an attacker must meet or beat to hit the target and an armor score that an attacker must meet or beat to inflict full damage. If their attack roll falls between those numbers, the target is hit, but their damage reduction is subtracted from the damage they suffer.

The target's defense score would basically be their unarmored AC, probably 10 + Dexterity modifier + shield bonus, etc.. The armor score would be the defense score plus the coverage score of their armor.

The easy way to convert armor to this system is to just give everything a coverage of 10 and a DR equal to its normal AC modifier—double the DR value on my previous system, since it can now be bypassed. And that's the assumption I made when putting together my comparison spreadsheets.

Armor-as-DR experiment v.2.0

I've improved my testing methodology. Instead of rolling 5,000 actual tests and averaging the results, I'm just calculating the average damage in a single formula (and factoring in a double-damage-on-natural-20 crit mechanic). I'm basing my numbers on Old-School Essentials in ascending AC mode. I charted out the expected damage for large ranges of attackers and targets with the standard AC mechanic, my simple armor-as-DR mechanic, and my limited-coverage DR mechanic. (In order to make my life easier, I just gave the targets in the simple DR test 100 armor coverage instead of coding coverage out all together.) Then I compared the results, first comparing AC to armor-as-DR, then comparing AC to the limited-coverage version.

In those comparisons, I'm subtracting the expected damage in the AC system from the expected damage in the variant system. So, a number (and redder cell) means that, in the variant system, more damage would be done in that particular attacker-target pairing. The occasional negative number and slight trace of blue shows where the variant system results in less damage. The green areas are where good old AC and my DR hack produce very similar results.

Interestingly, while both DR systems result in armor being less effective against high-damage attacks, the charts for the sicko version (limited coverage) are a lot smoother and greener than the simple DR mechanic. Which does make me wonder how my gaming group would feel about having to compare every attack roll to two different target numbers.

Anyway, the real reason I'm interested in this more complicated armor system is that coverage gives you another fun number to use. Instead of just giving every kind of armor 10 coverage, I could include low-coverage stuff like breastplates and leather jackets. If I wanted to go really sick, I could model piecemeal armor by adding even more armor threshold scores. Damaged armor could lose coverage (which seems more appropriate than losing DR). Helmets could add coverage—I could do a whole post about how helmets are cool and we should think about helmets more!

But yeah, I dunno if I'd really want to inflict this on my friends. I've got a higher complexity tolerance (and more dire simulation perversion) than most folks, and I am very aware of that. For now, this is just thought experiment territory.