Armor as DR presents an interesting opportunity that never really worked laid over the top of the game but can find its place if designed in from the outset. There were two main failing with armor as DR that I have encountered in the past. At low levels, armor tends to provide too much DR, making many enemy attacks almost irrelevant. This was largely a problem because low level PCs had so few hit points (one hit die), requiring low level monsters to deal little damage, but DR coming in Light, Medium, and Heavy “tiers” of armor. Second, armor becomes too necessary. When magic armor is granting 10 DR, all characters are overly incentivized to get as heavy of armor as possible. This isn’t fun when it shuts out legitimate character archetypes.
My working solution is to install a tradeoff. Armor grants DR but also lowers Reflex (the AC equivalent). This way, as monster damage increases, the reduction in defense becomes more detrimental at almost the same rate the DR increases.
When DR is acquired without penalty, it is always a no-brainer. Acquire as much as possible. This is because the effectiveness of straight DR is equal to [amount of DR] x [average monster hit rate]. If you have 10 DR and monsters hit 50% of the time, every time you are attacked you can expect an average saving of 5 damage (assuming monster damage is above 10).
The formula when you introduce a tradeoff is somewhat more complex. The armor saved more than it cost when the damage dealt < [ ( [DR x hit rate] / [old hit rate – new hit rate] ) + DR ]. Here it is graphically.
The top bar shows a normal character. The second bar shows the same character with armor that reduces defense by 2 for 1 DR. The area in which the character is damaged extends to be both the red and pink, but the entire band is reduced by the black and grey. As a result, the expected damage is indifferent. What we see is that the real tradeoff lies between the pink and the black. If the area of the black is larger than the area of the pink, the DR provided a net gain. DR, then, is most valuable when the enemies are likely to hit because a longer initial red zone means any DR will automatically provide a longer black bar. Here is a quick example.
If we set Light, Medium, and Heavy armor to -1, -2, and -3 defense, respectively, we can eek out a modest DR progression that makes each of them attractive across levels. This requires an understanding of expected monster hit rates (which I plan to have relatively consistent) and expected monster damage rates (which are also established). Since each level of armor represents a larger investment, I plan to make the payouts for higher armors similarly better.
The nifty thing is that even though over the course of the campaign armor will average out to be better, it might vary battle to battle. If you face a foe that hits rarely but hits hard, you’d be better off without armor and rely on your ability to dodge. The reverse is also true. This creates opportunities for cool scenes where the knight removes his armor before riding into battle, or, better yet, cuts it off during battle.