A parry in a given fight will nullify, on average, X damage prior to non-passive mitigation. Determine, perhaps based on prior fight (HP) resource graphs, the likelihood that the parry could leave one at a survivable health value through that 20% (previously up to 25%) mitigation and thus forgo any healers' average maximum spammable GCD heal. Then, look at the potency of an average healer's filler attack. The product of that chance any attack potency is, pretty much, the parry's rDPS value.
The problems with additionally defensive builds are primarily just thatYes, every parry could end up affecting only an auto. But that is what is what the averages are for.
- they have traditionally been tuned too low to be competitive in rDPS (not a fundamental issue at all, but a balance issue), and
- the value of any GCDs of healing saved is impossible to tap into when no optimized healer is using GCD healing regardless --at least outside of ReciAdloDeployment (which is, again, not a fundamental issue, but a contextual one).
Yes, there is a further issue of "illusion of choice": As content hits more than X over a fight, increasing the rDPS value of a random parry therein, it becomes optimal and short of that point it's inferior. But that is already true for every stat, and typically with far less context-dependence. That's not a problem fundamental to defensive builds, but an incidental problem from content tuning relative to available (free) healing.
