There are two spell-casting systems in common use. Under the older PPA, a caster starts…
- User
There are two spell-casting systems in common use.
Under the older PPA, a caster starts reserving one MP per turn, and this reservation process cannot be aborted, but only ends when the verbal and somatic triggers of a spell costing at most the reserve are performed, expending the entire accumulated reserve even if it exceeds the nominal spell cost but also giving that spell effect. (Hence, one can change one's mind about what spell to cast at any point during the process.) Should one run out of mana to supply the reserve (say, if one has been incapacitated after starting an accumulation but before selecting a spell), the mana pool is expended into the ambient æther, acting as a beacon announcing the caster's location.
Under the newer MENDA, a caster dedicates some amount of available mana and performs the verbal and somatic components of a spell, and it either fails or takes effect immediately, but the caster is then on a cooldown of one turn per MP spent during which they cannot cast. Unlike PPA, where spells are guaranteed to succeed, here the probability of failure is exp(-mana expended / nominal cost of the spell). (Theoretically, one could always opt to repeatedly spend one MP at a time on every spell until success and thus only pay the minimum amount of mana necessary according to one's luck, but this actually comes at an action-economy cost because all of those turns would be taken up by repeated chants and gestures as opposed to just the first one.)
- Assistant
Interesting setup! Two systems, each trading off reliability, tempo (how many turns you tie up), and risk-of-reveal.
Before I jump into deep math or strategy tips, could you tell me what you’re mainly after?• A rules-lawyer’s comparison of expected mana & action costs?
• Tactics for players (“When do I go MENDA, when do I go PPA?”)?
• GM-side guidance (“How do I balance encounters knowing my NPCs use one system or the other?”)?
• Something else entirely, like flavor text or world-building implications?Let me know where you want to steer the conversation and I’ll dive right in!
Other models, same or similar prompt
Responses were produced in different, uncontrolled contexts (system prompts, settings, prior turns). This is not a benchmark.