Hiram McDaniels wrote:I admit I’m not very familiar with fatal's core mechanic.
I can't remember if this is from FATAL or some other awful system, but - first roll 1d100 to find out the probability to succeed, then roll 1d100 and see if it's below that to tell whether you did. That is, flip a coin in an overly complex way.
Hiram McDaniels wrote:What I’m wondering is what a bad output looks like, and what a good output entails.
A system has specific goals*. Those vary from project to project, but it probably includes at least trying to emulate a specific genre of fiction, resolve uncertain outcomes, and align everyone's expectations for what happens within the shared world. The outputs are what you get from the system by following its rules, and how good or bad they are depend on how well they meet the design goals.
For the standard goals, a "worse than magic tea party" system would presumably:
- Do even worse at emulating its genre than an average GM can make up on the spot.
- Resolve uncertainty even worse than a human attempting to make up random numbers (wow). You could do that by breaking your RNG so the answer is always obvious.
- Do worse at aligning expectations than trying to guess what's going on in the GM's head. You could do this with prose that's misaligned with the mechanics, or make the mechanics so opaque that nobody knows what to expect will come out the other end.
* I mean, hopefully. Having design goals is a prerequisite to having an actual design. Otherwise it's just someone's favorite mechanics in a blender.