Bill Zebub
“It’s probably Matt Mercer’s fault.”
One of my tests is can the GM always state firmly what percentage chance of success they are giving. Not so much because I think you need to set all the stakes retroactively based on the intended chance of success, but because if the GM doesn't understand the system and math well enough to quickly compute those, then chances are they are capricious and random without intending to be so.
I think we are deriving different conclusions because we are starting from dramatically different preferences.
I don't care if a GM can "firmly state percentage chance of success" demonstrating some kind of deep mathematical understanding of the system. I want the GM to offer choices and decisions that are interesting, with non-obvious 'best' answers.