The holiday season has left me mentally barren. I expect to recover soon.
In the meantime, I've been considering overhauling the basic underpinnings of magic use in my games, both wizard magic and cleric magic. My main intent is to have spellcasters rolling a d20 to cast a spell, just like players roll a d20 to make an attack. Since my main desire is some sort of equanimity, that means the higher the die roll, the better the spellcast.
Just like any system-fiddling, the first question is 'why bother, it works now.' Simple - I don't really like how it works now. I want my spellcasters doing magic. I'd rather it was more magic with less surety than less magic with more surety. I think that makes for a more interesting situation. Would it be any fun to play a fighter if you knew that you were guaranteed to do damage when you tried to hit.. but that you could only try to hit something once or twice a day? I'm thinking no.
I also have no interest in a wizard rolling 'to-hit' with their magic missile. Lame. I'm looking for something more interesting than that. I'm also thinking mostly about combat-related or time-sensitive magic. To be honest, if you can cast detect magic I don't really care how many times a day you cast it. Conversely, you probably don't care if it takes you 15 to 20 minutes to cast it. I think that's a pretty good trade-off, but I'd still like to see a d20 roll anyway, perhaps with a much easier target number, but with a chance to screw it up and make something bad happen, or even at least waste your time.
I love examples!
Charm Person: failure results in the intended target becoming insensibly enraged at the casting magic-user. They will immediately do their best to kill the magic-user, disregarding all other considerations to do so, including their own safety.
Detect Magic: failure results in a roll on the sensory mishap table.
Feather Fall: failure results in a roll on the minor physical mishap table, with all results visited upon the mage in the form of feathers and/or down.
One thing I really like about the take on this kind of system from James of A Dungeon Master's Tale is how (basically) target numbers get harder the more frequently you use magic. For what I'm after, the tables would look a lot more like one of Zak's tables, and the difficulty of successful spellcasting wouldn't increase quite as steeply as in James' system.
I'll keep thinking about this.. please throw ideas at me.