Design Notes
I've been toying for some time with a class proposal for 5e. I think it might make a lot of different play styles happy and that is what is intriguing me. I think you could add about any flavor you like to the concept so I won't dwell on flavor. I'm mostly going to concentrate on mechanics.
One problem rampant in the edition wars is the disconnect between the way the classes work and the differences of opinion on play balance and the importance of play balance. One common refrain is that wizards in prior editions had too much flexibility. They were too quick to steal the thunder of the rogue via knock or invisibility. The other side though groans at the loss the game takes when such magic is removed. D&D just isn't D&D without the utility magic.
My solution with the Mage is to use a concept 4e originated (at least in D&D) for tracking the duration of conditions. Instead of keeping absolute durations and forcing the chore of tracking everything you instead rolled a saving throw every round. This created an average duration but with some uncertainty. Practically there was a bell curve that represented duration and that worked well. So my idea for the Mage is to do something similar but with a twist.
Class Design
The Mage starts out with five spell points. Each time he tries to cast a spell he rolls a check against a spell DC and if successful nothing changes. If though he does not succeed then he loses a spell point. If the spell points ever reach 0 then the Mage can no longer cast spells. The mage is permitted to recover 1 spell point for each hour of sleep or period of total relaxation. Standing guard or walking is not permitted.
Computing the DC. It might seem at first glance that some linear formula based upon level of the spell vs. level of the caster would be the best way to compute DC. It is not. High-level spells are never easy to cast, no matter your level, and they always come with a risk. This will encourage the Mage to not cast spells needlessly especially outside of combat.
Ok so here is the formula for computing the DC.
Spell level * 4 - caster level.
Roll an unmodified d20 against this DC. The advantage to this system is that you can fully pre-compute the DCs for all your spells. Then when the Mage is ready to make a check he simply rolls against this pre-computed DC. Optionally a roll of one always causes a spell point loss but some may dislike it. No spell over 5th would ever be automatic anyway.
That’s the basics of the class. You could add a material components and give a minor bonus like +1 or +2 for an expensive component. You could give a ritual casting bonus for when you are not in combat. I wouldn’t want it to be too much but +4 might work.
Variations
For a bookish wizard type variation, I would limit the spells known at any given moment to their prime attribute. Thus an 18 intelligence could in theory prepare up to 18 spells. When a Mage qualified for a new level of spell, he would get learn one more and that's it. All the rest he has to find while adventuring. I would though allow the wizard to have as many spells as he wants in his spell book and swap them once per day with the one's he has prepared.
For a more sorcerous variation, I would allow more spells known at level up but no book.
Advantages of this system
If five spell points is the wrong number and even I think it might be a tiny bit low, I toyed with everything between five and eight, then just adjust. I liked eight because of eight hours of rest gets them all back. People wanting more or less magic though could get it my changing that one number.
A Mage never wants to cast a spell if he doesn’t have to because he could always lose a spell point. He won’t be casting knock if a rogue is in the group. But he might if the rogue fails the DC and the situation is desperate.
The Mage has a lot of the flexibility of older editions but not really the staying power. So practically he has eighteen spell choices but depending on the levels prepared he might find his spell points dropping fast if he goes nova very much or over does the utility. I hope this approach would satisfy the utility loving wizard people from 3e while not going so overboard powerwise that the 4e people hate the concept.
