Saturday, September 7, 2019

A/X: Fibonacci Spell Points

I've been thinking about spell points for A/X.  Originally I was thinking of rolling the resource model for spells all the way back to "spells are per adventure", but 1) this is a somewhat dissociated resource model which would require in-world explanation and induces weird edge cases around "well what's an adventure", and 2) I'm thinking that if HP and mercenaries can be recovered (slowly) in the wilderness in refuges, perhaps it makes sense for magic to also be recoverable in the wilderness (slowly).  But spell slots of particular levels are very quantized; doing slow recovery without breaking them up sounds like it would end up being complicated.

At 1d3 HP per day of bed rest in reasonably sanitary conditions, an average first-level fighter can recover full HP in about two days.  So putting magic at parity, it would make sense for the average first-level mage to be able to recover a first level spell about every two days.  This means first level spells need to cost more than one spell point - two seems workable.

This got me thinking also - most spell point systems make 2nd level spells cost 3, 3rd level spells cost 5, 4th 7, 5th 9.  This is pretty close to reasonable except at the very low end.  Is web really worth three castings of sleep?  Is fireball worth five castings of sleep?  I tend to think not really, with OSR sleep - a fireball wins a wilderness encounter with a band of beastmen by cooking their chieftain, two or three sleeps win an encounter with beastmen by knocking our the chieftain and a couple of gangs.  The main thing these higher-level spells have going for them is action economy.  So thinking about these factors, 1st level spells costing 2 points and the relationship in power between spells of various levels, I ended up at the Fibonacci sequence.

Spell level: spell points
1: 2
2: 3
3: 5
4: 8
5: 13
6: 21

A fireball once every five days is also pretty close to the recovery time for a badly-injured retinue unit (or for the beastmen you fireballed last week to recover from their burns), which suggests that it's about right for recovery of wilderness-level abilities.

Further considerations:

If NPCs follow the same rules, what does this do to the economics of hired spellcasters?

What does this do to the availability of Restore Life and Limb?

Can you recover spell points faster while resting in civilization, where you have all your incense and pentagrams in order beforehand?  Can you apply these same sort of things (building sanctuaries, consecrating altars) to improve refuges and boost spell point recovery there?

What of the Contemplation proficiency?

Mercenaries only recover partway (the dead don't rise, but the injured heal) - should there be some analog to that?

Maintaining indefinite-duration spells like continual light and wizard lock - reduce your max spell points?

Places of Power seem like a much more natural fit for spell point systems than for vancian casting - reservoirs of spell points that can be drained but recover slowly.

Are there known pitfalls to spell-point systems that I should be wary of?


  1. In my experience with spell points for 3.x (via psionics), compared to vancian casting, spellpoint casters are generally allowed and incentivized to blow all their daily budget on a small number of high-level spells. We found that, in general, this made them more effective but less interesting than vancian spellcasters. This is pretty embedded in the late-3.x paradigm of "small number of difficult encounters every day, then rest and everything bad goes away", admittedly.

    1. > This is pretty embedded in the late-3.x paradigm of "small number of difficult encounters every day, then rest and everything bad goes away"

      That makes a lot of sense. I could still see it being an issue during wilderness adventuring in OSR games, since wilderness encounters tend to be big and dangerous. I was thinking that slowing the recovery model might help, but thinking of a wilderness adventure as a series of (difficult) encounters with limited recovery between them, I am not so sure. Perhaps restocking of monsters between retreats to civilization is the key here, much as it is in the OSR dungeon.

  2. Just to stir the pot here a bit, a different approach could work here as well; shift from 1d3 hp regain per rest to disadvantaged d6 and have magic cast from hp. Take your Fibonacci advancement system and generalize it to cover things outside of magic so you have a ‘push yourself/gamble my safety’ system attached to up actions.

    Unrelated, I’ve been subscribed to you for a while, keep up the good work

    1. Hmm, interesting, but I think there might be some cascading consequences (things like making sure there are no healing spells that restore more HP than they cost to cast, to prevent infinite casting loops), and I should probably make only the minimal changes necessary to make wilderness adventuring work well. Even if I don't end up pursuing that rules proposal, your comment has provided me with insight on the scope of my project ("defining what a project isn't is just as important as defining what it is"), and for that (and the subscription and compliment) I thank you! Also I enjoyed your post on Pullman's work; I hadn't considered the episodic/transient nature of aspects of folktales.

  3. As far as I can recall in my times using them, the major "pitfall", if it can be called that, is just the sheer flexibility - exactly what you reference with a fireball not being worth 5 castings of sleep.

    I had a 3E wizard back-in-the-day using spell points, and it was a bit terrifying. And that gets back to the style of the game, really - I'd be much more worried about running a spell-point caster in 5E, let's say, vs any given OSR variant, as the expectations of encounter depth/frequency can vary (and I guess I'd be mostly looking mostly at published modules here).

    Tying together time to deal damage vs heal damage is interesting, though.

    There are a few OSR variants that use spell points that I believe tie speed of spell point recovery (for regular rest) to location - I want to say NGR has it recover faster in the wilderness (something about Law vs Chaos). That being said, I've always wanted wizards/etc. to be more powerful and dangerous when in their sanctums/etc., and what you propose would be a step in the right direction.

    1. Yeah, I agree that spell points are probably more dangerous in more modern games, given that expectations are better-defined.

      I was tempted to have spell points be more available in the wilderness on that Law and Chaos basis, yeah, but I think that would throw the resource model for wilderness adventuring all out of whack. I could see making extra spell points available in the wilderness in exchange for corruption ('tapping wild magic inimical to man') or other hazards; that could be pretty interesting.