Spells without levels

The title of this post reflects a recent design ambition of mine, which is to eliminate spell levels from OSR games without eliminating spells.  I have a few reasons for this.  First, as I've been recently discussing, I have an interest in sandbox play without character progression, and having spell levels starts to feel overly restrictive when characters don't even have levels.

Another thing I've noticed, however, is that the nature of spells and spell levels in OSR makes it difficult for such games to support multi-classing.  At first glance, it may sound like I'm talking about very unrelated things, so I'll explain a little more.  

Don't make me choose, OSR!

The problem

The essential problem with multi-classing in D&D-type RPGs is that a character improves more by advancing a class from level 9 to 10 than from 1 to 2.  At the same time, XP costs increase as one gains levels, almost exponentially at certain points.  So the obvious way to incorporate multi-classing, in this situation, would be to require that the character needs XP to gain a level appropriate for the class and level he's getting.  Right?

It's red so you can't see all the bloodstains

Not so fast.  One problem here is that hit point progression follows a completely different curve, which is absolutely linear at a given rate up to a high level, and then rises at a much lower rate after that.  The problem that this presents is that it makes multi-classing a really cheap way for high-level characters to gain hit points.  This is true even if gains remain strictly constant even at high levels.

The solution to that problem would be to have the cost to increase any of a character's class levels based on the character's total number of experience levels.  This, however, presents several problems.  Firstly, there's the fact that most OSR classes don't progress at exactly the same rate of XP; this can be remedied by making the classes relatively balanced, and besides, the standard XP rates aren't necessarily balanced, anyway.

But the other problem is more fundamental, which is the fact that, for most classes, the power curve makes higher levels more valuable.  This means that it would be a waste of time, usually, for a tenth-level wizard to grab a level of fighter, since he'd get a lot more bang for the buck by grabbing another level of wizard.

The way to solve this problem is pretty straightforward - keep the value of each experience level pretty constant throughout the entire progression.  And that's where OSR magic-users are a real pain in the ass.

He mis-cast his Polymorph Hat

Take a look at the tables for various OSR magic using classes, what they get at each experience level.  In BECMI, take a look at the Magic-User.  At third level, he gets a single 2nd level spell.  Wow, welcome to the big time!  By contrast, what does a twelfth-level Magic-User get?  A 2nd and a 3rd level spell!  That's unequivocally better.  The curve is shot.

It's a lot easier to flatten out the power curves for Fighters and Thieves.  Take a look at that favorite of mine, Lamentations of the Flame Princess.  In addition to their hit point and saving throw gains, which are basically constant, the Fighter gets +1 to hit each level, and the Specialist gets two skill levels.  That's already flat as hell.  But even in LotFP, spell-casters are on a different journey.

So can you see my motivation to get rid of or otherwise minimize spell levels?  Even if you keep the number of spells that a magician gets each level to a constant value, the fact that they get access to higher level spells over time means that those slots mean more.  If we want to have multi-classing in OSR that doesn't screw up the power curve, we have to change this aspect.

Possible approaches

I don't think I'm the only person thinking like this.  I've seen other systems, lately, that drop the idea of spell levels, or at least minimize them.  Even if you aren't looking to make it easier to multi-class, we can make less of a joke out of game balance where classes are concerned.

So here are a few ideas that I've come up with.

Why do wizards always wear robes?  They look stupid.
My theory is that they're just lazy.
Robes are the tracksuits of the medieval world.

Spell skill levels

The idea here is to treat spells similar to skills in Lamentations, where characters can invest in different skill levels as they can experience levels.  In this system, the level one version of a spell is a lot weaker than the level six version.

This approach is kind of cool, because each spell becomes kind of like its own skill.  You can gradually do more with it, with greater potency, as you improve over time.  I can imagine that being very appealing for a player.

The downside here is that the progression for spell levels is more exponential than linear.  So we get back to the same concave power curve that we were trying to avoid.  Avoiding this means smoothing out the effects of spells as they increase in level, which then either means that more levels are necessary (making things super-granular), or the upper end of spell power becomes far less impressive.  After all, a 9th level Magic-User spell is a lot more than nine times as powerful as a 1st level spell!

Not a simple production

Casting costs

This is the approach that I'm currently settled on.  The idea here is that spells still have a spell level, but characters are largely unrestricted about what spell levels they can learn, and learning a 1st level spell doesn't take up fewer "slots" than a 6th level spell.

So how do we keep players from stocking up on the most powerful spells in the game?  By making them impractical.  In Empress, we already have the idea that casting a spell has a price, if only in terms of an expenditure of time or physical fatigue.  This means that we can allow the power curve of spells to arc exponentially upwards, as long as we keep casting costs in line with that.

A higher-level magician may reduce the cost of casting their spells gradually, and at a similar rate across spells of all levels.  I think that this shouldn't disrupt the power curve we're shooting for, but I'm not positive about that.

As a bonus, since this approach has the potential to solve the problem of spell level scaling, you can play around with other stuff, like spell skills levels, without worrying about where that will leave your power curve.  All you have to do is make sure that the casting cost is commensurate with the effects of the spell.

Anyway, that's what I have for now.  The real trick is getting the casting costs right.  Empress is basically already doing this, but I have to re-examine the costs in light of a possible removal of other spell level-based restrictions.