Sorcerers were introduced to D&D in 3rd edition. At the time they could be best described as “Wizards, except they use Charisma instead of Intelligence, and they use spell slots to dynamically cast spells from a small static pool, rather than casting from from pre-defined loadout of spells from an indefinitely large pool”. Oh, there was some noise about where they got their apparently-innate magical ability – something about possibly being descended from dragons – but if so that didn’t seem to make any difference to their abilities – they were still learning and casting the exact same set of spells as Wizards.

I can’t speak to how Sorcerers were handled in 4th edition, but now in 5E they’re…mostly unchanged. They now have a choice of two power origins by-the-book, with each giving very minor abilities, as well as a pool of Sorcery Points, useful mainly for applying metamagic effects to spells. Metamagic effects may seems like a big step up in distinctiveness, but Sorcerers used to be the premiere spontaneous casters in 3E – Bards cast spells spontaneously as well, but they started with only 0-level spells, topped out at 6th, and had a fairly restrictive spell list. In 5E Bards start with 1st level spells and go up to 9th at the same rate as Sorcerers, and they can make a few selections off any spell list. And of course they use Charisma as well, so the Sorcerer seems a bit of an uninspired copy of the Wizard.

I think we can change this without totally redefining everything about Sorcerers – they just need a feature that helps set them off a bit both from both Wizards and Bards. Wizards are the easier place to start – the easiest way to create real differentiation here is to differentiate their spell lists a bit. My suggestion is to do this by leaning into the notion of the Sorcerer having an inherent link to power – they can’t learn certain things a Wizard can, but they can also use powers a Wizard simply doesn’t have access to. We’ll do this by giving them affinities. Let’s start by doing this for 3rd edition Sorcerers, then modifying the definition as necessary for 5e.

Affinities for 3rd edition Sorcerers

An affinity is a spell descriptor as listed in 3rd edition (the online System Reference Document lists these 19: acid, air, chaotic, cold, darkness, death, earth, electricity, evil, fear, fire, force, good, language-dependent, lawful, light, mind-affecting, sonic, and water. A Sorcerer has 1 affinity for every level of Sorcerer spell they can know (so 2 at 1st level, +1 at every even-numbered level through 18). A Sorcerer’s spell list – the list from which they can select spells known – is all spells whose descriptors match at least 1 of the Sorcerer’s affinities, regardless of what other lists those spells may or may not be on. A Sorcerer has no access to spells that don’t match their affinities, but can cross all other spell lists in learning spells that do match an affinity.

A Sorcerer may learn any spell on their spell list that they’re high enough level to cast, as usual, and they can cast spells known as usual. However, a Sorcerer can also cast a spell on their spell list that they don’t know if they’ve seen it used at least once, but doing so is effectively a metamagic effect that increased the level of the spell slot needed by 1. So a Sorcerer with Fire affinity who doesn’t know Fireball could cast it using a 4th level spell slot.

Affinities for 5th edition Sorcerers

This is basically identical to 3E Sorcerers except for the following:

1. The effective rate at which affinities are gained is 2 at first level, +1 at every odd-numbered level through level 17.
2. Spells in 5e don’t list any descriptors. Look up the descriptors from 3e spells; where there’s no corresponding spell in 3e the DM and player should agree on what the descriptors ought to be (these will mostly be pretty obvious).
3. A Sorcerer can cast spells they don’t know but which are on their spell list by spending 1 Sorcery Point. They can trade out one of their Spells Known for another one on their spell list by spending 2 Sorcery Points; this replacement lasts until they take a long rest.

Analysis of this change

Sorcerers will now have very different spellcasting abilities not only from members of other classes, but also from each other unless there’s a large overlap in their affinities. With 2 choices from 19 options , a 1st-level Sorcerer may have any of 161 different spell lists; by the time the Sorcerer receives their 10th and final affinity the total number of possible spell lists is 92,378! However, these numbers are a bit overblown since some descriptors overlap (language-dependent is a strict subset of mind-affecting; it might be better to change it to language-related?), and since some descriptors are basically used to describe the same spells with mostly cosmetic differences (the alignment descriptors tend to work this way…there isn’t much point in picking up more than 1 alignment affinity).

Even taking these considerations into account there should be quite a bit of variety amongst Sorcerers; the young would-be Queen of Air and Darkness should play fairly differently from the specialist in Plants and Death. There are a few other things to keep in mind: some affinities are almost certainly much better than others. This may not matter as much at high levels where the Sorcerer already has a large number of affinities, but at low levels trap option affinities are a real concern – the DM should be flexible if player asks to change their Sorcerer’s affinities because a choice they made at low levels isn’t as much fun as they thought it would be. Since the value of affinities will change over time as new spell effects are published and incorporated into campaigns (even for established Sorcerers due to their ability to cast spells they don’t know at a slight premium) a bad affinity might also be salvaged by adding a few spells – but an already-strong affinity can also become overpowered. This still shouldn’t be as big a problem as adding new Cleric or Druid spells to the campaign.

The biggest drawback to all this potential variety is that you have to build your Sorcerer’s spell list by yourself – there’s too many potential lists to predefine them. You can try filtering on this Google Spreadsheet I put together based on Andargor’s 3.5 database (though if you’re comfortable with the database formats Andargor offers you might prefer to work directly with them). Assuming you’re happy working with tools like that you have the flexibility to add new descriptors; you can also try finding out how various descriptors compare to each other (Mind-Affecting offers far more spells than any other, and a simple majority of spells have no descriptor).

In fact, this last discovery is probably the single biggest problem with the whole proposal – most descriptors just don’t cover very many spells. I suppose we could allow 3rd edition subschools (like Summoning and Compulsion) to be used as affinities, but that still leaves the original descriptors out there looking better than they actually are, and they really do seem like they should be better – that they should cover more spells, but I don’t see a good solution to this other than reviewing hundreds of spells to see that descriptors are more common. Note that the Sorcerer spell list in 3.5 core books comprises some 375 spells; about 159 spells in 5e. Affinities that only encompass a dozen spells (most of them at mid-to-high level) aren’t a good fit for this system; sadly that appears to be most of them, so this is probably a non-starter.

Featured image is from extremelyshysuzu.tumblr.com.