The first spellcaster relied on the Swords & Wizardry saving throw. The second relied on reaction rolls, very loosely based on Chainmail. This post will talk about using percentile dice like a Thief using his or her skills. Although I written about it previously, the last post in the series will use the Turn Undead table to determine success.

A simple way is to set a percentile chance of casting spells is to each spell level. Set an 85% chance to cast a 1st level spell and subtract 10% for each level after that. Yes, this means that there is a base -5% chance to cast a 9th level spell; there will be more on that later. The 85% number comes from the Thieves' chance to Climb Walls at 1st level.

To make things interesting, there will be two modifiers to the roll. One modifier will be based on the spellcaster's level, the other based on Intelligence. For each level of the spellcaster, the chance to cast the spell increases by 5%. For example, a 1st level spellcaster has a +5% chance of success while an 8th level spellcaster has a +40% chance of success. As far as Intelligence, a spellcaster with an INT of 15 or more gains a +3% chance of success. If the INT is 8 or less, the modifier is a -3% to the chance of success.

Again, the purpose of these classes is to provide tweaks, not rewrite spellcasting classes, so the spell table is still the main method of limiting this class' power. This prevents higher level spells from being cast too soon as compared to other classes. A 2nd level spellcaster without a spell table limit would have a 85% chance of success to cast a 2nd level spell. For that matter, the same spellcaster has an almost two-thirds chance to cast a 4th level spell. So, the spell table remains the limiting factor.

Despite rolling percentage dice, this class is only slightly less reliable than a standard Magic-User. As such, I think it would be fair that this class would lose a spell slot on an unsuccesful roll. This makes them even less reliable, though the risk at higher levels is still pretty small. a 7th level spellcaster can automatically cast 1st and 2nd  level spells. There will be a 90% chance to cast a 3rd level spell and an 80% chance to use their lone 4th level spell slot. Those odds are still pretty good.

Like any unreliable spellcaster, there's a good chance that he or she will attempt to find ways to guarantee success. The skill-based spellcaster makes temporary foci to guarantee success. The Chainmail(ish) spellcaster uses amulets to gain an advantage to cast a spell successfully. Neither of these classes, however, risk losing spell slots. If the magic focus or amulet fails, the spellcaster can try again, even if they don't have any more magic aids to boost their chances.

To make them different from the other two classes so far, let's have this class strive to save the spell slot, instead of increasing the chances of success. It's easy to think about increasing success as somehow stablizing the magic required to perform a spell or increasing the raw magical power being manipulated. The spell slots, though, measure capacity.

If you imagine a spell as being a semi-living creature crawling around in a spellcaster's skull, saving the spell slot can be compared to a trap that snatches the spell back in case of failure. A different way to think about it would be like the safety on a firearm. The spell has to successfully turn off the safety to be cast, but if not, no spell is used up. The spell is still in the chamber, ready for another try.

Stability still requires a cost, mechanically and financially. I'll stick with the tried and true 100gp per level of the spell slot to be saved. Mechanically, the cost will be a -5% modifier to spell success. The physical representation of the safety could be something ingested by the spellcaster before casting the spell. I have not read Brandon Sandersons' Mistborn series, but if you are a fan, this could be used as a basis to make Allomancers. Instead of metal, the spellcaster could swallow powdered gems, or just about anything else. For ease of reference, I'll call these ingestible items Mnemonics.

So this class has four tweaks:

  • Base percentile chance to cast a spell based on spell leve, modified by spellcaster level and spellcaster Intelligence.
  • If the roll fails, the spell slot is lost unless...
  • ... the spellcaster creates Mnemonics that are consumed so that the spellcaster keeps the spell slot in case of failure.
  • Mnemonics cost 100gp per spell slot level and lower the chance of success by 5%.

All in all, not too bad. Because of the ingesting, it is easy to think of this class as an alchemist. If you allow a standard Magic-User to create scrolls in your game, you should allow this class to make potions of their spells.

Fireball spells can be like dragon's breath, or the tiny bead of energy escapes the mouth, your preference. Charm spells can show themselves as songs (or not if you really really hate bards). There's all kinds of creative ways to have a spell come from a potion instead of a scroll. How about the Magic Jar spell or even the Prismatic Sphere?

Wrapping up, here is a spellcaster that uses a familiar mechanic (rolling percentage dice like the Thief) with two tweaks, that turns out to be a type of potion-making spellcaster. The spells may be the same, but he or she will play differently from the spellbook toting standard Magic-User.

Next time, the Turn Undead table.