OED: Spell Changes

In addition to the section on PC Generation, the OED v0.4 rules also added a section on Spell Changes, in 3 parts. In particular:

Sleep: Roll 1d6+1 for total hit dice affected (no figure over 3HD). Duration is 12 turns; slapping/shaking awakens 2-in-6 per round.

You'll see a lengthy analysis arriving at this a few posts back. I made the die roll 1d6+1 for elegance sake (it neatly results in exactly 1-3 creatures of 2HD, for example).

Missiles: For missile spells (fireball, lightning bolt), announce a range, and then roll 2d6 over/under for exact location (read lower die, ties indicate on target).

Now we're getting into the material that you might keep unmentioned until a player throws their first fireball and discovers this during play. This is originally an optional rule from Chainmail for catapults/field guns. (The fireball and lightning bolt in those rules, of course, simply reference back to the effect of such catapults/field guns.) Therefore I think it's both legitimate under the original rules, and scratches an itch of mine that the fireball/lightning combo are generally used with suspension-of-disbelief-shattering accuracy.

If there's any question about the rule, it's this: Announce a range to fire the missile-spell (say, 10”). Roll two d6 of different colors (say a red one for “over” distance that comes up 4, and a white one for “under” distance that comes up 2). Apply the result of the lower die (in this example, 2” under the declaration, placing the spell 8” from the caster in the desired direction; if the dice were tied then the shot would be exactly 10” away).

Permanence: Spells under 5th level cannot be permanent. Those without a listed duration fade after d6+6 weeks (such as charms, continual light, etc.)

Another itch of mine being scratched. There's a whole lot of mischief that can be done if you permit low-level spells to be truly, indefinitely, permanent. Charm person allows someone to gain infinite followers over time. Continual light allows those magic-street-lamp cities (yuck). Wizard lock may as well be used by cantankerous wizards on their downtime, locking every door they encounter forever, just to be nasty.

You can also think of this as a generalization of the principle introduced in Supplement I: Greyhawk, where charm spells are given additional saves to break, assessed over some weeks of time. Why not apply the idea equally to other spells where indefinite duration allows silly mischief?

This does, however, still leave the door open to higher-level permanent spells, and we should carefully consider if there are any game-breakers in the bunch. I feel that the 5th-level spells provide in-milieu limitations: wall of stone can be dispelled (so a wizard would prefer to have a real wall constructed) and animate dead has obvious drawbacks (collecting wagonloads of bodies for an undead army invites the wrath of more heroic adventurers). The 6th-level spells need more careful handling: invisible stalkers should get more and more perilous if used in great numbers; geas spells should have some drawbacks or risk if used with abandon (c.f. the works of Vance, for example).


  1. That chainmail die system has the weird problem of never being exactly 1" off target. Also, how do you adjudicate 'long' hits? E.g. will a fireball really miss and detonate 1" behind a Troll?

  2. Also this seems as good a time as any to say that I like your defaulting to 2-in-6 chance for many of the mechanics. That cuts through a lot of the subsystems in a way that "lookit this chart of modifiers/DCs" does not.

  3. "That chainmail die system has the weird problem of never being exactly 1" off target."

    I guess I should point out that what the Chainmail text literally says is "take the higher", while what I have here is is "read lower die". Taking the higher die is so batshit crazy that I'm convinced it must have been a typo. As you say, it generates no 1's, and even worse, a completely non-Euclidean probability curve. Taking the lower die results in a very clean bell-shaped (or technically triangular) curve.

    Sure, I'd adjudicate long rolls just the same -- the fireball pellet is presumably taking a ballistic (parabolic) trajectory just like the catapult shot that it's based on. Roll over-1 and under-4, it's shot 1" past the troll (thereby catching the troll in the 2" radius blast).

    Thanks for the other comment!

  4. ... although I really can't take any credit for the 2-in-6 mechanics, since they're all directly out of OD&D. Maybe the one thing I'm doing is collecting them in one place to show their commonality, and suggesting some modifiers to them.

  5. @Delta,

    Wouldn't simply adding both dice together provide the necessary 'spot-on' result while still allowing for a wide variance of location?

    i.e., +1,-1 = 0 exactly as intended, while any non-doubles result would determine the placement, with the higher roll dictating short or long?

    As usual, *confused*

  6. Wouldn't simply adding both dice together provide the necessary 'spot-on' result while still allowing for a wide variance of location?... As usual, *confused*

    The idea is mathematically equivalent. However, algorithmically (1) Most people find adding negatives (subtracting) more difficult than identifying the minimum, and (2) I always defer to original-book rules whenever possible (perhaps with typo corrections, as here).

    It's similar to the fact that there's 16 different equivalent ways of expressing the D&D "to hit" relationship -- and after a great deal of analysis I'm confident that the "add everything, beat 20" process is simplest for people to perform mentally.