Friday, June 3, 2022

Thieves in AD&D: variant rule


Most of the criticisms of the Thief that I've read aren't conceptual, but mechanical: it's not that sneaking and finding traps aren't cool or worthwhile, it's that (O/A)D&D Thieves are *bad* at these things, especially at low level.

Intuition tells me that the root of the problem is using % dice for thief skills, in conjunction with low starting values for all the skills.

I wonder whether a dice pool mechanic might not be a better way to model expertise. Something like the following five Thief skills, each starting at one die (d6):

Camouflage: exist/sneak without your presence being detected

Forgery: create documents, uniforms, etc. that look like the real thing

Hands: delicate manipulations such as pickpocketing, card magic, lockpicking, disarming fine traps

Perceive: notice large traps, hear clues, oppose Camouflage

Savvy: predict NPC actions, detect social manipulators, have already prepared for a given situation.

(Maybe a Climb Walls skill somehow too, or is that just Hands?)

I'm imagining that you get one extra die per level to add to one of these skills, or maybe two at high level. Tools might give you additional dice or subtract dice if they're substandard.

Example usage:

Picking a crude lock jury-rigged out of iron by a goblin blacksmith requires Hands 1:5 (read: at least one 5-6 in your dice pool). A thief who chose to specialize in Hands (2 dice) and has decent tools (no penalty, no bonus) would roll two d6s. If either is 5+ (55% probability) he picks the lock, otherwise that lock is beyond him until the situation changes (such as gaining more Hands skill or consulting a mentor about goblin locks). At 2nd level, he could bump Hands to 3d6 (70% chance of success on the crude lock) or invest in a different skill.

A lock in town might be harder, 2:5. A really impressive tamper-resistant lock might need higher rolls, such as 2:6 or even 3:6.A Hands 3:6 lock is impossible for a novice thief to pick, and tricky even for a master thief with Hands 10d6.

Similarly, forging a simple travel permit within the Hobgoblin Empire might be Forgery 1:5 (easy for anyone who's ever seen one) while access to a classified military location might require Forgery 3:6 documents.

If the PCs get thrown in a jail during a heist, "I've already bribed this guard with money. He's supposed to drop a copy of the key on the floor" might require a Savvy 3:5 check, but if the DM decides that this guard isn't corrupt, and money won't work, he could say "roll Savvy 5:5 instead" and on success, tell the player, "You found that this guard isn't corrupt and won't take money, but he's desperate to find his missing daughter [adventure hook]. You promised to help with that if he helps you. He drops the key on the floor just as you had planned."

I imagine that instead of various races getting +/- 5% to Hide In Shadows, etc., elves might get * or ** to Camouflage while halflings get * to Perceive, where each * means "you can add +1 to a die of your choice". Thus, Camouflage 2d6* might roll 4, 5, which you can use to satisfy Camouflage 1:6 (4, 6) or Camouflage 2:5 (5,5).

I think by boosting the low-level success rate and reorganizing the skills, Thieves would have a unique niche that isn't easily duplicated by a wizard with Knock and Invisibility. Having a Thief in the party would feel less like a tax begrudgingly paid (or ignored in favor of another priest to offset trap damage) and more like a role that brings an actual benefit to the party and is fun playing.

At any rate, this is about what it would take to get *me* to want to play a Thief sometimes.