Thief Draft

I think it may be an OSR rite of passage to redesign the thief class. Here is my entry.

First, the literature review. I have been heavily influenced by Robert Fisher’s On thief skills in classic D&D, Philotomy’s musings on Thieves & Thief Skills, Matthew James Stanham’s article Thieving Ability (also discussed here). The Jovial Priest has an interesting article, though I decided to go in a different direction (his 2d6 resolution table is too complicated for me, and his “favored skill” system introduces complexity during character creation). Also see Talysman’s recent thief. I’m also becoming quite fond of Talysman’s situation roll, as you can probably tell from my section on skills below. Grognardia’s thiefly thoughts post is also worth a read for a quick summary of why some people think the classic thief is problematic. The LotFP specialist (Grindhouse Edition free Rules & Magic page 10) must also be noted as an important replacement thief. I like how the d6 skills emphasize that all classes can attempt certain tasks (climb, hide, search, etc) even if the specialist is better at them, but I don’t like how those skills take up extra space on the character sheet and use a point-buy system, necessitating choice during character generation.

The Cook/Marsh Expert rules (page X8) suggest some ideas for extending thief skills for higher level play: climb overhangs, climb upside down, ventriloquism, powers of distraction, mimic voices. Some of my thief skills are based on these ideas, and I may work the ones I am not using into a later draft in some fashion (particularly the improved climbing abilities).


  • Thieves are masters of stealth. They are commandos, spies, assassins, infiltrators, and skulks. They solve problems by guile, cleverness, and trickery. They will never fight fairly if they can help it.

Most thief skills consume the resource of time (suggested by the B/X Blackrazor “automatic” thief). Absent distraction or complication, thieves will succeed if they are not rushed (time required is included with the skill description). Wearing any non-light armor (i.e., anything other than leather following the B/X rules) will result in penalties for the physical skills (likely a penalty of one or two on the situation roll, but specifics are by circumstance and referee ruling). I will probably formalize this through the encumbrance system. I use LotFP encumbrance rules, so it should be easy to base the penalty on the encumbrance level (see the free Rules & Magic book, pages 38 – 40).

Thieves use d6 for hit dice. They may use any weapon or armor. Since mobility, speed, and agility are important to thieves for many reasons, they rarely wear non-light armor. I use hit die based weapon damage (with 2DTH for dual wielding or two-handed weapons). So, thieves do d6 damage with weapons. They use the “medium” base attack bonus by level (shared with clerics). Though I’m not a fan of the “combat advantage” DPS super-strikers of 3E & 4E, I do like my thieves to be a bit more capable in combat than they are in the various early rule sets.

Regardless of level, all thieves have the following abilities:

  • Alertness: bonus to hearing noise (2 in 6 chance rather than the normal 1 in 6); this also makes the thief harder to surprise
  • Surprise attack: +4 to attack roll, double damage upon hit (in some ways this is the inverse of the alertness ability)

Skills are the essence of the thief class, and I have not neglected them. I am following my idea of giving the thief one awesome ability per level rather than many poor abilities all at once that develop slowly. This also fits my principle of introducing complexity slowly. I do think it is important to allow the thief to develop as levels are accumulated (this is why I am not following the B/X Blackrazor all-at-once automatic thief). The level advancement incentive is a big part of classic D&D play, and all the other core classes develop (cleric: turning & spells, fighter: attack bonus, magic-user: spells).

Thief skills by level (or roll d10 each level, re-rolling duplicates):

  1. Open locks (requires tools): Mundane locks take 1 turn to open. Complicated locks may require dice (5 in 6 succeeds, on failure dexterity save to avoid breaking the tools). Magical locks may not be opened.
  2. Move silently, 10’ per turn (20′ per turn is also possible but will likely require a dexterity save, or a 5 in 6 success probability, depending on circumstances)
  3. Hide in shadows, may not move
  4. Interpret languages, codes, and maps (5 in 6 success, on failure may only retry upon gaining a level)
  5. Climb walls, 10’ per turn, if distracted or attacked save or fall
  6. Voices: ventriloquism & mimicry
  7. Legerdemain: pilfer, distract, or amuse; can also be used to disable small mechanical traps (5 in 6 success, on failure dexterity save to avoid being caught)
  8. Brew poison: given 1 day and 100 gp, a thief can brew one dose of save-or-die poison sufficient to threaten the life of a human-sized opponent. How larger or smaller creatures react to poison is by referee ruling. 5 in 6 chance to identify and know effects of examined poisons. Other recipes (such as for a paralytic poison) can be found, or can be synthesized based on reverse-engineering an identified poison.
  9. Assassinate: successful surprise attack does damage in hit dice rather than hit points
  10. Use magic scrolls (5 in 6 success, on failure save vs. spells or backfire)

That’s right, if you roll for your skill, you could start with the assassinate or brew poison ability. I like that.

Am I missing any traits or abilities that are associated with the archetypal thief?

27 December 2011 edit: more good suggestions from Jeffro here.
11 February 2012 edit: added poison identification to the brew poison ability.
13 February 2012 edit: added reverse engineering to the poison ability.

8 thoughts on “Thief Draft

  1. The Jovial Priest

    Very nice.
    The summary at the start very helpful.
    The one ability per level adds tremendous flavor and very tempting.
    One suggestion because you call it draft-
    Move silently 10’/turn is very slow and would mean it can’t be used in combat. Perhaps move one third normal speed (120′ normal speed = 40′ / turn silently. Encounter speed 1/3 or 1/4 of normal)
    I’ll add to Links to wisdom.

  2. jeffro

    I like the Dexterity saves on failures– I like there to be some difference for varying attribute scores, but not overmuch. Also… explaining what the failures mean kind of matters– ie, broken tools.

    As to more thief abilities… all I can think of is UA’s Thief-Acrobat, Sherlock Holmes’s disguises, and maybe some kind of fast-talk skill. Maybe some kind of explicit fence access for selling stolen goods…. Urban contacts. I’m probably over-GURPSing it, though.

  3. Brendan


    Yeah, it is a bit slow for combat. I would not be opposed to a slightly faster movement rate. I do see move silently being used more for circumventing conflict though, and perhaps getting off a single surprise attack if things go pear-shaped. Also, the 10′ rate is for guaranteed success in most cases. A faster rate with some chance of failure would also be allowed. Given that combat situations are supposed to be inherently chaotic, that is probably the way I would handle that, I think.


    Thanks for bringing up the thief-acrobat from UA. I don’t have a copy, but I’ll have to check it out. I’ve never played with UA, so other than knowledge of its existence, I don’t have a sense of what the thief-acrobat can do.

    Disguises I would probably allow any class to attempt, though I suppose an almost supernatural disguise ability in line with a magic-user’s illusion would be a reasonable ability.

    The social abilities (fast-talk, fence, urban contacts) I would probably be more likely to handle via direct roleplaying and player creativity. I’m not very good at adjudicating those kinds of things as skills.

    I do have a some thoughts about a version of the monk (and here) which perhaps fits the martial artist and acrobat better, though I’m still on the fence about making classes other than the B/X core seven explicitly available at the beginning.

  4. jeffro

    I agree with the limitations class availability. Some have offered xp for just going to new locations on the big map. I’ve heard of guys running Stonehell that let people come back into the game with a new character starting at the lowest level that is currently being run by the PC’s. Finally… scenarios are the best methods for introducing new rules as systems. I suggest combining all four principles at once:

    New classes can only be introduced to the campaign based on what have been activated on the campaign map. (Hint: Wood Elf, Lizard Man, Bear/Man Skin-changer….) However… the party must first perform a single session adventure/quest for the new race/class/organization in order to allow new first level PC’s of that type to be introduced into play. After the quest is completed, the party has a one-time option to replace fallen PC’s with a character of the new class that will start the game at the lowest level that currently being run in the party.

    (This could be worded better, but hopefully you get the idea.)

    No new class should be created without a corresponding scenario to introduce it. New classes are not an extension of the splat book munchkin menu… but should rather be a perk for players that have lost nth level PC’s.

  5. Brendan


    I love any excuse for a quest; that’s a great way to “unlock” new classes.

    However, I do prefer to always start new PCs at first level. Due to the pseudo-geometric progression of the experience tables, by the time the current PCs have accumulated half the experience required for the next level, new PCs will have reached level N – 1.

    Also, that allows some benefit to be had from elevating a henchman that has been patiently levelling (given the 1/2 share of experience) to a full PC upon a character death. I’ve never actually played in a game where that has happened, but I would like to.

  6. Matthew James Stanham

    That is an interesting approach. One other half-formed possibility that strikes me from reading your “thief magic” idea is to give the thief all the abilities above, but a limit on how many times he can use them in an adventure, perhaps tiered like for spell casters.


Leave a Reply