Thursday, March 12, 2009

Skills: The Middle Road

Rules for non-combat skills in D&D have spanned a rather broad course over the years. In OD&D, there were no rules for non-combat related skills other than magic until the introduction of the thief--a point that most grognards put in its favor, since it encouraged everyone to try their hand at everything. However, this lack does present problems when trying to determine exactly what a character can do that's better than other characters, and as more classes were created to cover these non-combat niches (first thieves, then rangers) the implication set in that unless your class abilities said you could do a thing, you couldn't except by DM fiat.

AD&D attempted to fill this gap someone with randomly-determined secondary skills as an optional rule. However, it gave no real understanding of how a DM was supposed to adjudicate the use of such skills. Later guides introduced the concept of non-weapon proficiencies, which were enshrined in 2nd Edition, but which never really scratched the itch that other games systems had already relieved due to their clunky nature. In fact, the NWPs actually narrowed the possibilities for players by placing in our minds, however unintentionally, the idea that if you lacked the right NWP, you couldn't do an action.

3rd Edition went a different route, implementing a fully-developed skills system that at first glance has a marvelous elegance of design. However, it does have a singular flaw: Advancement in said skills is wholly dependent on class level. At low levels this does not overly imbalance the game, as an unarmored fighter with a halfway decent dexterity has at least a decent chance of sneaking past a pair of guards, even if the rogue is better at it, but at higher levels this rapidly breaks down, until characters are compelled to keep certain abilities maxed out at all times in order to just have a chance of detecting that pit fiend sneaking up on you. In the end, the creative use of skills is still stymied.

4th Edition attempts to fix this problem by having every skill advance with level; "having" a given skill simply provides a +5 bonus over those who have no skill in it. However, in the process the game has sacrificed any sense of customization on the altar of balance.

The extensive rules for dealing with non-combat skills in other systems, both classic and modern, speaks of the desire of players to be able to know in some quantitative sense what their characters are good at. However, if we are to come up with any such system as a house rule for OD&D, it has to meet several basic parameters:
  1. "Having" a given skill, NWP, or whathaveyou should not, as a rule, be a requirement for attempting any adventure-related action or for having a reasonable chance of success.
  2. The system must be scalable, allowing for characters to improve existing skills by the expenditure of time and wealth or as a reward for a successfully-completed adventure (as described in my previous post),
  3. And yet it must be simple enough that no OD&D product must be significantly altered to employ it (i.e., the referee should not have to go through every adventure and install DCs for every challenge or create a complete set of skills for every goblin) and that any referee can easily ad-hoc it during play.
With these guidelines in mind, here is the house-rule that I have created for non-combat, non-magical skills:

There are four levels of competancy for any given skill: unskilled, skilled, expert, and master. All characters are assumed to be unskilled at any given task unless it falls under their class (especially in the case of thieves) and/or background. Having a background that encompasses a particular action means that one is skilled only, since it is assumed that the character left his old apprenticeship to pursue the road to adventure before surpassing a journeyman's skill level.

Of all character classes, only thieves automatically advance in skill levels as they increase in character level, and then only in those areas directly related to thievery (i.e., the classic theif skills of picking pockets, hiding and moving silently, finding and removing traps, picking locks, climbing, etc). In all other cases, advancement or gaining new skills must come as a result of gameplay, typically purchasing training or obtaining it as a reward by some individual or group.

Before explaining further, a particular mechanic must be described. The skill die denotes the type of die rolled for a given skill level, to which is added the appropriate ability modifier (using the Moldavy scale) when a skill contest arises between two individuals. For example, a 1st level thief with a 16 dexterity is trying to sneak up on a 1st level cleric with a 15 wisdom. Since the thief is considered skilled at moving silently, he would roll a 1d8+2 (modified for dex) against the cleric's 1d6 (unskilled) +1 (wis) to determine if he succeeds.

The referee can apply additional ad hoc modifiers as he sees fit; for example, the thief might suffer a -2 penalty trying to cross a carpet of dry leaves or gain a +1 bonus trying to cross the soft carpet of the cleric's church (or, conversely, suffer a -2 penalty for the fact that he's attempting to assasinate a cleric in a church).

Regardless of modifiers, a character always succeeds in this type of contest if he rolls his maximum possible result against an opponent's natural 1.

In cases where there is no contest between two statted characters--for example, a thief sneaking up on a goblin whose wisdom score is unknown, a character climbing a cliff, or a character with a jeweler background trying to make a gift to impress a noblewoman--the referee should assign an ad-hoc possibility for success. This can be a number that must be surpassed by the roll of the skill die (the goblin is generally facing the direction the character is approaching from, so the referee rules the thief must roll a 10 or better to sneak up on him), a percentile chance, or any other sort of roll.

In the latter two cases, the character's skill level grants a bonus to attempting the action. For example, the referee (or module) has decreed that any character has a 50% chance of tracking the fleeing goblin through the dusty room, but a character who is an expert at tracking would have a 75% chance of doing so (50% plus a 25% bonus for the expert skill level). In this way, modules and other OD&D-compatible materials do not need prior modification before this house rule is used.

A character can be trained to a skilled level by either an expert or a master, and to an expert level only by a master. Achieving mastery takes long hours of self-study (hence the long training period). Each level of skill must be attained in order; e.g. to go from unskilled to expert takes 4,000 gp and four months, not 3,000 gp and three months. The costs of training may be varied or even waived by the referee depending on circumstance (as in the case of the grateful rangers), but the required times should be considered a minimum, and preclude adventuring or any other lengthy travel. Note that this system works best of the referee is awarding experience points based on the expenditure of gold rather than it's acquisition, since it forces players to make valid choices in terms of advancement, but can be made to work with any system of awarding experience with a bit of fiddling.

Skill Die: d6
No bonuses to die rolls

Training: 1 month and 100o gp
Skill Die: d8
+1 bonus when making skill-related rolls with a 1d4, 1d6, or 1d8
+2 bonus when making skill-related rolls with a 1d10 or 1d12
+3 bonus when making skill-related rolls with a 1d20
+15% on making skill-related rolls with percentage dice

Training: 3 months and 300o gp
Skill Die: d10
+1 bonus when making skill-related rolls with a 1d4
+2 bonus when making skill-related rolls with a 1d6 or 1d8
+4 bonus when making skill-related rolls with a 1d10 or 1d12
+5 bonus when making skill-related rolls with a 1d20
+25% on making skill-related rolls with percentage dice

Training: 6 months and 10,00o gp
Skill Die: d12
+2 bonus when making skill-related rolls with a 1d4
+3 bonus when making skill-related rolls with a 1d6 or 1d8
+5 bonus when making skill-related rolls with a 1d10 or 1d12
+8 bonus when making skill-related rolls with a 1d20
+40% on making skill-related rolls with percentage dice

In cases where multiple dice are used, use the upper limit of the die roll to determine the correct bonus. For example, an expert who is performing an action whose results are determined using 2d6 will get a +4 bonus to the roll. In cases involving combinations of dice with a deep bell-curve (e.g., 3d6), the referee may adjust the bonus downward to preserve the curve. For example, our expert might get only a +4 bonus instead of a +5 on a roll of 3d6 at the referee's descretion.

While certain types of skills are suggested in the background table or canonized by the thief class, no fully codified list of skills exists. This lies in the domain of the referee, with player input. Likewise, there is no list of exactly what each skill might accomplish. This is up to each group to negotiate among themselves, thus preserving the "free-wheeling" aspect of OD&D.


  1. It does not appear you update your blog anymore, but I think this skill system is brilliant. I am working on an alchemy document and plan on including the basics of this system (crediting you and your website) within this document. I have also included this post in my 'required OSR reading list'.

  2. This really is genius work. It's a pity you're not writing any longer, Benjamin David. I'm working on a simplified version of this for my LotFP game.

  3. Another person checking in. Found this through Courtney's site while trying to develop a skill system for my own game. This is brilliant. Subscribing on the off chance you write anything else again!