Jack Daniel
Legend
Forked from: How does a game work without skills?
What a perfect segue into this topic. And since I don't feel like hijacking the old thread, I get to use ENWorld's shiny new forked thread feature for the first time! This was posted at dragonsfoot.org and rpg.net, but I'm guessing the discussion will be more germane right here where the subject is being debated!
*****Cross-posted:*****
These rules are intended for the 1981 (B/X), 1983 (BECMI), and 1991 (BB/RC) revisions of the original D&D game.
Classic D&D has an official skill system: "general skills," first introduced in the GAZ series and then included in the Rules Cyclopedia. The system was very similar to AD&D's non-weapon proficiencies. All characters (regardless of class) received 4 + Int mod "general skill slots" which could be filled with skills, and then a new slot was earned at every 4th level above the 1st (so 5th, 9th, 13th, etc.).
I don't like non-weapon proficiency style systems for two reasons. First, the lists are big and unwieldy, without even being comprehensive. There's no way to tell whether a skill is important enough to be on the list, whether it should be its own category, or whatever. Nobody can remember all the skills on the list. Some of them are also unbalanced, since they have clear combat uses (like blind-fighting). 2e handled some of this by making the good proficiencies cost multiple slots, but Classic D&D's General Skills system has no such caveats.
Second, there's no mechanical benefit to knowing a skill. If you know the Deception skill and you want to try and fast-talk somebody, make a Charisma check. If you don't know the Deception skill and you want to try and fast-talk somebody, guess how that game action is still handled? Right. Roll a Charisma check. Okay, sure, there's a rule that says you can spend multiple slots on one skill to get a -1 bonus to the d20 roll for each extra slot, but somehow, I think taking the skill in the first place (and being able to learn many skills) should actually mean something.
**********
So what qualities must a good skill system have? Well for one, the list should be short. Like 4e. Say what you will about 4e, but it has a nice, short, easy to deal with skill list. But the list should also take a stab at completeness. 4e's list isn't great, because it's conspicuously missing rules for simple skills like crafting. Granted, crafting doesn't need much of a mechanic in a minis-based wargame, but it's helpful to have that sort of thing in an RPG!
Part of the reason to have skills in the game at all is to codify a player character's background knowledge---the things he learned before becoming a fighter or a mage or whatever. It helps to differentiate Bob Fighter from Joe Fighter. And in a game like Classic D&D, where there are no paladins or rangers or bards (at least, not before name level), having a "religion" skill or a "wilderness" skill or a "music" skill can help to give players that feeling anyway.
Another quality, vital for any skill system in Classic D&D, is that it should be simple and streamlined. Many, many old-school DMs do not believe in skill systems. They feel that character background knowledge and non-class skills ought to be handled by DM fiat in all cases, or that having rules for such things is too limiting. I am not that sort of DM. I like to have some rules to fall back on. Call it a crutch if you must, but I think that character knowledge really ought to be limited by the rules. Something must be in place to keep the fighter's player from deciding, in the middle of the session, just when it's vital to the story, "Oh, yeah, sure I can make a sword. I, uh, I was apprenticed to a blacksmith in my youth. Yeah, that's the ticket..."
Ideally, such a system should be based on a game mechanic already in place. Classic D&D exhibits several: ability checks, rolled on 1d20 (used for the General Skills system); thief skills, rolled on d%; and many actions are handled by rolling low on 1d6 or 1d10 (used for spotting, listening, and various demihuman abilities). This last mechanic is my preferred choice, for reasons I will explain below.
Lastly, skills ought to improve as the characters gain levels. 2e was notorious for the non-weapon proficiencies never getting any better. 3e is kind of the reverse problem, since some of the DCs are so high that you need to be a high-level character (or stack lots of obscure bonuses) to make them. A static skill system is no good, but then again, neither is a skill system with variable difficulty levels (at least, not for Classic D&D).
**********
A skill system can easily be made to work using a "roll lower than your skill number" mechanic, on any die type: d20, d%, whatever. I prefer to use the d6, because it preserves some of that OD&D feel. The game already makes the baseline assumption that when the player wants to do something tricky, the character has a 1-in-6 chance of success... unless the character is a demihuman trying to spot or hear something, in which case the chance is 2-in-6. That's a perfectly wonderful mechanical foundation for a D&D skill system!
It took a while to arrive at a skill list I liked. At first, I used a longer list, in the area of 24-30 skills. It was a little redundant and clunky. I'd also experimented with shorter lists, 6-8 skills. It didn't do enough to differentiate characters. I finally settled on a simple dozen skills, just enough to keep characters of different backgrounds, well, different. And it keeps the character sheet nice and clean. The skills (and their uses) are:
Academics (knowledge, language, history)
Athletics (climbing, swimming, acrobatics)
Business (professions, economics)
Crafting (tinkering, disabling traps, repairing, alchemy)
Diplomacy (negotiating, persuading, bluffing)
Medicine (healing, anatomy, herbs, psychiatry)
Outdoors (animal-handling, plants, tracking, riding)
Perception (searching, listening, sensing motives)
Performing (acting, playing music, oratory)
Piloting (driving, sailing, flying, navigation)
Politics (bureaucracy, law, making contacts, gathering information)
Stealth (hiding, sneaking, pickpocketing)
Not all of these skills are equally useful to a band of dungeon-delving mercs. But in BECMI, high-level characters usually run strongholds and dominions. In my current campaign, the tinker gnome has a widget factory and the knight became a baroness... and boy, do they wish they'd invested more in the Business and Politics skills now!
The Mechanics: This skill system starts with a simple rule, based on the default assumption already present in the game that a tricky action has a 1-in-6 chance of success. All characters begin at rank 1 in all twelve skills. This is the minimum possible rank; it represents being "untrained."
Characters generally begin the game with 4 (plus or minus Intelligence adjustment) skill choices, to add to their skill ranks as they see fit. At the 1st level of experience, a character may not raise a skill above rank 2, so all of the character's starting skill points must be spent on different skills. After that, the limitation is removed, and characters may build their skills as high as rank 5 (a 5-in-6 chance of passing a skill check, the best possible). One new skill choice is earned at every odd-numbered level (3rd, 5th, 7th, etc.).
Skill rolls are made by rolling your skill rank or lower on 1d6. Ability scores can impact skill rolls, but only minimally. A very high score (16+) grants a -1 bonus to the skill roll. A very low score (5-) imparts a +1 penalty. Regardless, a 1 always passes a skill check, while a 6 always fails.
Demihumans: Using a skill system like this allows some demihuman racial abilities to become more codified and streamlined. Elves, for example, instead of having a special chance to find hidden doors, simply get a +1 bonus to Perception rolls. Dwarf and halfling abilities can be re-interpreted to grant bonuses to Crating and Stealth (rather than special "find stonework traps" and "freeze in wilderness terrain" abilities).
Thieves: In my campaigns, I have been able to use this system to replace the thief class with a more customizable "expert" class. Instead of their traditional d% skills, they just get more skill points than other classes. Experts begin the game with 6 + Int mod skill choices (still limited to rank 2 at 1st level); and a new skill choice is earned at every single experience level. Once the expert passes "name" (9th) level, a new option is available: experts (and only experts) can spend a fifth skill point on a skill in order to achieve "rank 6." Rank 6 (or "mastered") skills are checked on 1d10 rather than 1d6, and only fail on a roll of 10.
A class like this can still feel very roguish if it keeps the thief's traditional backstab ability. In my own campaign, I replaced the backstab with a more generic "critical hit" ability, because I wanted the expert class to be more generic (and a little bit swashbuckley). In any case, a system like this does a handy job of fixing the old "sucky thief skilz" problem, with the thief-type characters building their skill ranks very rapidly.
The thief/rogue/expert type class really shines with skills, but other characters now have an outlet for describing background knowledge and present whims, and for learning things that often simply aren't covered by a character class system. A fighter could learn pickpocketing; a mage could learn to be a medic; a cleric could learn to forge swords. It gives a minor game-mechanical nod to differentiating characters of the same classes, without going way overboard like a proficiency or feat system. And after having used a system more or less just like this for several Classic D&D campaigns to date, I can affirm that it doesn't make the game feel any less "old school" (since that's important to some people). The skills and their uses aren't absolutely defined, leaving plenty of room for DMs who like to make up new applications for old rules on the fly.
What a perfect segue into this topic. And since I don't feel like hijacking the old thread, I get to use ENWorld's shiny new forked thread feature for the first time! This was posted at dragonsfoot.org and rpg.net, but I'm guessing the discussion will be more germane right here where the subject is being debated!
*****Cross-posted:*****
These rules are intended for the 1981 (B/X), 1983 (BECMI), and 1991 (BB/RC) revisions of the original D&D game.
Classic D&D has an official skill system: "general skills," first introduced in the GAZ series and then included in the Rules Cyclopedia. The system was very similar to AD&D's non-weapon proficiencies. All characters (regardless of class) received 4 + Int mod "general skill slots" which could be filled with skills, and then a new slot was earned at every 4th level above the 1st (so 5th, 9th, 13th, etc.).
I don't like non-weapon proficiency style systems for two reasons. First, the lists are big and unwieldy, without even being comprehensive. There's no way to tell whether a skill is important enough to be on the list, whether it should be its own category, or whatever. Nobody can remember all the skills on the list. Some of them are also unbalanced, since they have clear combat uses (like blind-fighting). 2e handled some of this by making the good proficiencies cost multiple slots, but Classic D&D's General Skills system has no such caveats.
Second, there's no mechanical benefit to knowing a skill. If you know the Deception skill and you want to try and fast-talk somebody, make a Charisma check. If you don't know the Deception skill and you want to try and fast-talk somebody, guess how that game action is still handled? Right. Roll a Charisma check. Okay, sure, there's a rule that says you can spend multiple slots on one skill to get a -1 bonus to the d20 roll for each extra slot, but somehow, I think taking the skill in the first place (and being able to learn many skills) should actually mean something.
**********
So what qualities must a good skill system have? Well for one, the list should be short. Like 4e. Say what you will about 4e, but it has a nice, short, easy to deal with skill list. But the list should also take a stab at completeness. 4e's list isn't great, because it's conspicuously missing rules for simple skills like crafting. Granted, crafting doesn't need much of a mechanic in a minis-based wargame, but it's helpful to have that sort of thing in an RPG!
Part of the reason to have skills in the game at all is to codify a player character's background knowledge---the things he learned before becoming a fighter or a mage or whatever. It helps to differentiate Bob Fighter from Joe Fighter. And in a game like Classic D&D, where there are no paladins or rangers or bards (at least, not before name level), having a "religion" skill or a "wilderness" skill or a "music" skill can help to give players that feeling anyway.
Another quality, vital for any skill system in Classic D&D, is that it should be simple and streamlined. Many, many old-school DMs do not believe in skill systems. They feel that character background knowledge and non-class skills ought to be handled by DM fiat in all cases, or that having rules for such things is too limiting. I am not that sort of DM. I like to have some rules to fall back on. Call it a crutch if you must, but I think that character knowledge really ought to be limited by the rules. Something must be in place to keep the fighter's player from deciding, in the middle of the session, just when it's vital to the story, "Oh, yeah, sure I can make a sword. I, uh, I was apprenticed to a blacksmith in my youth. Yeah, that's the ticket..."
Ideally, such a system should be based on a game mechanic already in place. Classic D&D exhibits several: ability checks, rolled on 1d20 (used for the General Skills system); thief skills, rolled on d%; and many actions are handled by rolling low on 1d6 or 1d10 (used for spotting, listening, and various demihuman abilities). This last mechanic is my preferred choice, for reasons I will explain below.
Lastly, skills ought to improve as the characters gain levels. 2e was notorious for the non-weapon proficiencies never getting any better. 3e is kind of the reverse problem, since some of the DCs are so high that you need to be a high-level character (or stack lots of obscure bonuses) to make them. A static skill system is no good, but then again, neither is a skill system with variable difficulty levels (at least, not for Classic D&D).
**********
A skill system can easily be made to work using a "roll lower than your skill number" mechanic, on any die type: d20, d%, whatever. I prefer to use the d6, because it preserves some of that OD&D feel. The game already makes the baseline assumption that when the player wants to do something tricky, the character has a 1-in-6 chance of success... unless the character is a demihuman trying to spot or hear something, in which case the chance is 2-in-6. That's a perfectly wonderful mechanical foundation for a D&D skill system!
It took a while to arrive at a skill list I liked. At first, I used a longer list, in the area of 24-30 skills. It was a little redundant and clunky. I'd also experimented with shorter lists, 6-8 skills. It didn't do enough to differentiate characters. I finally settled on a simple dozen skills, just enough to keep characters of different backgrounds, well, different. And it keeps the character sheet nice and clean. The skills (and their uses) are:
Academics (knowledge, language, history)
Athletics (climbing, swimming, acrobatics)
Business (professions, economics)
Crafting (tinkering, disabling traps, repairing, alchemy)
Diplomacy (negotiating, persuading, bluffing)
Medicine (healing, anatomy, herbs, psychiatry)
Outdoors (animal-handling, plants, tracking, riding)
Perception (searching, listening, sensing motives)
Performing (acting, playing music, oratory)
Piloting (driving, sailing, flying, navigation)
Politics (bureaucracy, law, making contacts, gathering information)
Stealth (hiding, sneaking, pickpocketing)
Not all of these skills are equally useful to a band of dungeon-delving mercs. But in BECMI, high-level characters usually run strongholds and dominions. In my current campaign, the tinker gnome has a widget factory and the knight became a baroness... and boy, do they wish they'd invested more in the Business and Politics skills now!
The Mechanics: This skill system starts with a simple rule, based on the default assumption already present in the game that a tricky action has a 1-in-6 chance of success. All characters begin at rank 1 in all twelve skills. This is the minimum possible rank; it represents being "untrained."
Characters generally begin the game with 4 (plus or minus Intelligence adjustment) skill choices, to add to their skill ranks as they see fit. At the 1st level of experience, a character may not raise a skill above rank 2, so all of the character's starting skill points must be spent on different skills. After that, the limitation is removed, and characters may build their skills as high as rank 5 (a 5-in-6 chance of passing a skill check, the best possible). One new skill choice is earned at every odd-numbered level (3rd, 5th, 7th, etc.).
Skill rolls are made by rolling your skill rank or lower on 1d6. Ability scores can impact skill rolls, but only minimally. A very high score (16+) grants a -1 bonus to the skill roll. A very low score (5-) imparts a +1 penalty. Regardless, a 1 always passes a skill check, while a 6 always fails.
Demihumans: Using a skill system like this allows some demihuman racial abilities to become more codified and streamlined. Elves, for example, instead of having a special chance to find hidden doors, simply get a +1 bonus to Perception rolls. Dwarf and halfling abilities can be re-interpreted to grant bonuses to Crating and Stealth (rather than special "find stonework traps" and "freeze in wilderness terrain" abilities).
Thieves: In my campaigns, I have been able to use this system to replace the thief class with a more customizable "expert" class. Instead of their traditional d% skills, they just get more skill points than other classes. Experts begin the game with 6 + Int mod skill choices (still limited to rank 2 at 1st level); and a new skill choice is earned at every single experience level. Once the expert passes "name" (9th) level, a new option is available: experts (and only experts) can spend a fifth skill point on a skill in order to achieve "rank 6." Rank 6 (or "mastered") skills are checked on 1d10 rather than 1d6, and only fail on a roll of 10.
A class like this can still feel very roguish if it keeps the thief's traditional backstab ability. In my own campaign, I replaced the backstab with a more generic "critical hit" ability, because I wanted the expert class to be more generic (and a little bit swashbuckley). In any case, a system like this does a handy job of fixing the old "sucky thief skilz" problem, with the thief-type characters building their skill ranks very rapidly.
The thief/rogue/expert type class really shines with skills, but other characters now have an outlet for describing background knowledge and present whims, and for learning things that often simply aren't covered by a character class system. A fighter could learn pickpocketing; a mage could learn to be a medic; a cleric could learn to forge swords. It gives a minor game-mechanical nod to differentiating characters of the same classes, without going way overboard like a proficiency or feat system. And after having used a system more or less just like this for several Classic D&D campaigns to date, I can affirm that it doesn't make the game feel any less "old school" (since that's important to some people). The skills and their uses aren't absolutely defined, leaving plenty of room for DMs who like to make up new applications for old rules on the fly.
