Wednesday, June 5, 2013

Social Skills and NPC Stats

On Monday, I introduced my approach to social skills in roleplaying games. Now, I'm pivoting to a related topic, that of non-player character stats, in particular ability scores and skills.

Some gamers don't like to bother with stats for npcs, for a number of reasons: too much work for too little return is a pretty common one - 'why should I spend all that time on a npc who may last only five minutes in the game?!' - though *waves hands furiously* comes up with some frequency as well - 'I just make something up in response to what the players do.'

I can definitely relate to the first: the event which lead to my personal 'old school renaissance' began with trying to prepare a supplement of generic non-player characters for d20 Modern - after looking at page after page of two-line characters stat blocs in a copy of 1001 Characters for Traveller, compared to what I was attempting to churn out for d20M, the less complex rules of earlier roleplaying games looked mighty good to me. In my experience, games which make non-player character stats simple to prep and manage in actual play are more likely to encourage often harried referees to actually use them rather than handwaving them.

I'm not fond of handwaving the rules of the game as a general rule. It's one thing to make a ruling based on a situation at hand, to apply or not apply a particular rule based on its applicability in the moment, or to improvise a target number when one isn't readily available; it's another thing altogether to consistently ignore the rules for an approximation of how an encounter with a non-player character 'should go.' As referee, I like to play the game, too, and if I'm just making stuff up that 'sounds right,' then I'm no longer actually participating in the same game as the players. The rules provide a more-or-less objective foundation to the game-world, and by holding myself to that foundation, it makes it easier for me to assume the role of neutral arbiter when the dice hit the tabletop.

Moreover, most players become invested in their characters' abilities as expressed through the rules: trading attribute points, choosing skills, selecting equipment, and so on. Note that this is not the same thing as optimising a character; all of these choices can be made for reasons, such as fulfilling a particular character concept, that have nothing to do with how effective a particular character is with respect to task or conflict resolution in the game. In my opinion, this investment carries with it the reasonable expectation of that objective game-world against which those characters may be tested.

So, my personal preference is to have appropriate stats for non-player characters, and to this end I select games which skew toward the less complex end of the spectrum when it comes to character generation and creation. Note that a game doesn't have to have identical rules for generating or creating player and non-player characters - consider the difference between, say, a 1e D&D player character and a 0-level non-player character - only that they should be comparable in some way through the rules of the game. To make them easy to use in game, I like to create a stable of generic non-player character stat blocs in advance of play - generic barkeep, generic guard, generic bravo, generic gambler, and so on - often using a stripped down subset of the rules for player character creation, with unique characters generated in more-or-less the same way as player characters.

That means that, in the campaigns I run, each non-player character has a set of attributes and skills that relate to the rules for social interactions; as a function of that objective setting, I'm not inclined to simply make up a target number on the fly if there's some other recourse open to me. Now, there are a number of ways in which the rules for social skills can be used to manage social interactions, and in my experience, at least two of them are pretty problematic. Social interaction rules may be simple or complex. Perhaps the simplest is the reaction or loyalty roll, modified by an attribute score or a skill; the Charisma modifier in 1e AD&D or the Liaison skill in Traveller each come to mind as quintessential examples of this. Years of playing with reaction rolls convinces me that, if you use no other social interaction rules in a roleplaying game, then use these, along with morale checks for opponents - it's pretty near bomb-proof as far as rules go.

Another approach is to roll-against-attribute: frex, a roll under the character's Charm attribute score indicates success. This is one of those rules which can be somewhat problematic, however; in essence, the character's attribute score becomes the objective measure of difficulty in performing a task in the game-world. A rope is as difficult to run across as the character's Acrobatics score makes it; a maid in a countess' entourage is as difficult to seduce as one's Seduction score. Static target numbers, such as the DC's for the d20 Diplomacy score, are similarly flawed, a problem exacerbated in those games by feat and class ability bloat resulting in extremely high skill values.

One way to account for this is using objective measures of difficulty to apply as modifiers to an attribute or skill roll. When I ran Traveller, I used the BITS task system for skill check modifiers; I particularly like the correlation between the descriptor and the value of the modifier, as it helps to establish a common, replicable frame of reference for both the referee and the players.

My preferred approach is to manage social interactions through opposed rolls, stat versus stat, such as Charm versus Wit to bluff a guard. For me, it is perhaps the most intuitive method of handling social interactions through dice rolls, as it pits the relative strengths of the characters against one another.

In the rules and published adventures for Flashing Blades, the examples of skill use provide for both direct rolls against an attribute or skill value as well as opposed rolls. I decided to preserve this approach in my own campaign, but I also wanted to avoid the pitfalls of the simple roll-versus-attribute. Toward this end, I developed an additional non-player character stat - Tractability - which serves a similar function to the BITS task system for Traveller. How Tractability works is the subject of my next post.


  1. I'm a little guilty of the hand waving approach to NPC stats - I have a set of "average" stats and skills in easy reach, then for each NPC, All I really do is make note of how they differ from the norm. It saves me a bunch of time, and has so far worked pretty well for me.

    Just read your last post by the way, and thoroughly enjoyed it.

    1. That's not really what I consider *waves hands furiously,* though - at least you have a set of base stats against which you can compare the player characters' numbers, rather than just looking at the die and deciding what a 12 'feels like' in that 'scene.'

  2. There is also always the every-handy "roll 3d6 when needed" approach.

    1. I've generated several barmaids' Charm attributes on the fly for our carousing Musketeer that way, or by rolling one die 'up' and one die 'down' against the generic barmaid's Charm score.

  3. I really enjoyed this article and the following does a great job of articulating why I find a totally subjective, hand-wavey style of play unsatisfying.

    "The rules provide a more-or-less objective foundation to the game-world...Moreover, most players become invested in their characters' abilities as expressed through the rules...In my opinion, this investment carries with it the reasonable expectation of that objective game-world against which those characters may be tested."

    In some works of fiction, resolution of tests is based on narrative choices – the heroes get beat up by or cannot defeat the villain(s) in the initial meeting as a narrative device to set up a later confrontation where the heroes will win. If done well it is dramatic and leaves the audience with a satisfying, feel good ending. When done poorly (as it quite often is) I am left with the feeling that the heroes have no objective ability. While they were strong, tough, and skilled in the one episode or scene in this episode or scene they are weak and inept. And this variability of ability is deeply unsatisfying for me. I prefer not to see it in my fiction nor in an RPG that I play.

  4. I'm really enjoying this series on social mechanics (one of my favorite topics).

    You wrote:
    My preferred approach is to manage social interactions through opposed rolls, stat versus stat, such as Charm versus Wit to bluff a guard.

    Moi aussi. I've also been fiddling with the Contacts table from Top Secret. As you know, this involves an opposed roll with a range of results that run the gamut from A-I, "A" being "Contact will listen to the agent and try to do anything the agent asks..." and "I" being "Contact will not listen to the agent and attempt to engage in Hand-to-Hand combat."

    You mentioned the TS contact mechanic in passing in your previous post. Have you adapted a swashbucklers' version of this table?

    1. The Contacts table is one of my favorite features of TS, especially with the revised rules in the Top Secret Companion. Adding it to my campaign is on the list right behind adapting the missions chart into a set of rules for rogues' criminal enterprises . . .


Note: Only a member of this blog may post a comment.