False Accuracy in Roleplaying Games
For some GMs, creating NPCs is a pleasure. For others, including myself, it is a chore. Trying it create an NPC on the fly is one of my greatest fears and when I need to figure out the stat of an NPC that I never expected to need it can grind the game to a halt. Most RPGs use a combination of stats to determine a characters skill. Should the shopkeeper’s haggling skill be a +6 or a +5? Now the PCs want to try to shoplift something, what’s his perception? Never mind, the bard is going to charm him, better figure out his mental defense.
One of my favorite things about Monte Cook’s Cypher System is the ease of creating NPCs and monsters. Everything the players roll against is assigned a level which is multiplied by three to obtain a difficulty for the player’s roll. This is great if I’m playing Numenera but what about the game I’m planning to play next, 13th Age? Actually, 13th Age uses a similar system of classifying areas into one of three tiers and basing all difficulty checks on that tier. For example, in a dungeon designated as “Adventurer Tier” all normal tasks are DC15, hard tasks DC20 and extremely hard DC25. This also applies to any required stat for an NPC. What’s the difficulty to haggle with an average denizen of an adventure tier environment? DC15 of course. How about the Mental Defense for the charm spell? 15.
DND Dice by Zamanu. Image used under CC 3.0 Attribution No-Derivative license.
At first this rule can appear to make the game bland and uninteresting. Everyone the characters meet in an environment is exactly the same, what’s to differentiate one NPC or challenge from another? The reality is that when making a single roll against an NPC, or to disarm a trap, there is not much difference between a 15 and a 16. The player will roll and they will fail or succeed without any real understanding of whether the chance of success of the check was 75% or 80%. This makes determining whether a DC is 15 or 16 a false accuracy.
In order to add some interest and keep the players guessing throw in the occasional difficult check. Perhaps that NPC is really good at haggling or the trap in the dungeon is particularily well-constructed. This is enough of a difference to make the world seem real.
Are you one of those crazy GMs who likes to fully stat out NPCs using character creation rules? Let me know in the comments or find me on Twitter or Google+. Don’t forget to subscribe for more.
I must admit, I do love stating out NPCs. For more elaborate games, that is. I often find it easier, though, to have preset stats for groups of NPCs rather than single ones – especially when playing a game where NPCs are incidental (hack and slash) rather than planned (mystery). Amateur-level merchants will have Haggle 3, while more seasoned merchants will have 5, and merchant kings 8, etc. I think it’s more or less a variant of the type you described above, and tackles the “bland and uninteresting” issue you mentioned as well.
I usually make their stats up on the fly: working out something they’re great at (because its their job for example), something they’re good at (because it relates to their job), and stuff they’re average (or bad) at (because they don’t relate to their job). I do occasionally do what you’ve suggested and give them an odd quirk where they have an odd skill quite high, but normally I save that for significant NPCs they will likely see multiple times and not the shopkeeper they’ll meet once
To put it in an example: a politician. His main “job” is to persuade people to agree with him and work out who he should agree with. Thus he’s great at persuasion, resisting persuasion, bluffing, reading people etc. Secondary to that he needs to have some knowledge of law (or at least the processes of passing laws) so he’s good at tasks relating to understanding laws, or passing laws. If he’s from a militant country he might have some gun related skills in the good range. Average and bad is everything else, with bad being areas I want him to have a weakness. Perhaps he really doesn’t get science, and thus knowledge rolls relating to nature, scientific processes etc are in the bad range.