Topic: Levels of NPC's throughout the population of your game world.

[Cory Magel]

Ditto with yamma and GOF.  I don't feel as if I need something telling me what the odds should be.  I should be able to say "Ok, well this guy travels a good amount and has been around the block a few times, so I'll say you need to hit a target of 80 to succeed." or "Ok, well this is just some backwoods yokel who digs ditches for a living, so you need a 50 to succeed." then the player will have to rely on their pick-pocket skill to get there.

[RandalThor]

I have done that, as well, and sometimes that is a fun way to go, but sometimes it is nice to have some stuff ready to go, too. Plus, some of that can still be done even if you have some of the numbers done ahead of time.

[jdale]

Just arbitrarily setting a difficulty for picking pockets is ok, if you are comfortable doing it that way, and you have earned the trust of your players. But the difference is that you are making a judgment call that is directly "how hard should this player's action be" whereas if you have an idea of the level distribution you have a more objective way to make that decision. In the latter case you can say, e.g. level is 1d10+3, and therefore alertness skill is X. In the former case you might be thinking how alert is the miner and how often does he deal with pickpockets, but you might also be thinking about how often you want the player to waste your time with pickpocketing rolls and is this your buddy and did he bring the snacks this time etc. All ways in which the decision can become subjective and therefore in which you can end up favoring some players over others, or some actions over others. You can to some extent address that by rolling but even then you are effectively defining a range subjectively and on the fly.

Again, this is not a problem for all GMs. But I think having a level distribution takes away those concerns, and is very useful to many types of GMs.

[Cory Magel]

I see figuring out the difficulty of the action as a better gauge than a level.  As I hinted at in my example above, an NPC might be 25th level and oblivious or 10th level and have developed their alertness as much as they could have.  Their level is of less importance than how much I think that particular NPC would have developed the skill in question... which hinges on life experiences.  So the farmer might be clueless while the street vender is quite attentive.

[GrumpyOldFart]

That's basically what I'm saying, but I extend it to the scenario and its setting.

Take the example of the mining camp. When it's as generic as "a mining camp," my instinct is to say the largest group will be in the 5-10 level range, the next largest group will be the low level apprentice types, and there will be a scattering of seriously experienced people.... but most of them will know nothing at all about anything but mining. There will be a certain amount of "service sector," keeping them fed, bringing in supplies, assaying ore, etc. They will be good at what they do, know something (maybe even a lot) about mining, and little else. Fine and good, until you put it in the setting.

When it's in the coal mines in the center of the country, they know mining, know hardly anything about combat... but boy, they sure know about fighting fires, dealing with explosions, and first aid for burns. When it's in the gold mines out on the frontier, where they're subject to raids at any time, yeah, they're middling good at combat, and probably higher level on average as well. But at the gem mine right at the edge of a major node's magical interference pattern, the actual mining is almost a sideline, the business of the day is surviving all the monsters long enough to get a chance to dig.

Whatever that particular camp has as a "service sector" for those miners changes in similar ways.

That's what I mean by having the setting and the scenario drive those decisions rather than a broad generalization.