Prestige Classes, Demographics, Clerics and Statistics.

[Silveras]

The DMG demographic info isn't well-suited to making judgements about anything other than individual communities. I did some work on this a few months back, and if I recall, PC-Class percentages ranged from roughly 6% of smaller communities to around 2% for cities.

Actually, it is not representative of much other than 'how big a town do we need to get to if we want to spend X gp or have spell Y cast ?'.

The demographics in the DMG include only "normal residents" (a term that remains rather loose) -- the DM is free to plunk down a whole monastery of monks, with any distribution of levels s/he sees fit, in addition to what the tables say.

Also, since Prestige Classes are an optional part of the game, they are not included in the tables. If they were, I doubt that you would see them reducing the numbers of other classes; I expect they would be in addition to the "normal" Clerics, Druids, etc.

1st and 2nd edition used a base assumption of 1% of the population was PC-Classed. Personally, I'm using 2% as the number for my current campaign. I could certainly see going a little higher but I think 5% is about the maximum I would see. It would have been nice to be given something a little more official to work with, but there it is.

Well, 1% of the number was classed AT ALL, the remainder being "0-level" non-combatants and generic soldiers. That model does not work as well in a game where the residents are ALL some sort of class with multiple HD. PC Classes remain the minority, the specialists and outstanding talented individuals of their times and communities.

Honestly, I think it is an exercise in frustration and futility to try to make a sensible demographic rule out of D&D. D&D is a game; it has needs that are as often dictated by "play balance". Demographics joins Economics and Physics as analyses that it is often more trouble than it is worth to try to use in the game.

That is not to say don't tinker ... I do, myself. I am just saying "don't agonize over it".

FWIW, I have revised the tables for communities to reflect cultural and racial trends in my homebrew world. Each area gets a different set of dice rolls; some classes are eliminated entirely, and others get bigger dice rolls than the DMG tables. I also use the alternate demographics in A Magical Medieval Society: Western Europe (2x as many at level -2, instead at 1/2 the level).

 

[Dogbrain]

Unskilled Peasanty (Extremely common)
Skilled worker (Very common, Experts etc)
Nobility (uncommon)
PCs-classed characters as a whole (very uncommon; only half of one percent)
PCs of a specific class (extermely uncommon)
PCs of a specific prestige class (very rare)
PCs of a specific prestige class and who are of a certain level (very very rare)

You, sir, probably do not exist, and I can mathematically prove that. In addition, I can mathematically prove that I also probably do not exist, and in all probability, we cannot communicate even if we did exist, as it would be probably impossible that we would ever come across each other in any way, even electronically.

I cannot have ever met my wife, since the odds of me meeting her, specifically, are less than one in a billion.

Get a set of percentile dice and make 20 rolls, recording each. Look at that set of numbers. It cannot have happened, since the odds of rolling that specific set of numbers in that specific order are 10^20:1 against. What you have done is so unlikely that no reasonable person would say that it could happen.

Have I made my point sufficiently?