NPC random statistics roll chart?

[Ragitsu]

Originally Posted by Fish

Hmm. Tricky. Europe is a pretty broad category, and its population is dominated by the really big players, leaving the small ones out in the cold. Europe, even more than the United States, varies greatly from region to region.

What did you have in mind, specifically? A chart for "Random Person Who Comes from Europe" won't look quite the same as "Random Chart for Person on the Street in London."

Not necessarily comes "from" Europe, but for those that are already in Europe.

[tantric]

Wasn't there a fairly systemless NPC personality trait table in the AD&D DMG? I've been looking for something like that...

[Fish]

I have made a few modifications to the randomizer. In particular, I have added randomizer rolls for Age and Gender, and fleshed out some more (but not all) of the Occupational sub-rolls. Trying to find a suitable statistical model for Local and State government is giving me some trouble.

For the sake of brevity I will simply include the new stuff.

AMERICA, Year 2000 (TL8)

HOW TO USE THESE TABLES

DICE.  You will need five six-sided dice.

WHAT TO ROLL.  Some of the tables call for three dice to be rolled and their
values totaled.  This produces a standard distribution, also known as a bell 
curve, in which the center values are most probable and the end values least 
probable.

Other tables will refer to a perhexile roll.  This is analogous to the use of two 
10-sided dice, or percentile dice.  Since we are using only 6-sided cubes, 
this gives us a result in a Base-6 number system.  For this type of roll, do 
not add up the total.  Instead, use the dice result in sequence.  If you rolled 
a five-dice perhexile roll, you would get results such as 11112 or 45264 or 
66352.  The sequence matters!  11112 is not the same as 21111 or 12111.

For best results on a perhexile roll, determine in advance how you will read 
the dice:  for example, left-to-right; nearest-to-farthest; or a single die 
rolled multiple times in sequence.  For obvious reasons, reading the dice in a 
lowest-to-highest pattern is not suitable!

WHEN TO ROLL.  The tables are not a model to be followed slavishly.  If you 
would rather decide that the character is a left-handed female Lebanese 
social worker, simply choose those variations and proceed.  Nobody will 
catch you.  Honest.

If you decide to skip one of the tables, it may be best to select a result from 
the table before moving on.  Each random table builds on what has already 
been chosen — for example, before rolling to determine the character's ST 
score, it is helpful to know the character's Age, Gender, Height, and Weight 
so the result may be modified realistically.

Optional tables may always be skipped without bothering to make a 
selection, hence the term “optional.”

HOW ACCURATE ARE THESE FIGURES?  The author does not recommend 
using these figures as an authoritative citation in a research paper, nor does 
the author suggest these figures are rigorous enough to be used as a basis 
for local taxation or exhaustive economic study.  The numbers of each group 
may have been tweaked or adjusted by a few percentile points one way or 
another, in order to fit the probability curve.

1:  ANCESTRY
(already posted)
			
2:  AGE

AGE.  Use a perhexile roll on four dice to randomly determine the character's 
age.
	1111-1334	6.79%	0 to 4 years of age
	1335-1613	7.33%	5 to 9 years of age
	1614-2252	7.33%	10 to 14 years of age
	2253-2525	7.17%	15 to 19 years of age
	2526-3152	6.71%	20 to 24 years of age
	3153-3421	6.86%	25 to 29 years of age
	3422-3655	7.25%	30 to 34 years of age
	3656-4352	8.10%	35 to 39 years of age
	4353-4643	7.94%	40 to 44 years of age
	4644-5315	7.09%	45 to 49 years of age
	5316-5532	6.25%	50 to 54 years of age
	5533-6114	4.78%	55 to 59 years of age
	6115-6236	3.85%	60 to 64 years of age
	6241-6352	3.39%	65 to 69 years of age
	6353-6461	3.16%	70 to 74 years of age
	6462-6555	2.62%	75 to 79 years of age
	6556-6634	1.77%	80 to 84 years of age
	6635-6666	1.54%	85 and up


3:  GENDER

GENDER.  Based on the age of the character, roll on the appropriate chart for 
gender.

	AGE 0 TO 29.  Make a perhexile roll with three dice.
	111-413		51.4%	male
	414-666		48.6%	female

	AGE 30 TO 44.  Roll a single die.
	1-3		50.0%	male
	4-6		50.0%	female	

	AGE 45 TO 59.  Make a perhexile roll with three dice.
	111-364		49.0%	male
	365-666		51.0%	female

	AGE 60 TO 74.  Make a perhexile roll with three dice.
	111-353		45.8%	male
	354-666		54.2%	female

	AGE 75 AND UP.  Make a perhexile roll with three dice.
	111-321		36.6%	male
	322-666		63.4%	female	

SEXUALITY.  This optional chart is for determining the character's sexual 
orientation.  Since reliable statistics are difficult to obtain, the author has 
inserted his best guess, based on several contradictory studies, and you're
stuck with the result.

	MALES.  Make a perhexile roll with five dice.
	11111-11112	0.02%	male-to-female transgender
	11113-12113	2.79%	bisexual
	12114-13222	3.29%	gay
	13223-13432	1.00%	asexual
	13433-66666	92.88%	straight

	FEMALES.  Make a perhexile roll with five dice.
	11111		0.01%	female-to-male transgender
	11112-12322	3.79%	bisexual
	12323-13221	2.28%	lesbian
	13222-13423	0.95%	asexual
	13424-66666	92.94%	straight

EDUCATION.  
(already posted)

WEALTH.
(already posted)

OCCUPATION.  
(new tables found below)
		PROFESSIONAL
		3-5	4.7%	accounting and bookkeeping
		6	4.6%	management and technical consulting
		7	6.9%	legal services
		8	9.7%	architectural and engineering services
		9	11.6%	management of companies and enterprises
		10	12.5%	temporary help services
		11-12	24.1%	employment services
		13	9.7%	computer systems design
		14-15	11.5%	services to buildings and dwellings
		16	2.8%	business support services
		17-18	1.9%	waste management

		EDUCATION AND HEALTH
		3-4	1.9%	outpatient care center
		5-6	7.4%	social assistance
		7-9	28.2%	ambulatory health care		
		10	12.5%	educational services
		11	12.5%	offices of physicians
		12-13	21.3%	nursing and residential care
		14	6.9%	nursing care facilities
		15	4.6%	home health care
		16-18	4.7%	child day care

		LEISURE AND HOSPITALITY
		3-11	62.5%	restaurants, food and drink sales
		12	11.6%	amusements, gambling and recreation
		13-14	16.6%	accommodations and hotels
		15	6.9%	other
		16	4.6%	performing arts and spectator sports
		17-18	1.9%	museums, historical sites, zoos and parks

		FINANCIAL
		3-4	1.9%	funds, trusts, and financial vehicles
		5-6	7.4%	rental and leasing services
		7-8	16.7%	real estate
		9	11.6%	securities, commodity contracts, investments
		10-11	25%	insurance carriers		
		12-13	21.3%	depository credit mediation
		14-18	16.2%	commercial banking

		TRANSPORTATION
		3	0.5%	pipeline transportation
		4	1.4%	water transportation
		5-6	7.4%	transit and ground passenger transportation		
		7-8	16.6%	warehousing and storage
		9	11.6%	air transportation
		10	12.5%	couriers and messengers
		11	12.5%	other transportation support activities
		12-15	32.8%	truck transportation
		16-17	4.2%	rail transportation
		18	0.5%	scenic and sightseeing transportation

[Ragitsu]

How about disadvantages like Intolerance, Odious Personal Habit and the like? I am sure we can fit those in somewhere :-).

[Fish]

That's an interesting idea. The difficulty is that so many habits overlap; a person can easily have more than one, or even none (that is, none that reach the level of a point-value Disadvantage).

The more random fields one adds, the more one risks obtaining a result that makes no sense: an aerobics instructor with Callous, Paranoid, Distinguishing Feature (tattoo), One Leg, and Deaf.

I'll stick to the model I've got working for the moment; the nice thing about these characteristics is that they don't overlap much. You can't be both male and female, both 15 and 40.

[Dwarf99]

yeah and when you're done perhaps a word document or a PDF to circumvent those annoying code boxes

[Fish]

I've gathered some statistical data on prevalence of some interesting GURPS traits. I'll add them to the list soon. Among those traits:

Albinism, dwarfism, gigantism, epilepsy, autism, Downs syndrome, obsessive-compulsive disorder, lactose intolerance, lazy eye.

Allergies: bee stings, pollen, peanuts, penicillin, topical skin allergies (eczema), athsma.

Fraternal twins, identical twins, identical triplets, quadruplets, quintuplets, and conjoined twins.

Still to come: obesity, hair color, eye color, handedness.

What else can you think of that might be fun?

[Ragitsu]

The luck/unluck based Advantages/Disadvantages could be fun.

[Fish]

Originally Posted by Ragitsu

The luck/unluck based Advantages/Disadvantages could be fun.

What, in your opinion, is the statistical distribution for luck and unluck? I could design the distribution any way you want, but how do we agree what is realistic? 5% of people are lucky, or 0.5%?

I'm not trying to make a complete NPC generator, with all advantages, disadvantages, quirks, perks, skills, names, income, appearance, date of birth, blood type and so on included in the model. All I'm trying to do is find a way to model with dice a given population profile: how many of Group X are A, B and C?

Take Europe, for instance: if you choose randomly from the set "people in Europe" (830 million people) you're very likely to choose people from Russia (140 million people, about 75% of whom live in European Russia). After that, Germany (82 million), Turkey (75 million), and France (65 million).

But "people from Russia" isn't the same as "people born in Russia," which isn't the same as "Russian-speaking people born in Russia to Russian parents." A certain percentage of people who live in Russia will be German, Polish, Greek, American, etc.

That can be modeled with ease. The demographics exist, and we agree what "reality" is.

Take autism: some studies say that 1 in 150 people in the U.S. suffer from autism. I would venture to say, however, that this number is high, if we're speaking of "autism so severe that it's worth points in GURPS."

I have no idea how to model things like Intolerance, Luck, or OPH, and whether they overlap with other conditions. Are famous Hollywood actors and athletes more likely to have Luck? Are white people more likely to be Intolerant, or less? It's hard to find the right mathematical model.

[Dwarf99]

Originally Posted by Fish

What, in your opinion, is the statistical distribution for luck and unluck? I could design the distribution any way you want, but how do we agree what is realistic? 5% of people are lucky, or 0.5%?

well I'm not the guy that you're really asking about this but you might say:

5% of people have Luck 1
1 % of people have Luck 2 or Serendipity 1 (only because serendipity gives limited control over a situation, in that the character can make a plausible story for what happens)
.1% of people have Luck 3 or Serendipity 2 or higher

keeps the numbers small and doesn't require research