Post/Author/DateTime	Post
darkwave 09-16-04, 02:16 AM	Had a situation come up where I had to generate the alignments of NPCs - ALL the NPCs of a city. At first I came up with an ad hoc system, but have tried to refine it. Here is my presentation of those efforts. Needless to say, this is EASY on a spreadsheet, plus I can make it spit out easy answers, but it takes a bit to explain. :) The NPC Profile of the "Average" Neutral-Neutral Government of an "unvectored" race. Okay, let's first get some terms out of the way. 1) An "unvectored" race is one that doesn't have some proclivity towards a given alignment. While some D&D books call humans a "good" race, due to their pantheon's general balance, it would seem that at least in many of the "worlds" humans would be best considered a Neutral race. 2) Neutral-Neutral Government. Some may argue that a government can't be neutral in terms of Law/Chaos. When used this way, -> I'm <- having it mean that the government has "relaxed" (had time to become stable), has no good/evil leaning, and has a balance of governmental order and people's freedom. 3) "Average". When termed to a government is the government doesn't give preferences to a given alignment. To clarify a bit: this doesn't mean that they let criminals go, it means that a good character that breaks the law will be in trouble, and an evil character that DOESN'T break the law is not (or even vice versa). Another way to look at it is action counts more than intent. As the Government in this case is Neutral-Neutral, PR is "free", and there is no attempt to change society. 4) NPC Profile. There is assumed to be some variation in personal beliefs in a population. The profile is a fraction of the whole populace that shares a particular "alignment" of the 9 standard codes: LG, NG, CG, LN, NN (N), CN, LE, NE, and CE. The BIG underlining Assumption: It takes personal energy to "choose" an ethical standard. On the other hand, in a "relaxed" society, there has been a stability generated somehow that may allow some to explore moral choices. One can chose a particular standard (LG and True Neutral for examples) one can chose a particular level on a given "axis" (choosing to be good, or one must balance law and freedom), one can choose to NOT be some axis (Don't want to be evil for example). Or one can sort of passively drift towards a standpoint. The more "picky" a NPC is, the more personal "energy" they are using. Using all this, I got for a profile (I spaced them funny as I have no idea how to make charts here) of 24 "averaged" NPCs: 2 3 2 3 4 3 2 3 2 Place the numbers in the standard G-N-E, L-N-E grid. So in this, there is a greater chance of finding a neutral-neutral character than any other "ethos", and the "extreme" ones (no neutral attributes) are the most rare. This example has no "PR". So far, this isn't probably too useful, as the government outlook, government intent and the people are all neutral. Not too often is this seen (at least in my campaigns). So how does one change the ratios. First, let’s look at changing the government outlook. For this look at what types of folks make up the government. However here comes an important "value judgment". Does governmental survival place a greater value on the Law-Chaos axis, the Good-Evil axis or are they equal. I normally pick that for a people without inclinations, a government would find that playing around with the "good-evil" axis is twice as "expensive" than the law-order, mainly as the government, no mater how ridged or loose it is, is DEFINED to be on that axis, and as such playing around with the other is an indirect process (theocracies are one of the government types that would be on both axis). In any case, I then divide the axis into 5 parts: the normal Good - Neutral - Evil axis would become a Good - Good/Neutral - Neutral - Neutral/Evil - Evil for mortal governments. Now for the next step. Next, literally do a "weighted mean" of the government. I'd use true neutral as the origin (0,0), and then simply call the extremes of ethos as +/- 1. Then I'd take a guess of the "influence" of different NPCs in the government. Unless the PCs are going to interact with people in the government, or the balance of power in the government is fairly clear, I'd generally shorten this step to whatever the "general atmosphere" of the government's "behind the scenes" makes sense. For the most part governments (just like the people are assumed to be) don't hit extremes. So while a Paladin may rule a city, unless the Paladin is scouring the governmental ministers for deviations or it is due to the city being a theocracy, the government will probably be Neutral/Good - Neutral/Lawful or "less", meaning even more biased towards neutral-neutral. When there are multiple governments for a location, it is somewhat even easier. Simply take the Ethos of the various governments, estimate their "influence" and average them. The different governments have the tendency of collecting people of "their" ethos when they differ, and are better at finding the "faults" of the other governments when they agree (an example would be it is easier for the LG government finding those that aren't LG in another branch of government claiming such than finding those that deviate in their own branch). Okay, then take the average and find the resulting "governmental ethos". Remember this, plus how one weighted the 2 axis, as this will be useful near the end. The next step is find what the government wants to Project. This DOESN'T have to match what the government is composed of, in fact it often seems historically that they DON'T match. Often this is simply a case of PR. "We're a Good, Fair nation" would indicate a preference for NG, while "We're an orderly society" would just be a preference for law. PR takes effort for a government (and more the more deviated it is from the truth of the government's make-up), and it can't overwrite the underlining reality, but it too has an impact on the people's profile. This is also what the PCs have heard about the place, unless the PR, racial proclivities and governmental make-up noticeably mismatch. In that case, they STILL have probably heard the PR, but have heard "rumors" as well. Different governments for a location may have different PR efforts. Resolve the same way before, but this time take the mid-point of the INTENDED effort (for an example, a group saying "there is no evil here" would be pushing the midpoint of neutral and good, with the place of order-chaos matching that sub-group's ethos, but a group spending the effort saying: "We are a good, law abiding society" has spent the effort to exactly place the result, regardless of their actual ethos type. Note: PR stating a singular government's own ethos only costs effort if the governmental average differs. If the PR in turn ONLY applies to the one government, it shows the fracture of the governments, but it also "free". Another reason that all this can be important is it can give a guild line for how much "feelings of cutting through layers of lies" the PCs get if they start investigating the government(s). Even better, you'll have used it for the NPCs, used it to generate the mood of the town, and can use it for quick reference later as needed when PCs deal with the Government(s). Once this is done, and the nearest "ethos" of the 25 mortal choices for governments has been decided, you're ready for the NEXT step. Vectored Races. Pretty much if the DM has made decisions on certain races having a "natural proclivity" for certain ethical choices that differ from the MMs, this is where to put it in (example: is the race a certain way due to nature or nurture). Also the DM has to determine if the tendency for those races is "firm" - as in the members of that race all share that ethos, or situational flexible. Races that are determined to be "firm" are VERY heavily weighted when in this phase, as it changes the basic ethos chart (as shone in the beginning). One has to build up ones chart at this stage, but it comes in 2 parts, one being those that are "firm" (who will not change) and one with those who are flexible. Those who are "firm" are still part of calculating the placement of the overall racial proclivity, but are removed before the various upcoming transforms are done to the populace. Here is an example: The DM has decided that race X is inherently LE "firm" and the location (a city) is half race x and half Unbiased Neutral. Relative to the "Racial proclivity" point, one first generates the CHART. (I hope this looks okay) 2 3 2 3 4 3 2 3 2 + 0 0 0 0 0 0 24 0 0 = 2 3 2 3 4 3 26 3 2 In this example, We get a clear N/E by N/L result (24 of 48 by -24 by 48). However since the one race is "firm", they get removed before the transformations. Generally it is unnecessary to make a race "firm" to get the societies described in the game universe, a general bias coupled with government type/PR generally distorts a populace sufficiently. For DMs, my advice to use this tool (and generate charts) is to say: "If I made them look human, and erased many of their ethical memories, what do I feel they would "pan out" on an ethical chart?" I have already shown the "neutral" chart. Below I present a one axis and 2 axis "soft" biases towards "extremes". Rotate as needed. 1 axis (bases used: 24) 1 2 1 2 4 2 3 6 3 2 axis (bases used: 48) 1 2 4 2 7 8 4 8 12 Okay, now one has 3 different "positions" on a 5x5 matrix. Plus a "baseline" for population transforms. How to use them? First, recall if one axis is more "valuable" or "costly" than the other for the governments. If they are, I suggest that shifts due to the government type go through the below processes twice for that direction in the “governmental outlook” section. The "governmental" processes: For the Governmental "PR" (1st step). Remove 1 forth of the smallest "value" in the modified ethos chart from EACH value in the chart. It is spread evenly into any "boxes" that the PR value touches (either 1, 2 or 4 boxes). For example, a Neutral government has non-biased neutral population, but they put on a mild PR campaign to look both lawful and good. 2 3 2 3 4 3 2 3 2 Becomes: 1.5 2.5 1.5 2.5 3.5 2.5 1.5 2.5 1.5 with 4.5 in the "kitty". The campaign was mild in claims but very pervasive (as the population has no major issues), so the result was on the border of neutral/good and Lawful/good - i.e. a 4 way split. Adding the results back in (and multiplying by 8 to make things not TOO ugly looking). 21 29 12 29 37 20 12 20 12 A fairly effective campaign, as the population is fairly malleable in this case (fairly even in values). For the "governmental outlook" take the initial ethos chart you generated for your population (don't forget to remove those that are of a "firm" ethical standpoint). Then break down placement of a government into "steps". Each "step" from true neutral will require an "iteration" (or 2 if that axis is chosen to be more "costly"). For each "step" a government is off of neutral, increase each value of the "preferred" side by one third. The "middle" is decreased by one sixth, and the "opposite" is split in a third (this will often require renormalization later). For an example, a Lawful, Good/Neutral government (where the Good/Evil axis is twice as "costly" would have to run through the process twice for the Lawful direction (as they have 2 steps in that direction) and twice for the "good" direction (as it is twice as costly). The last part is due to Racial Vectoring. Use the same steps as the “governmental outlook”, but do not bias the axis differently (assuming one had before). The next to last step is converting the values in the 9 boxes into percentages. Simply sum all the 9 values, and divide each individual box by that resulting sum, and multiply by the percentage of races that WEREN’T “firm”. Then add the percentage of the “firm” races into the appropriate boxes. That will give you the percentages (by alignment) for the alignment of that location. I’d suggest rounding to the nearest percent, and then converting it into a chart that can be rolled against with percentile dice. There you have it, a way to generate the alignment of NPCs on all your locations, for however many governments, different population types and alignment vectoring tendencies. The only step that required a bit of difficulty programming was the PR segment. On the one hand, it is very “digital” in nature. On the other, it can look a bit like “reality following hype”. However if the government or population is noticeably biased against the PR, it pretty much is washed under, or if the population is already biased from some other force, the effect is minimized. “Mild” races can therefore be greatly influenced by the PR. If one wants to have a “PR resistant” race (though I would think the PR people would be stubborn as well) that somehow DIDN’T have some other form of bias, I’d suggest either 1/9 (somewhat stubborn) or 1/36 (almost resistant to outsiders input for alignment) of the smallest value on the chart at that step. Smaller than that the race doesn’t follow some of the other basic assumptions and simple elegant ratios (half, third, sixth = whole). Hope I didn't miss anything, system went down once while typing this up. Please, I know this is probably long winded, but any comments, suggestions or questions?
darkwave 09-16-04, 02:22 AM	Oops, don't want to edit it just in case the system goes down again, but I also wanted to say that issues, or halfhearted attempts at PR will ALSO result in the lower fractions (not just "PR resistance").
s4dfish 09-17-04, 02:47 PM	That was a fun read :) (statistics junkie). I'll have to sit down and play with it to give you any more input, but first impressions are positive!