NPC Traits

JohnnFour

Game Master
Staff member
CL Add-on Dev
Demonplague Author
Wizard of Adventure
#1
I've collected a few resources that have lists of NPC traits, including an RPT blog post with a whole bunch of them.

I now want to collate all into a spreadsheet and categorize them (eg physical, mental, behavioural, etc.).

And then I want to add a brief blurb (say, 10 words or so) that quickly describes the traits.

Ideally, perfect world, there's another column with a system agnostic game effect of the trait.

Unfortunately, I don't have time to do more than collate the traits, delete duplicates, and put them into a spreadsheet.

@ELF Do you think this is a good candidate for crowdsourcing help from RPT readers? Or maybe there's a different approach to getting the traits organized and fleshed out a bit?

My end goal is to turn this spreadsheet into a generator.
 

ELF

Generator Sage
Staff member
CL Add-on Dev
#2
I've collected a few resources that have lists of NPC traits, including an RPT blog post with a whole bunch of them.
I have been collecting a spreadsheet like that as well (my Twitter bot utilizes some of that, although not very efficiently). I have found that for complex generators, a manual processing stage on a spreadsheet can indeed be quite helpful. With YAML support that has become even easier.

I now want to collate all into a spreadsheet and categorize them (eg physical, mental, behavioural, etc.).
I would categorize them in a more fine-grained manner, so that you won't generate NPCs that would be, for example, lazy and energetic, curious and apathetic, etc. I have tried to find the best approach for that, and using the modern Big Five personality model would seem to produce better results than, say, Myers-Briggs, not to mention pseudoscience like enneagrams or horoscopes - although they can have their uses, too.

And then I want to add a brief blurb (say, 10 words or so) that quickly describes the traits.
I think that the most usable format would be 'the end of a phrase', i.e. not including the subject or the predicate. That can then be combined flexibly with different sentence structures, tenses and persons as required. As in the table B below (pardon the lacking table formatting):

A)
  1. "The innkeeper is"
  2. "I am"
  3. "You have been"
B)
  1. "forgetting names and details"
  2. "offended if not addressed properly"
  3. "friendly to everyone in the same trade, and rude towards others"
Ideally, perfect world, there's another column with a system agnostic game effect of the trait.
That could be nice, but many traits work best just as characterization tools, and forcing them into a game mechanic could be quite artificial. That could work better on the category level (i.e. agreeable traits would give bonus to something, and neurotic traits would mean a penalty to something else). This would also negate the need to invent mechanics for hundreds of individual personality traits.

Unfortunately, I don't have time to do more than collate the traits, delete duplicates, and put them into a spreadsheet.
Collating the data is the big chunk of work, but a good text editor software make the rest of the tasks relatively easy. And if you need a hand, let me know.

@ELF Do you think this is a good candidate for crowdsourcing help from RPT readers? Or maybe there's a different approach to getting the traits organized and fleshed out a bit?
Crowdsourcing could be useful, but would probably require quite a bit of editing. There is already so much collected material available that I would maybe start with that, see how well that works, and then crowdsource areas that would need more content. But diminishing returns are something to keep in mind here - a generator with 2000 entries is not necessarily twice as good as one with 1000 entries.
 

JohnnFour

Game Master
Staff member
CL Add-on Dev
Demonplague Author
Wizard of Adventure
#3
I have been collecting a spreadsheet like that as well (my Twitter bot utilizes some of that, although not very efficiently). I have found that for complex generators, a manual processing stage on a spreadsheet can indeed be quite helpful. With YAML support that has become even easier.
Cool! Should I maybe feed stuff into your existing system? Avoid duplication of effort and build your data pools up.

What's the best way for me to see your data and compare with mine so I could expand it?


That could be nice, but many traits work best just as characterization tools, and forcing them into a game mechanic could be quite artificial. That could work better on the category level (i.e. agreeable traits would give bonus to something, and neurotic traits would mean a penalty to something else). This would also negate the need to invent mechanics for hundreds of individual personality traits.
Great idea. Maybe I group traits into categories, and then offer several mechanics ideas for each category. GMs could apply or not as they like.
 

JohnnFour

Game Master
Staff member
CL Add-on Dev
Demonplague Author
Wizard of Adventure
#4
Poking around in the GitHub repo now for possible tables to add data to.
 

ELF

Generator Sage
Staff member
CL Add-on Dev
#5
The personality table has a skeleton framework for this, but it was apparent that a one-dimensional listing was not a good way to collect this kind of data. A two-dimensional spreadsheet works better when still figuring things out.

Cool! Should I maybe feed stuff into your existing system? Avoid duplication of effort and build your data pools up.
That would be great, and make the thing less of chore for both. I'll send you a Google Sheet link.

Great idea. Maybe I group traits into categories, and then offer several mechanics ideas for each category. GMs could apply or not as they like.
That sounds like a neat solution. Personalities are multifaceted and people are likely to have different interpretations of these traits, so it will be good to have options and multiple suggestions on how to handle nuances like this.

Currently I have set up the Big Five axis each with a scale of 5 slots for variations between minimum and maximum expression of that axis (producing columns like minimal extroversion, medium stability and maximum imagination). So far it seems that these axis indeed are independent and variations on different axis combine well, so that the end results are believable personalities. This also makes it possible to create different description patterns so that the components do not clash with each other (i.e. if you choose multiple traits, pick them from the same position on the relevant axis). But this is still on the testing stage, so let's see if something needs to be tweaked.
 
Top