Input ability scores manually or via your rolling method of choice; get a character description. Functions most naturally for first-level PC types in Generic Fantasyland, but can function for other purposes as needed. May get less repetitive as I fill it with more Content.
It shows things like "Moral foundation scores: Care/harm 0.79, Fairness 0.94, Authority 0.39, Loyalty 0.70, Purity 0.48, Freedom 0.61,
ReplyDeleteHEXACO scores: Honesty/humility: 0.71, Emotionality 0.01, Extraversion 0.53, Agreeableness 0.69, Conscientiousness 0.60, Openness 0.57.[...]" and a long list of what looks like variables ("starting with bard
checking index 0 which should be bard
checking index 1 which should be gunslinger").
Is it supposed to work this way?
No, you've caught it with its pants down - that's backend, some of which is legacy code from a thing I lazily adapted it from and some of which is doing actually useful work. Do you mind my asking what browser you're using and what buttons you're pressing?
DeleteI am using Firefox (with at least two different adblockers installed - maybe this is a problem?) and I got that result from 'Describe me' button as well as all of the others buttons such as "Set all to 10".
DeleteIt works fine from Chrome and Opera
Huh, I'm using Firefox with adblockers by default too! However, I've wrangled about in the code and the part that prints that should be gone. Let me know if it's still there, or if other problems arise.
DeleteIt is still doing the backend in Firebox but aside of two adblockers the browser also has a bunch of other restrictions on scripts and such.
DeleteI ripped through your source code and I can see how the different traits are in fact somewhat conditional on the attribute values, which is very cool, but it's a bit tricky to keep track of all the particulars from the JavaScript. Any chance you can add those tables or some way to see, like a graph approach, how the attributes affect the other traits?
ReplyDeleteSo the main backend variables are the six variables under Hexaco, which is a variant of the five-factor personality model, and Haidt, which is Jonathan Haidt's six moral foundations. (I'm skeptical of how well they (or personality psychology more generally) hold up as science but the foundation of five-factor stuff in the lexical hypothesis is that it should at least be useful for capturing elements of how we think about what a consistent personality is like, which is all that a character generator needs to be plausible.)
DeleteSo these backend variables are the main mediators between the mechanical frontend (ability scores, plus alignment - it's in there, maybe I should add a button to let you control it) and the descriptive frontend of the biography. Ability scores and alignment are generated randomly, and then there's a complex dance with the Haidt/Hexaco variables:
Goodness makes you value Harm and Fairness more. Lawfulness makes you value Authority, Purity, and Loyalty more, and Liberty less. Then in turn, valuing Harm more makes you more Agreeable (Hexaco.aa), and valuing Fairness more makes you Honest/Humble (Hexaco.hh). Intelligence and being chaotic both make you more Open to Experience/Ideas. Wisdom and being lawful make you more Conscientious. Being charismatic makes you more extraverted. Emotionality/neuroticism doesn't vary with anything.
Then it goes through possible entries for each sentence, puts appropriate ones in a bin, then chooses one of those at random. Life events can make other life events available but nothing sophisticated like changing your personality or ideals dynamically, which would probably make for better stories.
Oh that's cool that you mapped it to frameworks from personality psychology. I used to do cognitive neuroscience research and before that learned some personality psych in undergrad, and I'm inclined to agree with you. Many of the frameworks from personality psych are good general heuristics, but don't really hold up empirically in a way that is scientifically meaningful.
Delete