Intro
Here I am going to explore the possibilities of implementing personality traits in the game A Fighter’s Festival, where you - as the player - coach fighters into winning fights and growing throughout their careers.
Traits
Looking at the wikipedia page for Trait Theory we notice that there are quite a few personality traits in standard thrait theory. In fact, there are 17 traits:
- Openness to experience
- Conscientiousness
- Extraversion
- Agreeableness
- Neuroticism
- Honesty-humility
- Self-esteem (low)
- Harm avoidance
- Novelty seeking
- Sensory processing sensitivity (SPS)
- Perfectionism
- Alexithymia
- Rigidity
- Impulsivity
- Disinhibition
- Psychoticism
- Obsessionality
We now have a problem in our hands: while the number of available traits by itself is not an issue (well, as far as having a \(R^{17}\) personalty space is not an issue; maybe it is), if the player cannot know what something means (e.g.: what does it do), then it is only detrimental for gameplay.
My gut feeling here is that we should cut down the amount of traits to a manageable number, like maybe 6 or 10. But I do not want to do this blindly: traits should have effective gameplay characteristics, meaning every trait must have outcomes - or else they are just empty and don’t actually mean anything.
Putting it in another way, traits must be first defined by their gameplay value and only then we must think where does it fit in the standard theory; this is not because there is anything wrong with the theory, but because the game is not real life and it needs not to be.
Trait Space
Instead of having to code in 17 different, non-overlapping behaviours, it’s usually better to have a low dimensionality space for game systems, for two reasons.
First, it’s easier to code a few different rules than a bunch of completely different things. That’s the implementation benefit. Second, it also also allows you to reinforce things to the player and make it feel like a solid system instead of a bunch of rules that must be memorized or else nothing works. In short: prefer systemic designs, which are just low dimensionality spaces where the player can infer what they are from the sampling points (in our case, the traits).