this was an awesome video:

https://www.youtube.com/watch?v=r2whuz6tkb0

but i have my own thoughts on designing rock paper scissors systems.

one of the drawbacks to “type” systems is that they’re a bit cheap. the essayist here implicitly values the high symmetry of RPS (cf. the section on pokemon), but to me, the symmetry sucks! it’s a necessary evil for the desirable properties of these systems, but pokemon having a matrix with a diverse eigenvector is what gives it a unique texture imo

notably, one of the common design goals cited is forcing variety in team compositions and builds, but a precondition for that is that the different types need to distinguish themselves as more than flavors of the same thing. “the optimal team needs a red member and a blue member” doesnt reduce monotony if red and blue characters play the same in practice.

near the end of the video, RPS dice get a cursory mention, but i think they’re actually super instructive. long time followers know i have a post analyzing how they work, but the short version is that the rps effect emerges when you have “low”, “medium”, and “high” risk dice. low risk is more consistent than high, but medium risk has an edge over low, but the worse case scenario for medium is worse than high’s average case. thus it is intransitive q.e.d.

what i find most useful about these dice is that the intransitivity emerges from the texture of a simpler system. “high number wins” is far easier to get a grasp on than memorizing three arbitrary type interactions, and now grasping the underlying intransitive cycle becomes something that rewards system mastery.

anyway, what i most want to apply this thinking to is “PvE” scenarios, which are imo more common in games. see, many games see you blasting through waves of fodder enemies and confronting bosses, and both of these encounter types are asymmetric; your enemies operate by effectively different rulesets and mechanics.

this matters, because when designing this encounters, i think it’s less fruitful to think in terms of RPS per se and more… rock builds beat encounter type A (big brute enemies), scissor builds beats encounter type B (cutting through hordes), and ideally you want a build that can handle both

here’s an example to illustrate what i mean.

in slay the spire and similar games, you can analyze damage cards as offering frontloaded, area of effect, and scaling damage.

if you balance these in the intuitive way (i.e. frontloaded damage have the biggest numbers), then you create its familiar strategic terrain

• in short solo fights, frontloaded damage cards do best; aoe is inefficient and scaling is too slow
• in crowd fights, aoe shines while frontloaded damage becomes slow
• in boss fights, scaling damage finally shines

the watchword, i think, is fungibility. these different “types” cash out into the same underlying currency

so what i’m curious is what it looks like to recreate the equivalent of a 2‍-​paradox using this sort of intransitive fungibility. a set of encounter archetypes where for each one, you need to efficiently achieve at least two unique tactical objects to consistently win.

creating seven (!) qualitatively different types of damage feels like a tall ask though…

maybe adding another axis‍ ‍‍—‍ factoring in different kinds of defense‍ ‍‍—‍ would make it more tractable?