>>56375484I calculated all the way up to 18th order averages and got these results. The sequence of averages is quite consistent with most of the change happening in the first two steps. Perhaps this isn't so surprising; a type being good defensively depends on how many types it resists and how many of those resistances are against "good" types. Resisting types that are strong against types that resist types that are strong against many types doesn't seem to have a big impact (thankfully).
As a sanity check, I did this same analysis on a type chart consisting of just Grass/Fire/Water. Unsurprisingly, the sequences of multipliers were constant since the type chart is perfectly balanced. Continued in the next post.
