[NoOne]

The game uses overall stuff per PA. "They've" said as much in the forums, if not mistaken. (a proxy of some sort at the least)

pitch types do add some sort of flavor to the results. We can see that in the more overt examples with a smaller sample, and our eyes. most are not this way.

e.g. a good knuckleball and suitable ratings otherwise will provide a well below average babip against, all other factors the same.

some pitches may increase hr/9, some may increase bb/9, etc. or lower any of that stuff.

Stuff itself is a component of a few different things. individual stuff ratings, velo, and whatever else the manual and the in-gam editor mention. you can have high individual raw pitch stuff in editor, but have mediocry stuff in profile (profile shows the composite of each pitch as well as overall stuff as is used - if 100% accurate ratings on)

if you turn 100% accuracy on, you can see the resulting stuff per pitch and overall in a precise way and could calculate the weights of each factor.

while you are there, pay attention to the splits of each pitch. that will also help you determine which pitchers will perform better at similar ratings compared to others with similar ratings.

it's possible your leagues LTM/LTs influence how much flavor somethign adds, too. this is why it's so difficult to know anything for certain outside of very specific environment and a large enough sample. whether it translates is another question.

e.g. if a pitch happens to give up more hr/9, but your league only has 500-700hr a year (30tx162g), then that pitch's weakness is minimized and may be a more dominant pitch in that league than in a league with 7000 home runs in same # of games.