I don’t have a lawyer, although I probably should.

Watched (live-action TV): Leverage: Redemption 3.8-9: The one where Hurley has to cool the mark, and the one with the Effective Altruism polycule. Parker, or maybe Riesgraf, had way too much fun with that.

Read (manga): Dandadan vol 15 (Yukinobu Tatsu): That particular alien invasion might be over, but something is definitely up at school. First, though, the grandma and her metal exorcists have to help a haunted student. It’s still not clear to me whether anyone except the PCs notices any of the rampant destruction. Also, a lead on the missing golden orb.

Written (game design): 246:

Looking at option #2, you could roll 3d6 for any effect roll. If
you make a 8-, great, you do well above average (5 Stun per die).
If not, if you make an 11-, good, a little above average (4 per
die). If not, but you still make a 14-, then a little below average
(3 per die). If you can’t even make that, then it’s well below
average (2 per die). Body is always Stun/3. (For an even distribution
it should be more like 8-/10-/12-, but 8-/11-/14- numbers most Hero
players already have in their brains.)

Or if we wanted to combine with the success roll, say every odd number
showing on one of the dice gives you a bump, and a 1 gives you another
bump. (Or reverse if you think only rolls that were easy to make deserve
to do better.) That’s 4/6 bumps per die, so 3d6 gets an average 2 bumps,
min 0, max 6. That range is too large for 1 per die, so it’s 2.5 Stun
per die plus 0.5 per bump. Body could be Stun/3, or dice-2 +1 per bump
with hard bounds of 0 and dice*2.

These both use a lot fewer dice, and the calculation can mostly be done
ahead of time (except when something like a haymaker or moveby adds
dice), but it’s more table lookups and the curves don’t match that well.