I’m not suggesting the eye test alone is fool-proof. I’m refuting the idea that a stat w/arbitrary parameters that doesn’t account for critical variables has any meaning, much less is *superior* to observation.