Data and stats can tell you whatever story you want to promote
Seen this so many times at my work. There’s some bone-headed decision and the people in charge are like “look guys we ran the numbers”. But the methodology is messed up somehow, or they just ignored / misinterpreted the numbers while pretending they were following the data, or it doesn’t bear out in the real world; etc.
When data and common sense disagree you’d better be damn sure in the data.
It’s called “data chauffered”: instead of following the data, tell it where it needs to take you