kind of obvious, no?!
The New York Time of Aug 28 had a lengthy analysis of the above graphs… Which in essence sums up to the fact that, unless Y is a deterministic function of X, cutting the dataset at X=x⁰ or at Y=y⁰ does not produce the same partition of the population. Graphs used to point out the difference between body mass index and body fat, i.e. that body fat is not a deterministic transform of BMI. An intriguing question that may just be a coincidence: how comes the red dots for both women and men constitute 18% of the population?!