You are making some flawed observations. For example, 3 seconds in the Tanev example doesn’t count the fact that the Jets had zone time for a fair time before he regained possession and then lost it with a bad pass (and subsequent collision with his own teammate).
The other flaw is that lots of advanced stats aren’t meant to be used for single games because of such anomalies. Do you assume that birth rates for a city are zero if there happen to be no births on the particular day you look at? Or do you maybe look for a larger sample size.
|