Binning is a bit arbitrary by nature. 10 games, 16 games, 1st half/2nd half, all well and good. 10 is a nice round decimal number. Very Metric. (BTW: the Jets' December 6-4 is really 6-2-2, so in points% terms it's .700 and January was 6-3-1/.650 - 6-4 makes it look like they were just .600 in both bins, but they were better).
16 game bins make sense because that isolates the crazy hot streak to start the season - and why would we want to do that? Because we're cherry picking, of course! Wait, no - it's because we recognize that hot streak as an unusual event - so separating that from what happened after might give us a better picture of what's more realistic for the team.
The fact that the slump came directly on the heels of the streak is a good example of regression. Then take the two bins together and you get a mean: 15-1-0 + 7-8-1 = 22-9-1 = .703 which is closer to reality than the .937 the Jets were rocking in their first 16 games, OR the .469 in their second 16. Right now they're .718, so pretty close.