Reply to thread

Message: <blockquote data-quote="jon" data-source="post: 878367" data-attributes="member: 15113">Thanks for the good depth.Is this not bias, though?  If sacks and kneel downs affect GB's numbers then they affect every team's numbers. So this is an error (or, at least, a push) that is present in every data record so ultimately not meaningful (or, at least, not enough to annul the finding).  That is, we think the pack at 58/42 is normal next to the league's 57/43.  If, correcting for bias, it is actually more like 65/35 for GB and ~64/36 for the NFL, then GB is still right in the middle.  If we assume NFL playcallers know what they're doing, then MLF's run/pass play selection is fine.  And then there's standard deviation, a measure of how far off of average you can go and still be considered "about average".  I'm not going to nerd wrestle these numbers, but in this data set, with a range of 42.5 to 65.4% in pass plays, I'd guess the standard deviation would describe a range most on this board would consider a very significant difference in play calling.  So if MLF shifted to, say, 53/47 (uncorrected) most fans and commentators would talk about a huge shift that really isn't huge at all--  statistically.  It would be huge, in my view, if by "pass" you mean Aaron Rodgers and you decide to do less of him and expect to win.</blockquote>

No members online now.

Total: 306 (members: 0, guests: 306)