StatsDetective 9 | NFL MVP + QB Wins

This analysis was fairly straightforward and the results sat directly on the fence. It is not surprising that measures of QB performance (in this case passer rating) have a big impact on the outcomes of games. It is also no surprise that these same measures do not make up the whole story or even half of the story. However you feel about the issue, these results will probably not sway you.

One part of the argument that I did not touch on because it is fundamentally philosophical, is how believing that quarterback play contributes to winning is a paradox. Let's say you believe that quarterback performance is worth 40% of a win (because you watched the video). That means that the opposing quarterback's performance is also worth 40%. So don't the two influences cancel each other out? Framed in this way, the argument opens up an entire landscape of problems. Do we have a reliable way of showing a player's 'performance against' like a strength of schedule for one player? The answer is no, but the problems don't go away even if we did.

Even the argument that better QBs will get more opportunities to throw and thus have an outsized impact falls flat. Do a few extra runs decimate the effect of a QB? Something for a much longer video, Maybe Greg or Nikki will have a thought on this.