
The Dallas Cowboys aren’t doomed if they pay Dak
Point 1: Sample Size
There are a couple important things to point out right away: First, the bad “science”. What separates this from worthwhile analytics is this isn’t analytics at all. It’s just reading a simple historical fact from a ridiculously small sample size all while completely ignoring context and other variables
Would it be nice to pay less rather than more at the QB position? Of course, but not at the cost of declining at the most important position.
When we use Super Bowl wins as the only successful outcome for a team, we’re calling all other outcomes and their blueprints, failures. Consequently stating the Super Bowl winning team has the only acceptable blueprint for victory. Never mind this completely ignores the boatload of variables that go into winning and losing a football game.
Also consider we’re only looking at Super Bowls under the last CBA, and now we’re dealing with an n of 9. No mathematician would accept any conclusions drawn from this.
Point 2: Randomness and other variables
Go back and watch an NFL game. Count all the funny bounces the ball takes any given week (both literary and figuratively). Every game displays a handful of random events and/or turning-points that can keep the better team from winning. Obviously the better team wins more often than the lesser, but that just illustrates why a larger sample size is critical when we try to draw conclusions.
Rodgers was the top QB in 2014. Would his team have recovered that onside kick if he was cheaper?
— Mina Kimes (@minakimes) May 11, 2020
Brees was near the top in 2018. Did he miss that PI call?
If Brady's ranked 9th instead of 12th/15th--is that the difference btwn the Pats winning and losing?
Mina Kimes cites in her tweet above just a few of random events that have prevented top-paid QBs from winning at the highest level. The Falcons switching to prevent offense and allowing a handful of fluke plays has nothing to do with Matt Ryan’s compensation at the time. A missed pass interference call has nothing to do with Drew Brees’ percentage of the cap either.
The point is this extremely small sample size would have been busted repeatedly if randomness just went a different direction – Randomness that so clearly had nothing to do with the QB’s salary.