No Sean McVay in PFF’s Top-10 coach rankings is just wrong

The expansion of data analytics into virtually every aspect of an NFL team’s operation is beginning to cross the line of meaningful data into meaningless nonsense. The LA Rams use the comparative forms of data analytics to compare and contrast players from known NFL career tracks to project into the unknown.

For veterans, that can take the form of valuing their future performance to set the range for a new contract. For rookie prospects, that can take the form of who is drafted and who is not.

But there are other uses for the data, and in the end, the misuse of that data is just as easily presented as ‘proof’ of incorrect conclusions. Can I give an example? Well, I actually can. ESPN created a new data analytics algorithm, and concluded that LA Rams All-Pro defensive lineman Aaron Donald was ‘just average’. Let’s be brutally honest. If you design a data algorithm that leads you to believe that Aaron Donald is just average? You’ve buggered up the algorithm.

The reception for that report was a social media firestorm. Bottom line? Have a control (test conclusion) set up to ensure that you don’t get lost in the forest of data for the trees of ‘isn’t this awesome?’ data points.

Just creating mathematical manipulations of data is not necessarily a way of adding value, meaning, or narrative, to data. if you did a study and discovered that you got a pay raise in February and that on each of your last three pay raises it snowed on that day, should you then conclude that snowy days compel your employer to increase your pay? If you do, then you are ready for a job analyzing NFL data it seems.