This year’s NFL MVP race ultimately came down to two players whose careers were somewhat in doubt at the end of the 2011-12 season as a result of major injuries. Peyton Manning missed all 16 games after undergoing neck surgery, and Adrian Peterson tore his ACL in the Vikings’ final game. Peterson ended up taking home the hardware, but there’s a reasonable case to be made for Manning’s candidacy, especially with the difficulties of comparing players across positions. In this instance, how does one accurately compare value added to one’s team?
Todd Cioffi, director of technical training at Navis Learning, set out to answer this question at the Globe-hosted NFL Hackathon on January 19. The Hackathon, an eight-hour event that paired analysts from all over Boston with ten years of NFL play-by-play data, prompted participants to investigate some topic of interest and work on their projects in a collaborative environment. Cioffi’s approach to the problem was to consider both what kind of supporting cast Manning and Peterson were working with and how their peers around the league performed relative to them.
The graphics below, produced using the Stats Lab on nfl.com, compare Peterson’s and Manning’s production in 2012-13 with the league average at their respective positions, illustrating by just how much they outperformed their competition.
Compared to the league’s other running backs, Peterson’s numbers were far and away superior. He led the league in four of these categories, and his explosiveness is perhaps best captured by the substantial gap between him and the league average in rushes over 20 yards. While Manning was near the top in almost every relevant statistical category at his position, he didn’t exceed the average by nearly as much—though part of this may be due to the realities of playing the running back position, which means dealing with injuries and splitting carries.
But the key information Cioffi used to make his argument was what Peterson accomplished when Percy Harvin went down in Week 9 with an injured ankle. To that point, Harvin had been one of the league’s most productive players with nearly 800 total yards, bolstering what was otherwise a substandard passing attack and taking the focus of the defense off of Peterson. Without Harvin, Peterson was essentially the only serious offensive threat the Vikings had.
Yet, in the last seven weeks of the season, Peterson’s numbers actually increased. Peterson rushed for 1,140 yards with Harvin missing from the offense, good for 162.8 yards per game, as compared to 957 yards (106.3 per game) with him in the lineup—a 53.2 percent increase, as Cioffi noted. For his ability to raise his game when his team had a need for it, Cioffi determined that Adrian Peterson should edge out Peyton Manning as the league’s most valuable player.
But Peterson and Manning were in contention for another postseason award: Comeback Player of the Year. The criteria for deciding this award were not strictly performance-based; one also had to take into account the obstacles each overcame to return to the football field.
Both players sustained substantial injuries that Cioffi mentioned might have ended their careers in a different era of medical technology. But one of the main things that differed between their two comebacks were the situations to which they returned once they were healthy. For Peterson, it was essentially like he had never left; he was injured in Week 16 of the 2011-12 season and was on the field for Week 1 the following year.
Manning’s return to the NFL was not nearly so smooth. Not only did he miss an entire season after undergoing neck surgery, he came back to a new team with new teammates, new coaches, and some new schemes. But after a shaky first few weeks, it was like Manning had never left the Colts. He was back to his old self, reading and dissecting defenses on the way to a 13-3 record. Given these circumstances, Cioffi deemed the Comeback Player of the Year award a little more ambiguous. In terms of raw performance, he believed Peterson had the better season, but Manning surmounted greater odds in putting up the numbers he did. Manning ultimately won the award, a result with which I imagine Cioffi had no problem.
Cioffi’s project, which ended up winning the Hackathon, was a good example of the proper role of statistical analysis in sports—not slaving oneself to a spreadsheet, but rather combining data-driven arguments with a healthy respect for the context in which that data appears. We may not be able to quantify how much Harvin’s loss hurt the Vikings offense or the extent to which Manning’s new team threw him off, but qualitative factors will always have a place in NFL front offices and beyond. “Stat geeks” are not the enemy; they’re just a source for more information.
The author is solely responsible for the content.
Stats Driven features a closer look at statistical analysis, sports strategy and trends within Boston sports. Andrew Mooney, a student at Harvard College and an active member of the Harvard College Sports Analysis Collective, is the primary contributor. Email him at firstname.lastname@example.org and follow him on Twitter at @mooneyar.