clock menu more-arrow no yes mobile

Filed under:

No, Brent Grimes is not the best double-move cornerback around

If you buy something from an SB Nation link, Vox Media may earn a commission. See our ethics statement.

NFL: Carolina Panthers at Tampa Bay Buccaneers Kim Klement-USA TODAY Sports

The past days one statistic keeps making the rounds: Brent Grimes’ supposed awesomeness on double-move routes, as per Pro Football Focus’ charting. You see, their Eliot Crist broke down cornerbacks’ performance by route types, and Grimes allowed no catches on double moves. He even had three passes defensed on those routes.

Grimes had deep responsibilities on all five of his double-move targets. On one play he was the man defender, while on the other four he had deep-zone responsibilities. Grimes broke up three of the five passes on his way to earning the highest PFF grade for any corner versus double-moves.

Here’s the thing: this is based on five plays. Five. Brent Grimes did a good job on those five plays, apparently exclusively as the deep player, but this doesn’t really mean anything. For all we know, those five plays just featured poor throws, or pressure, or great underneath play that had nothing to do with Grimes. Or he just got lucky.

Even if all five plays had a positive outcome because of Brent Grimes’ play, that still doesn’t mean much: it’s really easy for a player to do well on five plays, and less well on a slew of other plays. In fact, given the amount of cornerbacks in the NFL, that will be true for a few of them every year — and perhaps Grimes just got lucky this season.

Stats like this could be useful in aggregate, but not on tiny sample sizes like that. Which is often the problem with Pro Football Focus: they take some potentially useful (if often subjective) data and then hype it as the be-all end-all to football evaluation. And somehow, then, their grades are used as ‘the’ objective ranking of players. Which, well, no.

None of this is meant to take away from Grimes, by the way. He had a good season. Just...use different statistics to prove it, please.