The validity in the "yards allowed" statistic for corners

I have always been pretty back and forth on how valid talking about how many yards guys like Justin Blackmon got on Prince Amukamara in the Nebraska Oklahoma State game this year. Does Blackmon getting 157 yards in that game really mean anything? Or is Amukamara often lining up on one side of the field, and is he not necessarily facing Blackmon? Well, I decided to test the validity of that statistic. I figured that if there is validity in the statistic, then corners who that allow a small amount of yards per game according to the statistic should be drafted higher than corners whom allow a lot of yards per game according to the statistic. I decided to see how many yards per game each corner expected to be drafted allowed to number one receivers, excluding games against FCS teams or games decided by 21 points or more, in which a corner may be subbed out pretty early in the game and his backup used to avoid injury, or the team with the lead stops throwing the ball because their lead is so great. I figured, that if the top corners in the FBS did well in this statistic, then the statistic is valid. Here are my results.

Some quick background information on the statistic; there is no need for it at the NFL level. To my knowledge, ESPN’s K.C. Joyner was pretty much the first to start using how well receivers did against a team as a way of measuring how well that corner produced. But the Football Scientist took it one step further; he started watching every pass thrown during the entire NFL season and seeing how many yards each corner allowed, how many yards per pass he allowed, how many balls were thrown toward each corner, how many catches allowed, the percentage of balls that corners allowed receivers to catch, and even every one of these statistics for receivers and tight ends, like the amount of passes thrown in their direction, etc. Actually watching how many yards each corner allowed during the season is the ideal statistic for corners; however, it is extremely time consuming, and it is almost impossible to do at the college level without because of the 119 games and the difficulty involved in getting game film on teams like Western Kentucky and games against FCS opponents. I am trying to create the next best thing.

Once you factor in the fact that guys like Johnny Patrick and Brandon Hogan play in the Big East and their stats shouldn’t be taken quite as seriously, it’s pretty clear that corners that did well in the statistic were projected to be drafted higher than corners who weren’t, with only four players that could be considered exceptions: Prince Amukamara, Patrick Peterson, Darrin Walls, and Jimmy Smith.

I am not going to sit here and bash Amukamara. If Amukamara is examined more closely, it’s clear that he wasn’t as bad as the stat suggested; 42% of his yards allowed came on 3 single catches. The 80 yard pass to Justin Blackmon, for example, was a flea flicker, and it would not have gone for quite as many yards if Amukamara had made the tackle that he missed, Oklahoma State wasn’t at its own 20, or if the safety that was fooled by the flea flicker had stayed in position to make the tackle. Though I do think Amukamara is overrated, it’s actually because he has mediocre production for a first round prospect, and he doesn’t have the jaw-dropping physical tools that can make up for the above average overall production, but still mediocre for a first round pick. He is 6’0 with a 4.49 40 yard dash, which is completely average, but far from the “once in a lifetime,” talent a corner should have to be a mid first round pick. Now, if he allowed 35 yards per game I would concede that he was dominant enough to make up for those average physical tools, but he didn’t, and he isn’t dominant enough to make up for those physical tools.

Peterson frankly doesn’t surprise me. I have always said he was extremely overrated, and, to tell the truth, I could have booked him for 81.4 yards allowed per game if I had counted the North Carolina game against him. I couldn’t really get an idea whom North Carolina’s number one receiver was during that game. Though Dwight Jones ended up leading that team in receiving yards, he only got 104 yards in the first 5 games combined, and didn’t appear to be the number one receiver at the time. It was either Jheranie Boyd or Erik Highsmith. Boyd ended up getting one more yard than Highsmith on the season, and if I had counted Boyd’s numbers against Peterson in that game, it would have been an extra 221 yards to his tally, and an extra 23.2 yards per game allowed. When I look online, it appears that Boyd was the number 1 receiver for UNC at the start of the season. Even though Highsmith is listed as number 2 on the depth chart, he is listed as the staring number 2 receiver, only knows the routes that he has to run as a split end (weak side wide receiver). So, when Little was suspended, as Little’s backup that knew the routes that Little ran, Boyd probably would have taken over. But I don’t think I am going to count the fact the Boyd got one more yard than Highsmith against Peterson. But what’s undeniable, it would be pretty tough for Boyd to get 221 yards in a game without being the number one receiver.

And finally, Darrin Walls always seemed really underrated to me. Not only did he master pretty good production during the 2010 season, but he is 6’0 tall and runs a 4.45 40 yard dash. I realize that isn’t amazing, but guys like that don’t grow on trees, either, and they’re hard to find in the 7th round. He has always seemed pretty underrated to me because of that production, solid physical tools, and the fact that Notre Dame plays a pretty tough schedule. I’ll get more into him later.

The only truly inexplicable exception to this rule is Colorado’s Jimmy Smith. Looking at these statistics, and Colorado’s pass defense statistics in general, I thought Smith was an extremely overrated prospect. Since my mock draft last week, I have seen every Colorado game available on ESPN3, and Jimmy Smith actually looked brilliant in coverage. He allowed about 2 or 3 catches in all of the games combined.  The only true explanation for why this happened is that most Big 12 teams run a spread offense, and many number one receivers actually play in the slot much of the time (Ryan Broyles is a big one that comes to mind). Also, I mentioned that I didn’t count games decided by three scores or more because one team would start passing on every down and starters would even get subbed out if one team was really dominating. By sheer coincidence, the 3 games in which Colorado allowed under 15 points were all games decided by 3 scores or more because Colorado’s offense showed up to play.

I realize that people will doubt the validity of the statistic. Don’t write me off quite yet. The statistic appears valid according to those results. I know people are going to ignore the statistic, and say that corners too often line up on one side of the field, but I don’t think this is very valid. Typically, a team’s best corner will cover the strong side outside receiver (aka flanker), who is usually the team’s number one receiver (except in the Big 12 South). Try to remember the fact that coaches are going to want a real dominant corner on a real dominant player. For example, if you are Urban Meyer (or, I guess Will Muschamp), and you are facing Georgia, you are going to want a corner as dominant as Janoris Jenkins matched up against A.J. Green as much as possible, and your second corner, sophomore Jeremy Brown, matched up against Green as little as possible. If you watched the Florida Georgia game closely, you would clearly see that Jenkins never leaves Green. He gets and interception off of him and holds him to a mere 42 yards, his worst game of the season. Without watching the game (even though I did), should I even have to ask my readers if they are going to contest the idea that Jeremy Brown played a large role in stopping Green? I don’t think so. It was clear that Jenkins was matched up against Green, and using that game in Jenkins’ favor is completely valid. Now, if a coach has two corners that are practically interchangeable, such as Virginia’s Ras-I Dowling and Chase Minnifield, then the statistic is invalid. Each corner will match up against the number one receiver at different parts of the game. But let’s be honest; take the Alabama LSU game for example. Les Miles would eat grass before he let true sophomore Morris Claiborne line up against Julio Jones when he has Patrick Peterson on the roster (okay, that was a bad example). But still, there is no way he would let Claiborne line up against Jones if he can make Jones line up against Peterson. Julio Jones got 89 yards in the game. For the sake of proving my point, after I wrote this article, I watched the game. I used KC Joyner’s sabermetrics of figuring how many yards Peterson allowed and how many yards he gave up per throw, and watched every ball Greg McElroy threw. Julio Jones had 6 of his 9 catches against Peterson (he had 10 catches on the game, but one was while Peterson was on the sideline, taking a one play breather), for 52 yards, and Peterson also allowed a 37 catch to Marquis Maze in the game (Technically, the catch didn’t count. Peterson did a blatant pass interference on Maze, pulling his right arm around his back, and the officials ruled he made a one-handed catch, but when the play was reviewed, they ruled it wasn’t a catch. Peterson still got a 15 yard pass interference penalty called against him, but at the NFL level, that’s a 37 yard penalty. Considering I’m evaluating him for the NFL, I’m considering that a 37 yard catch). So let’s add that up; 37 yards for Maze, 52 for Jones… that’s 89 yards! Coincidentally, the exact number the stat counted against him. That stat isn’t usually going to be that perfect, but I think it is good enough to be considered valid. Miles did try to make sure that Peterson was lined up against Jones as much as possible. Jones still got yards. It was Peterson’s fault that he did. Now, are you going to tell me that stat is invalid? Please.

Overall, I think there is enough validity in the statistic for me to feel comfortable about using it to evaluate corners to some extent during the draft process. Clearly corners that did better in the statistic were considered better players by the draft community, and I think it makes the stat reasonably valid.

Tags: 2011 NFL Draft Cornerbacks Jimmy Smith NFL Scouting Combine Patrick Peterson Prince Amukamara

comments powered by Disqus