http://www.washingtonpost.com/blogs/foo ... s-playing/
Stats confirm how well Robert Griffin III is playing
Posted by Brian Burke on October 4, 2012 at 7:00 am
Seven point one. That’s Robert Griffin III’s net adjusted yards per pass attempt, which accounts for sack yardage and docks 45 yards for each interception. That number by itself may not mean much, but it’s the very best in the league so far—better than the total compiled by Tom Brady, Aaron Rodgers, Drew Brees and everyone else. Griffin hasn’t faced the toughest defenses by any stretch, but it’s obviously a very good sign. Stats like these merely confirm what we’ve seen with our own eyes here in Redskins Country, but they’re essential to put a player’s performance in context and compare him with all the other quarterbacks around the league, and the other rookies in particular.
So let’s dig a little deeper. First we’ll look at Expect Points Added (EPA), a stat that measures productivity according to the down/distance/yard line situation. Griffin ranks 5th in the league with 0.24 EPA generated per play–passes, runs, and scrambles included. Fellow rookie Andrew Luck is 12th, and Ryan Tannehill, Brandon Weedon, and Russell Wilson rank 30th through 32nd.
Success Rate (SR) is a stat that filters out the effects of large plays, which are often the result of broken coverage, tipped balls, or other relatively random causes. SR simply measures the proportion of plays that make it more likely, rather than less likely, that a player’s team will score. Griffin is 9th in SR at 52%, well above average. The fact that he is 5th in total production (EPA) but 9th in SR indicates his production is coming disproportionately from big plays rather than long strings of short gains. Luck is 17th in SR, Wilson is 22nd, Tannehill is 30th, and Weedon is 32nd.
Win Probability Added (WPA) measures each play’s impact on the chances of winning the game. It largely measures heroics, because it is highly context sensitive, accounting for score and time remaining. Griffin is 16th in the league with 0.19 WPA/Game. This means that Griffin’s performance (in concert with the rest of the offense) would, on average, take an otherwise average 8-win team and make it into an 11-win team. In full disclosure, that’s a little overstated because of how offenses have taken a firm upper hand over defenses in the last couple of years. Luck tops the rookie class with 0.32 WPA over 3 games and is 4th overall. Tannehill is 28th, Wilson is 32nd, and Weedon is 33rd.
Griffin is clearly moving the ball with both his arm and legs. He has generated just over 30 net EPA (net points) on pass plays and more than 13 net EPA on designed runs and scrambles. On a per play basis, his runs and scrambles are more potent, and by a large margin–about 50%.
The only statistical weakness Griffin’s numbers show is a high reliance on receiver yards after catch (YAC). Although his overall pass efficiency leads the league, he is 13th in “Air Yards” per attempt–passing yardage minus receiver YAC. Of the quarterbacks who have 100 attempts through week 4, Griffin has the third highest percentage of YAC. That’s not a bad thing in itself—more yards is always better. But a large percentage of YAC is often an indication of an over-reliance on screens, dump offs, and short routes, which themselves can be symptoms of poor pass protection. Griffin has thrown deep, defined as attempts of greater than 15 yards through the air, on only 13% of his attempts, 30th among league quarterbacks. This is also likely the largest factor in his very low interception rate.
He’s only bound to improve throughout the season, but it might not show up in the numbers as he faces tougher opponents in coming weeks. Even rookie quarterbacks who turn out to be franchise-level players don’t rarely have the kind of early success that Griffin has enjoyed. If one thing is certain, it’s that, for now, the Redskins have found the right formula for their rookie superstar.
Brian Burke is the creator of Advanced NFL Stats, a Web site about football, statistics and game theory.