Advanced Metrics: Interceptions & Atlanta Falcons’ Matt Ryan

A couple of months ago I did a piece on advanced metrics regarding Falcon’s offensive-line and Atlanta’s decision making to replace both coordinators, among other coaches they let go. From time to time I come across statistics on Football Outsiders that I feel absolutely must be shared with the Falcon’s Faithful. This is one of them.

The stat is entitled ‘Adjusted Interceptions 2011’ and basically it gives a perfect case scenario of interceptions. Every quarterback throws passes that hit the defender right in his hands, and should be an easy pick, but for whatever reason, the defender cannot hold on. In those situations (in this statistic) an interception is added to the quarterback’s season statistics. By the same token, if a pass is directly to a receiver who should have easily made the reception, but is tipped and a defender makes the pick, an interception is deducted from the QB’s season total. The same thing is done for Hail Mary passes or certain types of ‘desperation end of game’ passes; they don’t accurately reflect the ability of the quarterback and therefore the quarterback should not be punished for it.

This set of statistics doesn’t have quite as much involved in it because there is really no way to determine how play-calling affected the interception numbers, and certainly doesn’t lend any credibility to an argument that Matt Ryan’s performance led to the firing of Mike Mularkey. It shows that the interception totals that Ryan had this year are probably what we can expect from him for the rest of his career, although they may improve slightly.

Ryan was intercepted on 2.1% of his passes in 2011. He threw 12 interceptions.  That is actually a remarkably low percentage and total number of interceptions. Granted there are individuals who had off the chart seasons like Aaron Rodgers who was at 1.2% and 6 overall, and Alex Smith who simply didn’t turn the ball over in 2011 (1.1%, 5). But according to the adjusted interception rate, Ryan was only intercepted on 2.0% of his passes, Rodgers 0.8%, and Smith 1.3%. Since the numbers changed so little for these quarterbacks, it shows that their numbers weren’t artificially inflated or deflated. By that, I mean that there were not a significant number of easy interceptions that were dropped, nor were a significant number of their interceptions last second in the game, desperation type throws. For the most part, their interception totals were a fairly accurate portrayal of their performance.

The same cannot be said for the vast majority of the league. Most quarterbacks that had a very close Interception rate-Adjusted interception rate-ratio threw a whole lot of interceptions. To pick on Josh Freeman, he threw 22 interceptions. an additional two easy interceptions were dropped, but also two ‘desperation’ interceptions were deducted from his total. Hence, his Adjusted Interception Number is 22 and his Adj. Int. percentage is 4.0%; exactly the same as his true Interception percentage (4.0%).

Other quarterbacks had a great many interceptions dropped that should have been intercepted. I will pick two quarterbacks for comprehension purposes. Cam Newton threw 17 INTs on the season. A whopping 9 of those were dropped, while only 1 was considered a ‘desperation’ type of interception. As a result, there is a huge step up from his Interception percentage (3.3%) to Adjusted Int. Percentage (4.9%). If football were executed perfectly, which is reflected in a quarterback’s Adjusted Interception Number, Newton would have had 25 Interceptions on the season. That is a bold increase. The same goes for Matthew Stafford. He threw the ball a heck of a lot in 2011, and ended the season with 41 TDs. However, things would not have looked so great if his Adjusted number were his true interception total. With 9 dropped interceptions, and zero tipped/’desperation’ passes, his Adjusted Int. Percentage rose from 2.4% (very respectable) to 3.8% (worse than Mark Sanchez) with an Adjusted total of 25 interceptions.

This brings us back to Ryan. He threw 12 INT’s on the season; he had 1 ‘desperation/Hail Mary’, 2 dropped INTs, and 2 tipped passes. His Interception percentage went from 2.1% to an Adjusted Interception percent of 2.0% , and  from 12 INTs to 11. This tells us that Ryan is very consistent, forces bad throws far less than other quarterbacks, that his receivers are guilty of a pair of interceptions that went in a bad column of Ryan’s statistical line, and that he isn’t afraid to throw the ball up to his receivers in the end-zone late in the game (they just may not come down with it). These statistics reward QB’s who try to win games, and not the ones who try too hard to pad their stats and avoid interceptions when their teams are losing.

What is the bottom line here? What actually happens on the field is what matters, and that the team that executes wins. It doesn’t matter how you finagle the numbers to make them say what you want them to say, some teams come up with the key interceptions, and some don’t. Some quarterbacks get lucky and don’t get intercepted, propelling their teams to victories; others don’t put themselves and their teams in bad situations where they could turn the ball over. Overall, Ryan is very good about not turning the ball over; we can only hope that trend continues for the long haul.