You Can Expect Philip Rivers’ Interception Rate to Drop in 2017

I’ve written in the past about the Chargers’ turnover struggles in 2016, but today I’m going to show you the leading cause of INTs (a proof of concept, if you will) and how the moves the Chargers made this offseason could act as preventative measures.

When it comes to interceptions and trailing, many think of it as a Chicken and Egg scenario – which came first? In this article, I’m going to show you that INTs are (generally) a symptom of trailing, not a cause. In addition to the fact that INT rates are notoriously inconsistent (not surprising given how many dropped INTs we see year after year), I had always read this claim but never looked into the nuts & bolts of it all (don’t leave! I promise there’ll be no more puns!).

I gathered 2016 (regular season) play-by-play data and looked at the circumstances of every INT (point differential (aka Game Script), down, etc). To start off, let’s take a look at overall league INT% (INT/pass attempts) by Game Script (points up/down/tied), with the total INT% represented by the red line.

It’s already clear that INTs are more likely to be thrown when a team is trailing by any amount. A team trailing by 1 point is 40% more likely to throw an INT than one that is tied. What is interesting is that the INT rates are not drastically different between any of the trailing categories, or between any of the tied/leading categories. It does a good job illustrating the concept behind the data – the simple fact that a trailing offense is more likely to have the idea that they have to take great risks in order to gain the lead.

Next, let’s take a deeper look. The following table breaks down the 415 interceptions of 2016 by the order in which they were thrown in a game (ie: the QB’s 1^st INT, 2^nd INT, etc). For each category, it includes the average & median game script (point differential), the % of those INTs thrown with a GS of 3 or more, -3 or less, and thrown on 3^rd or 4^th downs.

The first INT thrown in a game should be a good indicator of the factors behind the average INT (and I say average, because obviously over a career QBs have different rates of INTs based on their decision making), and we can see here that the average game script of initial interceptions is negative – trailing by almost a field goal, on average. In addition, 50% of all of these INTs come when the offense is trailing by at least 3, while only 25.4% come when they are leading by 3 or more. Further INTs are even more likely to be thrown while trailing, and less likely while leading. Of the 2 categories with the largest sample size, the % of INTs thrown on 3^rd/4^th down aren’t much off of 1/3, which leads me to think that teams aren’t any more likely to make a decision based on desperation on the “desperation down” than they are 1^st or 2^nd down. Of the 276 1^st INTs, 37% led to a 2^nd interception. While the majority of teams are able to right ship, it’s not uncommon for teams to become even more desperate once they’ve thrown their first INT. (This is not surprising given that some portion of these INTs result in TDs, or great field position for the opposing offense à leading to trailing even further and even more desperation).

Lastly, I want to take a look at a similar table for Philip Rivers’ 21 INTs in 2016

While a single season for a single passer is a very small sample size, we can still see that Rivers’ interceptions were thrown while trailing by even less than the NFL average. His 1^st-to-2^nd “conversion rate” was also much larger than average (54.5% to 37%), as well as his 2^nd-to-3^rd rate (50% to 25.5%). This shouldn’t be overly surprising given El Capitan’s gunslinger mentality, as well as the fact that for a number of years now he has felt he has to carry the offense himself. All of this results in risky decisions, especially when the team is trailing.

I don’t want to end this on a sour note though because I believe the offense is on an upswing. Adding starting-caliber offensive linemen in FA and the draft (in Russell Okung, Forrest Lamp, and Dan Feeney), as well as drafting a receiver sure to add value near the goal line (in Mike Williams), will only help the offense control the ball and maintain a lead. (Add in a likely stronger focus on the run game, and the stability is even more likely). When you consider that last season only 4 of River’s INTs, 19%, came when the Chargers were tied or leading – while 35.9% of league INTs came during the same conditions, fielding even just a stable offense should result in a much lower INT rate for Rivers in 2017. As I illustrated in my last article, a Rivers with a decent supporting cast puts up strong passing numbers (and doesn’t have to make decisions out of desperation). I believe we can all look forward to more of this in the coming season.