Why Are Interceptions Overblown Now?

Is social media or efficiency stats hurting NFL offense

Honestly I had a little writer’s block after I wrote this. I rarely if ever get writer’s block, but there was just something about the emotional payload following that piece. Whatever followed it in drafts felt awkward by comparison. So to combat my block, I’m drafting off a smart thought sent in by a subscriber.

Interceptions

In reaction to my Dieter pod, Commenter Grant has a theory on why interceptions now take up such disproportionate focus in NFL commentary.

Great conversation. The interception take here is spot on, but it’s a symptom of a much larger problem. In recent years, the NFL has decidedly moved toward offensive efficiency metrics at the expense of traditional productivity ones, and the interception is caught up in this unfortunate vortex. The media has become obsessed with interceptions and their belief that they are the single greatest offensive evil in football.

This inaccurate view has become so pervasive that it has polluted public perceptions and even infiltrated NFL offensive game strategy. By believing that every single INT is a game-changing cataclysmic event (the vast majority are not – essentially punts as pointed out by Ethan), the media has overstated the true negative impact of an interception, while ignoring the many positives that come from deeper throws - including stretching the defense even when unsuccessful – despite carrying some risk.

One direct impact from this shift is that the all-in QB metrics (Passer Rating, QBR etc.) have taken on mythical proportions. It is not a coincidence that NFL offenses have become more stagnant - less productive in terms of yardage and points - over the past 5 years at the same time these all-in efficiency metrics have risen in importance to become the sole source of media truth for QBs.

Efficiency is a critical lens for sure, but only when examined in conjunction with productivity. If you aren’t productive at gaining yards and points who cares how efficiently you are failing. Efficiency without also evaluating productivity is misleading. Both are needed.

The all-in QB metric formulas are negatively over indexed for INTs and positively over indexed for Completion% and TD%. In fact, it has been reported that a single INT under the NFL’s Passer Rating equates to a value loss equal to 100 passing yards. The fact that Aaron Rodgers was aware of these metrics and adjusted his play accordingly as mentioned here is a proof point for how important these metrics have become.

Thus, the easiest ways to optimize around these measures – and increase their ratings - is to throw shorter and safer less productive passes….precisely what has happened. The focus on efficiency and away from productivity metrics has led to an avalanche of shorter, safer routes, quicker releases, increasing checkdowns ironically applauded as “thoughtful progressions” and more back shoulder throws in the corner of the end zone when a run would seem the superior and lower risk choice.

Fans now routinely decry their team’s “dink and dunk” offense that fails to produce points or wins.

Take the now ubiquitous WR screen ...where the QB tosses the ball to a WR near the sideline 2-3 yards behind the line of scrimmage. They are almost always stopped for little gain or even a loss as the DBs easily identify them and attack the WR with the sideline serving as an added defender. Since these screens showed up in large numbers a few years back, data analysts have consistently called them out as being negative from an EPA productivity perspective and a waste of limited offensive opportunities in a game.

Yet, OCs continue to call them – multiple times each game - week after week, with the exact same poor outcomes. Why is this the case when the League we are told has become more “analytical”? Because these throws are safe – essentially extended handoffs – that count toward the passing efficiency metrics as a safe completion without an INT. They reflect an efficient if wholly unproductive play call.

Over the past 5 years Yards per Attempt (YPA), Air Yards/pass, EPA/pass, and INT% have all gone down – along with yards and points. What has increased? Gaudy all-in metrics for QBs that play it safe in the passing game and toss fewer yards per game for increasingly unproductive offenses.

The key QB stat for offensive passing productivity has always been some variation of yardage and YPA with excessive sacks and INTs accounted for - or the answer to the question, “what is my return for investing downs in the passing game versus alternatives?” Yet, over the past 5 years in the NFL, productivity measures have steadily gone down as the focus has moved elsewhere.

To increase offensive production, teams need to refocus on productivity too – a balance that includes yards per game, yards gained per down and resulting points. This requires higher EPA, YPA, Air Yards, Average Depth of Target (ADOT) and, yes, a few more INTs...risk/reward.

If you risk nothing, you risk everything regardless of what those all-in QB efficiency metrics are telling you.

This could be another sports instance of trouble made by Goodhart’s law, the theory that, “When a measure becomes a target, it ceases to be a good measure.”

Grant and I are completely aligned on how overblown interceptions have become, but his theory on causation had never occurred to me. I suppose that’s because I perceived QBR, however it’s calculated, as a beneficial reform. It’s at least better than antiquated (but still pervasive) Passer Rating, which fails to incorporate sacks, fumbles and rushing gains. How insane is that, by the way? This is a pedantically nerdy cause, but it drives me crazy that a stat as obviously bunk as Passer Rating still gets regularly cited. For the record, if forced to choose, I’m an EPA guy (Expected Points Added).

My theory on modern interception trepidation is less grounded in these nouveau stats and more in the social media effect bad plays garner. Nobody ever wanted interceptions, and coaches have always been obsessive about mistake avoidance, but back when I was a kid, many interceptions happened in the sports periphery. Your main (only?) source of highlights was SportsCenter and those clips trended positive. If a quarterback outside your regional TV window threw a bunch of picks one Sunday, it was almost like it never happened. Now? If, say, Justin Herbert has a few embarrassing plays against the Raiders, it’s all over Twitter/X for people to gawk at.

NFL coaches would probably say I’m being ridiculous to assert that this newly accessible embarrassment has a strategic impact, just like they’d say Grant’s misguided in blaming efficiency metrics, but human beings are quite sensitive to criticism, at scale. By the way, while correlation isn’t causation, the decline in NFL offensive production appears to mirror the rise in social media.

To be clear, interceptions aren’t good, and there are contexts for when a player should be especially avoidant of them. It just so happens that moving the ball forward is usually the main priority and it’s actually possible to have an offensively explosive game if three of your 11-12 drives result in picks. The conventional football wisdom on how horrific interceptions are reminds of when baseball professionals were overly worried about strikeouts. The analytical revolution in baseball tempered those overblown fears. Why hasn’t this happened in football? I’d hazard that a strikeout almost never makes for an entertaining video clip, whereas an interception often does.