The Curious Incident of the Pass On the 1st Down

Passing is king in the NFL, but did the Super Bowl Champion Philadelphia Eagles exploit a market inefficiency on 1^st down or did they pass a threshold that limits their efficiency? The eternal debate concerning run:pass ratio is one the football community will likely never settle. If passing is king, why do teams continue to run at a higher rate on 1^st down and 10? If the success rate is higher for passing in that situation, it would stand to reason that more teams would make a hard shift in that direction. But is too much of a good thing actually bad? While airing it out is clearly more productive on a wide-scale, it does have warts with a fresh set of chains.

You can’t pass ceaselessly, it will never work like that. For example, play-action is an effective weapon for an offense. While there’s no proven correlation between “establishing the run” and play-action efficiency, a threat still needs to exist. The conversation has shifted so far towards pass heavy being the way, that there’s been an over-correction and oversimplification of the impact of the run game on wins, losses, and overall offensive efficiency. The ground game is effective on 1^st down as a lower variance option for offenses, despite what “success rates” will tell you. I’m a relative newcomer to applying success rates to research, but I find its application fascinating and useful, yet slightly flawed.

There’s another discussion to be had about “success rate” regarding which threshold is best and the important of those successes. Yards are yards and each small gain makes an impact on future success. Analytics in football is still in its initial stages and it’s possible a new, more effective method of calculating these gains will come to light soon. In the meantime, I’m using Football Outsiders method of 45% of the yards needed on 1^st down as a success, 60% on 2^nd down, and 100% on 3^rd and 4^th down.

Looking at the entire league in 2017, I isolated 1^st & 10 plays, non-red zone, excluding non-plays (spikes, kneels, etc) and then added league-wide 2^nd down conversion rates on the resulting down and distance. “Positive failure” reflects gains of 1-4 yards.

There are some interesting takeaways from this 1^st down research:

• Teams are twice as likely to achieve a 0/negative result when passing
• Nearly half of runs are “positive fails” (gains of 1-4)
• 4 in every 5 runs is for positive yards
• The conversion rate on the ensuing down increases from 22.4% to 34% before becoming a “success”
• Pass plays are 14% more successful and achieve 10+ yards at 19% higher rate

The high variance of passing against the stable positive production comes as no surprise. Passing averaged 7.08 per attempt while running came with a much lower 4.32 average, further proving that passing is more fruitful. The Eagles had a narrower gap on 1^st down, despite leaning heavily on the pass (56.07%, 3^rd in the NFL). Through the air they averaged 6.20 (24^th in the NFL) with the ground game churning out 5.35 yards per carry (2^nd in the NFL).

With their 1^st down on 1^st down conversions ranking lower than league average through the air but 1^st in the NFL on the ground, the question remains why the Eagles would lean so heavily pass when running was the more efficient option for them based on league performance.

Ultimately, there’s more context needed. As I’ve pointed out before, the Eagles decision-making is heavily influenced by box numbers, so this outlier could be a result of box numbers. Dealing in the data we have at the moment, this is the second year under Pederson that the Eagles offense has ranked in the bottom half of the league for converting 2^nd downs into 1^st downs. The same goes for converting a 1^st down on 1^st down through the air. It’s not as a sexy of a stat as 3^rd down conversion, but the Eagles may look at the analytics and decide to morph their philosophy on 1^st down to improve upon their overall efficiency. Perhaps they already have.

In the playoffs they flipped their run:pass ratio on 1^st & 10, running 54% of the time, an 8% swing. They also increased their 2^nd down conversion rate to 37.7%, up from 13.2%, 3^rd best out of the 12 playoff teams. With a smaller sample size it’s hard to tell if that was a shift in philosophy or mere coincidence. Time will tell, but the Eagles are nothing if not receptive to their own analytics.