In week 1, Russell Wilson completed 70% of his passes, going 14 for 20. But in the modern NFL, completion percentage is meaningless. What really matters is first down percentage. Wilson took 4 sacks, and only 6 of his 14 completed passes picked up a first down. Therefore, Wilson picked up a first down on only 25% of his 24 dropbacks! Passing first down percentage and completion percentage are supposed to measure the same thing — how well an offense can consistently move the ball via the pass — but there can occasionally be a huge chasm between those numbers.

To be sure, passing first down percentage is hardly a perfect stat. Wilson completed an 8-yard pass on a 1st-and-10, a 7-yard pass on a 1st-and-10, and even had an 11-yard completion to Chris Carson that doesn’t get recorded as a first down because Carson fumbled. On the other hand, 5 of his completions were clearly not successful: a -3 yard completion on 2nd-and-long, a 2-yard completion on 3rd-and-17, 5- and 6-yard completions on 3rd-and-9s, and an 11-yard completion on 3rd-and-26.

Using success rate instead of passing first down percentage would be an improvement — you’d move the two long gains on 1st-and-10 into the success range, and say that 8 of his 24 pass plays were successful. And, of course, this isn’t all the quarterback’s fault or blame: it’s not his fault that Carson fumbled, it’s possible some of the 4 sacks weren’t his fault, and almost nobody is completing a 3rd-and-26 (on the other hand, he probably shouldn’t get credit when his receiver makes a great catch, or his offensive line blocks for much longer than average, or his receiver gets wide open, etc.).

So let’s not let the perfect be the enemy of the good: passing first down percentage is much better than completion percentage, and let’s leave it at that. You can view some of the outliers in NFL history in these two stats here.

A year ago, I looked at the relationship between those two statistics, and so I wanted to update that through the 2018 season.

This first chart shows every season in the NFL since 1960. In blue, we have completion percentage, which you can see has been steadily rising; last year, it reached an all-time high at 64.9%.  In orange, we have the percentage of completed passes that actually pick up a first down.  This number has been steadily declining, and in recent years, we can expect less than 55% of all completions to go for a first down.  When analysts say that the short pass has replaced the handoff, this is the evidence they are looking for: more completions per attempt, but fewer first downs per completion.

This is why I prefer to look at 1st down completions per pass play as a measure of success for a passing offense, rather than completion percentage. The graph below shows the 1st down per pass play rate for every season since 1960. To be sure, this is a number that’s also rising, but it’s not quite as subject to era-adjustments as the other stats. More importantly, you want to look at teams within a season, rather than just the league-wide numbers as a whole, which aren’t that interesting.

A good rule of thumb: if a team isn’t picking up first downs on one-third of their dropbacks, they probably are not a good passing offense.