Arizona rookie Kyler Murray has some pretty strong counting stats so far this year: he’s averaging 276 passing yards per game, which would lead him to breaking the mark for passing yards by a rookie if he keeps this up for 16 games. He “only” has three interceptions, and he is completing a whopping 28 passes per game: only Drew Brees (5x), Peyton Manning (2x), and Ben Roethlisberger (2018) have ever done that in a season while playing in at least 8 games.

But Murray’s efficiency numbers have been, well, bad. He’s averaging just 4.42 ANY/A this year, thanks to an awful sack rate (10.5%) and a poor yards per completion average (9.9). But we don’t need an advanced stat like ANY/A to see Murray’s struggles: the most basic of stats will suffice.

Murray has thrown for only 37 first downs so far this year, despite having 153 passing plays (pass attempts plus sacks). That is an ugly 24.2% rate, while the league average is 32.6%. First down rate is one of the most basic stats to measure passing success, but it’s much better than completion percentage and is a handy way to quickly and easily check in on quarterback performance. The table below shows the passing stats for quarterbacks prior to today: [continue reading…]

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. [continue reading…]

Completion percentages continue to skyrocket, and part of the reason for the increase is the rise of low value completions. As a result, the best way to measure the consistency of a passing game isn’t with completion percentage, but with passing first down percentage. That’s defined as simply:

Passing First Downs / (Pass Attempts + Sacks)

As noted last week, the Giants and Texans provided good real world examples for 2018. Through four weeks, Houston ranked 25th in completion percentage but 8th in passing first down percentage. Meanwhile, the Giants ranked 2nd in completion percentage but 22nd in passing first down percentage. My proposition is that passing first down percentage is a better reflection of the passing game and a more meaningful statistic than completion percentage; and that’s why, through four weeks, Houston ranked 15th in points/game, while the Giants ranked 29th.

But does that argument hold up over time? How do we *know* that Passing 1st Down Rate is more important than Completion Percentage? Here’s a simply study I did.

1) Calculate the completion percentage relative to league average for every team since 1970.

2) Calculate the passing first down percentage relative to league average for every team since 1970.

3) Identify teams that ranked two percentage points better than league average in one metric and two percentage points worse than league average in the other metric.

4) Compare those teams. [continue reading…]

Eli Manning ranks 2nd in the NFL in completion rate with a remarkable 74.2% average. Houston’s Deshaun Watson, on the other hand, ranks 27th out of 34 qualifying passers with a 62.2% completion percentage. But that’s a bit misleading, because as I’ve written about before, first down percentage is a much more important metric than completion percentage.

For Watson, 62 of his 92 completed passes have gone for a first down, meaning just 33% of his completed passes didn’t pick up a first down. For Manning, just 53 of his 112 completions have gone for a first down, which means a whopping 53% of his completions didn’t pick up a first down.

Completion percentage is supposed to measure how consistent a quarterback is at making positive plays. A better formula than (completed passes/pass attempts) is (first downs gained via the pass / [pass attempts + sacks] ). In other words, the numerator should be passing first downs, not completed passes, and the denominator should include sacks. [continue reading…]