Impact of Age on NFL Player Performance: Efficiency Stats (Part 5)

Here’s a riddle: Player 1 leads the NFL in passing one season with over 5000 yards. Yet the very next year, his team gets rid of him and replaces him with Player 2. Player 2 goes on to throw for almost 500 fewer yards, yet no one would deny he was an upgrade. Why is this? Bonus points if you know the situation I’m referring to.

Left photo (Winston) by Julio Aguilar/Getty Images, right photo (Brady) by Nathan Ray Seebeck-USA TODAY Sports

Here’s the answer: the 2020 Tampa Bay Buccaneers replaced Jameis Winston with Tom Brady. Winston lead the NFL in passing yards in 2019 with 5109. So why’d the Bucs move on? There were probably two reasons:

• Almost any team would replace their starting Quarterback with Brady, a future Hall of Famer who many regard as the greatest football player of all time.
• Winston’s 2019 season wasn’t actually as good as his passing yards would indicate.

That season, Winston became the first player since 1988 to throw 30 interceptions. The significance of this cannot be understated. The 2019 Bucs were 6th out of 32 teams in yards per play allowed, and 1st in the league in rushing defense. Yet, they gave up the 4th most points in the league, mostly as a result of the interceptions Winston was throwing. Because the Bucs were allowing so many points, they were throwing the ball a ton: Winston lead the NFL in passing attempts with 626. In a strange way, Winston’s interceptions were allowing him to throw for more yards (and even more interceptions). The Bucs had a talented team (despite Winston’s play, they finished a somewhat respectable 7–9), but were trapped in this endless cycle. If they could simply find a Quarterback who wasn’t throwing an historic number of interceptions, they’d be pretty good. In Brady, they got this and more. Not only did the 2020 Bucs improve by 4 games to finish 11–5, they went on to win the Super Bowl.

The Winston/Brady Bucs are an extreme example of why counting stats (such as yards) can be misleading. A player can throw (or run or catch) for a lot of yards simply because they have lots of attempts. This creates an issue for evaluating player performance (and ultimately aging). That’s why efficiency statistics are preferred by the sports analytics community. Efficiency stats simply take a counting stat (like yards) and divide by attempts. This looks different for each position:¹

Quarterbacks

There are four basic counting stats for Quarterbacks: Yards, Completions, Touchdowns, and Interceptions. All of these can be converted into efficiency metrics when divided by passing attempts.²

TD% by age for Hall of Fame QB Peyton Manning. Manning’s efficiency was fairly consistent besides a sharp spike at age 28 (when he broke the single season pass TD record with 48), and a steep drop-off at age 39.

Running Backs

The three basic counting stats for Running Backs are Yards, Touchdowns, and Fumbles. Again, each can be converted into efficiency stats when divided by carries.

Yards per carry for Emmitt Smith, the leading rusher in NFL history. Smith’s most efficient season came at age 24, after which his performance declined.

Receivers

The three basic counting stats for Receivers are catches, yards, and touchdowns. To properly determine how efficient a Receiver is, we need to divide each of these numbers by targets.³ One thing to note: targets were not officially counted until 1992, so seasons before this will not be analyzed.⁴

Catch% numbers for Terrell Owens, the third leading receiver in NFL history. As he got older, Owens caught fewer of the passes thrown his way.

Kickers

Kickers only have two basic counting stats: makes and misses. However, there are many ways to categorize these two basic numbers. First, there is the difference between field goals and extra points. Second, there are many different lengths of field goals. Typically the longer the attempt, the more difficult a field goal is to make. Because of this, I split field goals into five categories: less than 20 yards, 20–29 yards, 30–39 yards, 40–49 yards, and 50+ yards. I also kept each Kicker’s combined Field Goal %. I thought it would be interesting to analyze the effect of age on each of these stats. For instance, it’s possible that as Kickers grow older, their leg strength decreases and it becomes harder for them to kick 50 yard field goals. But perhaps this doesn’t affect their ability to kick extra points. This is one of the hypotheses I will be testing in the coming weeks.

Adam Vinatieri is the leading scorer in NFL history. He remained fairly consistent from age 24 until age 46, before experiencing a steep decline at age 47.

One more quirk in the data must be discussed before predictions are able to be made. In Part 6, I will discuss inflation: the overall trends in NFL statistics from 1978 to today.

Footnotes

1: Code for the graphs in this article can be found here.

2: Some people like to combine these metrics into one number: Passer Rating. My footnote explaining why I didn’t use it was too long, so I wrote a short article instead.

3: Targets are the number of times a pass was thrown to a certain receiver.

4: This change will cause a smaller sample size, making my Receiver predictions less reliable. For this reason, I thought of dropping Receivers from my analysis altogether, but decided they would still be interesting to analyze.