Home > News > No, there’s not a significant Taylor Swift effect on Travis Kelce’s stats
312 views 3 min 0 Comment

No, there’s not a significant Taylor Swift effect on Travis Kelce’s stats

Beware of small sample sizes in media narratives.

- October 31, 2023
Travis Kelce and Taylor Swift
Left: Travis Kelce (cc) All-Pro Reels; RIght: Taylor Swift (cc) Paolo V

“Travis Kelce’s stats are significantly better with Taylor Swift in attendance,” proclaimed the New York Post’s Oct. 22 headline immediately after the all-pro Kansas City tight end had a monster game (12 receptions for 179 yards) in front of her en route to the Chiefs’ 31-17 victory over the Los Angeles Chargers.    

The narrative picked up even more steam on Sunday when the pop superstar was absent for the Chiefs’ surprise loss to the last-place Broncos. Kelce had just 58 receiving yards against Denver, and a CBS graphic went viral showing that he’s now averaging over 50 yards more in the games Swift attends.

While Kelce’s production has clearly been better in the four games that Taylor Swift has attended this season than it was in the three games she did not (averaging 108 vs. 50.3 yards per game, respectively), that difference in means does not reach conventional levels of statistical significance (p = .153). Or more simply put, the so-called Swift effect could merely be the result of random variation in Kelce’s performance across games.

A standard statistical analysis of Kelce’s 2023 stats, for example, is riddled with uncertainty about the Swift effect’s actual size. Indeed, a simple difference in means test suggests that Kelce performs anywhere between 30.3 yards worse to 145.7 yards better per game with Swift in attendance than he does without her in the stadium.

The massive 95% confidence interval for the size of the Swift effect grows much larger, too, after accounting for the fact that three of the four games she attended were in Kansas City; Kelce is also averaging far more receiving yards per game this season at home than on the road (124 vs. 52.8 yards respectively). That uncertainty serves as another important reminder of how difficult it is to make meaningful statistical inferences from small sample sizes.

It will take more than seven games, then, to confidently conclude that Kelce performs significantly better with Swift in the audience. But that caveat is unlikely to get in the way of a splashy narrative about the enormous impact that Taylor Swift’s presence ostensibly has on her boyfriend’s performance.