Derek Jeter recorded his final Yankee Stadium hit just over a week ago. It wasn’t just significant because it was Jeter’s final game at home. The opposite-field single also produced the game-winning run on a walk-off. Classic Jeter. Yankees and baseball fans alike joined in the celebration of a remarkable career culminating in the walk-off single. Conspiracy theorists had other ideas. @SmittyBarstool believes that Jeter’s at-bat was rigged because Evan Meek, the opposing pitcher, threw a low velocity fastball.
He uses a series of statistical maneuvers to back up this theory. He finds the average velocity of Meek’s fastball to be 91.457 MPH for the past 7 months. The standard deviation of his fastball velocity is 2.984 for the same time period. The pitch to Jeter registered at 86.1. Using those 3 numbers he finds a z-score to discern how many standard deviations away from the mean the 86.1 MPH pitch was. It was about 1.8 standard deviations from the mean which leaves about a .04% tail based on the standard normal curve. The math thus far is entirely correct. The next step is where the math goes awry.
Multiplying Jeter’s batting average for the season (.255) by the chance that a Meek pitch will be 86.1 MPH (.03593) does not tell us the percentage chance that Jeter recorded that hit naturally. Regression models are necessary for determining if Jeter’s hit was rigged.
Aside from not setting up a proper hypothesis test with an associated p-value to see if the result is statistically significant, @SmittyBarstool used inappropriate inputs for the fastball velocity. He uses Evan Meek’s season average fastball velocity from 2014. However, in the game on that Thursday, Meek threw 3 fastballs with velocities of 88.37, 88.29, and 86.82 according to brooksbaseball. His velocity in this game was down from his seasonal average of 91.5 that the author uses. 86.82 (or 86.1 that author uses) doesn’t look like a huge outlier when put together with the 2 low 88 readings from the same game.
The other input, Jeter’s 2014 season batting average (.255), is not a measure of his true talent for that at-bat. MGL has done extensive research on the topic of using projections rather than season-to-date numbers to glean the true talent level of a player for a given stat category. Using Jeter’s Steamer or ZiPS projection, would better measure the chance of him getting a hit at that point in time. Finally, for what it’s worth, the pitch wasn’t right down the middle and brooksbaseball didn’t classify it as a grooved pitch.
As a Sabermetrics subscriber, I love an argument supported with data that challenges the general consensus. The popular ideas of lineup protection, momentum, and closer-mentality have been debunked by thorough research and evidence. Unfortunately, the idea that Jeter’s walk-off single was rigged, seems like a specious argument at best. Perhaps most importantly: Evan Meek, a 31-year-old journeyman replacement-level reliever, is highly unlikely to groove a pitch. The guy has spent the better part of his career in the minor leagues and is fighting to maintain a job in a big league bullpen, not become the answer to a trivia question.