Ben-Porat: Using fly ball distance to find sleeper prospects

From Eli Ben-Porat at The Hardball Times on October 11, 2018:

One sunny, beautiful afternoon in April, while sitting at his computer a mere 25 kilometers north of Lake Ontario, this writer submitted a bid for Franmil Reyes in the Ottoneu FanGraphs Staff League. The bid went unopposed. Glossing over the abysmal performance of said author in said league, this bid, which remains as yet mediocre, was based on one stat: Franmil’s adjusted fly ball distance in Double-A at a young enough age to be meaningful.

I suppose, technically, it is actually one statistic and one biographical detail, namely the player’s age. To be even more specific, it’s a maximum age of 22 years. Why 22? We’ll look at the data in a bit and demonstrate why. It’s not a Deep LearningTM method. Rather, it’s a human look at the data, identifying the cluster that gives us a meaningful, predictive subset. The small group of athletes who graduate to the majors are all outliers who should exhibit outlier ability. In this case, the ability to combine launch angle and power to drive the ball consistently farther than their peers.

Adjusted Fly Ball Distance (FB Dist+) is a relative measure of fly ball distance hit, based on all fly balls hit for an out or a home run, with 100 being the average. An average FB Dist+ of 100 translates to roughly 275 feet. However, the important metric is the number relative to the average, rather than the raw number. The methodology is covered in greater detail in my first piece on the subject. You may want to skip my second piece, where I tried to convince you that Dylan Cozens is the next big thing. I’ve since tweaked the data in this piece to adjust for stadium bias in fly ball distance, which has improved the reliability of the metric. We’ll look at all levels of the minor leagues where we have reliable data, starting in Low-A and work our way up to Triple-A.

Read the full article here:

Originally published: October 11, 2018. Last Updated: October 11, 2018.