The ump is blind!

That wasn’t a strike!

Yes, that’s what we often hear when we are at the ball park.   Especially if that ballpark is Citizen’s Bank Park.  Because Philly phans are extremely vociferous.

There are about 90 umpires that work for the major leagues, all of them male, with an average of 13 years of experience (average age 46).  Unless it’s the world series, each game has four umps- one for each of the bases. (There are 7 to adjudicate the championship games.) And, to prevent folks from trying to bribe an ump, game assignments are never announced until about 10 minutes (occasionally as long as 20) before game time.

Now, the job of the umpire is pretty tough.  (At least as hard as it is for a baseball player to decide when and how to swing at a pitch.)  In the course of about 43 milliseconds from when the ball is released by the pitcher to the time it crosses home plate, the umpire has to decide if that pitch was over the knees, under the shoulders, and directly over home plate ( a 17 inch square) and, therefore, a strike.  Otherwise, the pitch is a ball.

There’s even been some discussion about removing the home plate umpire and replacing him with those devices we see when we watch baseball on TV.  A camera that determines exactly where that pitch was when it crossed home plate.

Most traditionalists are against this sort of replacement.  (And, traditionalists are exactly who are the biggest baseball fans).

But, here’s some data that may change how we think about umpires.

A new study determine that it wasn’t the most experienced umps that are the accurate arbiters of balls and strikes.  It’s those with about 6 years on the job.  Mark Williams (Boston University, Questrom School of Business) and some of his master’s level grad students [specifically Tianyang Yang and Brandon Cohen, among other students] analyzed some 11 seasons of baseball data (4 million pitches).  Obviously, this is a data mining and analysis study.

According to Williams, it was the American League Division Series that caused him to consider such a study.  Some bad calls at first base during the Yankees- Red Sox game (specifically the third match-up) were overturned by replay reviews. But, since the most complete data exists for the situation at home plate, he opted to study the home plate ump and his calls.

The data was obtained from StatCast and PitchF/X.  Statcast is the system of camera and analytics that has been installed in every baseball stadium in 2015 to enable the sort of replay reviews that led to the overturned calls discussed above.   The PitchF/X system is much older, having been introduced in 2006; it’s a camera-based system that provides data about the trajectory, speed, spin, break, and location of each pitched ball.  (This is how that pitching map depicted above was produced.) The PitchF/X data is exactly what we see when we watch a ballgame on TV.  But, the system was really developed to better analyze the performance of batters and pitchers.

Williams was blown away with the data he obtained.  He expected the more experienced umps to be more accurate.  Instead, the data demonstrated  that the worst ten umpires over the 11 season study were those with more than 20 years of experience, with an average age of 56.  The best ten umps had only 6 years on-the-job experience with an average age of 38.  In other words, umps could be just like the ball players- they have a prime age after which their performance decreases dramatically.

But, given 4 million pieces of data, there were other findings.  Williams did find that bad calls decreased each year over the 11 years of the study.  Bad calls were made about 1/6 the time in 2008 (16.4%) falling down to 1 in 10 (9.2%) last year [2018].

That provides me to discern a completely different interpretation of these results.  Perhaps, since the younger umps learned their jobs (or is that honed their craft) as technology began allowing them to see more clearly where a ball crossed the plate, they could more clearly adhere to the proper strike zone definition.   Those umps that started earlier (and, therefore, were considered to be more experienced) developed their craft by determining their own strike zones- and within that self-determined strike zone (which may not be totally accurate), they ensured they provided precise, consistent results.

[Maybe this is a good time to re-acquaint you with the difference between precision and accuracy.   Precision is a description of random errors, while accuracy is the measurement of the difference between the result and the true value.  Oh, and the definition of the strike zone shown above in the diagram has never been tightly defined.]

(I should also let you know that Major League Baseball [MLB] takes great umbrage at Williams’ results.  Oh- and MLB stresses that calling balls and strikes is only a small portion of the umpire’s responsibility; it’s game management- the handling of the situations that develop and communication- that are the critical factors where an umpire is needed.)

In the meantime, you will probably be hearing me call for an optometrist to prescribe better glasses for those blind umps!

Go, Phils!