March 30, 2023
By Brandon Cohen, Senior Analyst
The 2023 season is set to begin, and UmpScores is excited to welcome everyone back to baseball with an update to our product and analytics. This year, we have improved our scorecards, and are thrilled to announce that we have added in a brand new metric!
Sports are games of margins. The adage has always been that “football is a game of inches” but the same is true for baseball included. The sport’s history is filled with rule changes, measured in inches, that have profoundly affected the way the game is played. Pitching performance and run scoring were heavily impacted by a series of rule changes in the 1950s and 1960s to the definition of the strike zone and the height of the mound. In 1968, when the run scoring environment was at its lowest, baseball lowered the mound from 15 inches to 10 inches, where it stands today. Runs per game have never been that low since the rule change.
Even today, subtle rule changes continue to have an impact. This year, to promote more steals and fewer injuries, baseball expanded the size of a base from 15 inches to 18 inches. Coupled with pitch clock rules and limitations on pickoffs, analysts expect stolen bases to rise this year as more teams try to take advantage of the new rules introduced.
In the same context, an inch or two around the edge of the strike zone can be the difference between a true strike or a ball. In an interview full of hyperbole last year, former MLB umpire Joe West discussed the controversial performance of Angel Hernandez in an April game, claiming that MLB graded his performance at 96%. UmpScores calculated his BCR as 12.40% for the game, putting his accuracy at 87.60%, well below MLB’s metric. West questioned in the same interview how it is that independent auditors such as us can have metrics with such different scores. The answer is actually quite obvious.
MLB does not count pitches which are within a baseball’s diameter (just shy of 3 inches) of the plate in their metrics. Those pitches are considered too difficult to include. This piqued our interest. How many pitches fall within that zone, and why throw those out? Wouldn’t it actually be better to track those calls and see how umpires perform when pitches are difficult to assess? These are actual pitches with real consequences for the players and the game. To better measure these close calls, UmpScores has developed the cBCR metric.
cBCR, like BCR, is a ratio of incorrect calls to all calls made during the game. During an average game, of the almost 300 pitches thrown, about half are called by the home plate umpire. The difference is that instead of including every pitch, we throw out all the easy ones which are at least a baseball diameter inside or outside the zone. Those are softballs and we should expect any trained official to get those right.
This new metric is designed to eliminate much of the noise and measure umpire’s performance around the crucial close calls. BCR was created to be pure and interpretable, but finer detail can matter when considering a full evaluation of an umpire's ability to get calls correct at the plate. This is where cBCR steps in, measuring the real skill and training umpires have.
There are times when some well known, tried and true statistics are not necessarily the best metrics for evaluating player performance. ERA for relief pitchers, who pitch far fewer innings than starting pitchers, stands out as a metric that can be dramatically affected by a single poor performance. Instead, various metrics such as shutdowns, meltdowns, and WPA (win probability added) are often used to better assess reliever performance. We hope cBCR can help provide all of you with a context to the overall BCR an umpire have by showcasing whether the umpire was truly a master of the zone that day.
cBCR in 2022
Now that we have defined the close call zone and cBCR, the next logical step is to apply our new statistical measurements to last year’s data and analyze umpire performance. In 2022, umpires had a league-wide BCR of 7.29% on almost 379,283 calls made. But of those nearly 380,000 calls, 43.2% of all calls were within the close call range.
These results are surprising and raise important questions about MLB’s reporting of umpire performance. MLB appears to be excluding nearly half of the sample size of pitches. Even worse, 95.7% of all missed calls from 2022 fall into this zone. Now we know the answer to Joe West’s question as to how our metrics differ from those from the league office: MLB has been self-selecting its sample size to remove almost every error made, resulting in highly inflated scores. UmpScores excludes calls at a far lower rate in calculating BCR (mainly as a way of dealing with potential error from Hawkeye), resulting in much lower accuracy ratings.
When further dissecting this data and looking back at Hernandez’ game, umpires had an accuracy rating of 99.5% on pitches not in the close call zone. This represents the sample size that MLB uses in its audit of umpire performance. Suddenly, Hernandez’ 96% accuracy for the game looks awful. Even without the context of cBCR (which was 23.7%), that performance appears to be one of the worst called games of the year.
When we looked at the calls in and around the edge of the plate in 2022, the number of mistakes went up by more than double the rate: cBCR for the league was 15.3%. The range for individual umpires was also much greater, with umpire performance grading out between the mid 11% to mid 19% compared to 5% and 9% for BCR. This allows us to more easily differentiate the best performing umpires behind the plate from the worst.
The umpires with the best cBCR in 2022 were Jeremie Rehak (11.6%), Pat Hoberg (11.7%), and Alex Tosi (12.6%). All of those umpires were also tracked as top 10 performers by BCR last year, so it is no surprise that they would thrive when it comes to calling close pitches. Tosi in particular stands out, as he was a replacement ump who only recently received a well-deserved promotion to full time for the start of the 2023 season. The worst cBCR scores, making incorrect calls when it really counted, belonged to CB Bucknor (19.6%), Andy Fletcher (19.4%), and Jerry Layne (19.3%), all members of the bottom 5 performing umpires based on BCR highlighted in last seasons Final Call.
To provide early insight into this season’s home plate umpire performance, UmpScores has begun posting spring training game results. In these results we have also started to include the new cBCR rating. Additionally, based on user feedback, we have updated our cards so the furthest miss will now calculate missed pitches inside the zone too.
Finally, our app, first launched in fall 2019 is being updated and we anticipate launching it by mid May. As we have more news on this exciting launch, we will keep you updated.
Thanks for your important feedback and together we can make the game better.
We hope you are as excited for the 2023 MLB season as we are!