Statisticians of the baseball world are faced with a conundrum: Until recently, more and more home runs have cleared the fences, and nobody knows why. A report released in June showed the efforts of nine MLB-commissioned physicists to find the root of the home run problem, but the findings raised more questions than answers. Amid theories that the balls themselves or the mud rubbed on them are to blame, Jason Wilson Ph.D., an associate professor of mathematics at Biola University in Southern California, proposes that the answer lies in a new metric: the QOP.
The June MLB report showed that between 2015 and 2017 the number of home runs in a season had increased dramatically from 4,909 to a record-breaking 6,105. Wilson’s new analysis, which he’s presenting at the Joint Statistical Meeting of the American Statistical Association next week, rules out explanations for the home run spike like the “juiced ball” theory (a decrease in drag on the ball due to manufacturing) or ideas about the mineral content of baseball mud.
Pundits missed this, he posits, because they didn’t have the right statistics to notice or describe the change.
Wilson proposes a new baseball metric, called quality of pitch — QOP — to help them do so. “We used our QOP to measure the pitching quality in 2017 during the home run surge,” Wilson says. “And what we found was that the quality of pitching is down and you have this high spike in home runs. When you move to 2018, quality of pitch is up, and the home runs are down.”
QOP Measures What No Stat Has Measured Before
The statistics that we already use to measure the effectiveness of pitching generally fall into two buckets — defense-dependent statistics and defense-independent statistics — neither of which describes QOP. Because pitchers are the ultimate defensive player — no runs can be scored unless the ball gets thrown — statisticians are tasked with distinguishing which runs are scored because of defensive mishaps, which runs are simply unpreventable, and which ones really come down to bad pitching.
Metrics like walks, strikeouts, and home runs allowed are some basic examples of defense-independent baseball statistics. On the flip side, ERA (earned runs allowed) is an example of a statistic that’s defense-dependent. Categories aside, however, all of these metrics have one thing in common: They judge the quality of a pitch by what happens after it’s thrown. QOP, on the other hand, is all about how pitch travels through the air.
How QOP Solves the Home Run Mystery
QOP assigns a number between zero and ten for every pitch, calculated according to six key characteristics of the pitch defined by Wilson and his student, Jarvis Greiner, pitcher for Biola University’s baseball team. Two of those characteristics, they found, are tightly linked to pitches with low QOPs — that is, those that lead to home runs.
The six factors that determine QOP, which the team identified by watching videos of pitch after pitch in a lab, generally fall into three categories: speed, ball movement, and location.
Faster pitches are harder to hit, and so they earn a higher score. Pitches slightly outside the strike zone are also harder to hit, so placement around the corners of the plate increase a QOP score as well.
Vertical rise is a “tell that a pitch is a curveball,” docking points from a pitch’s score. Swerving to the side, in contrast, is called “horizontal break” and can help improve a QOP score.
Then come the two most important components: vertical break, which is how much the ball falls downward during its trajectory, and total vertical break, which is when during the trajectory the ball begins to fall.
It’s in these final two components that Wilson and Grenier found their answer to the 2017 home run mystery. Of the 700,000 pitches they analyzed, they noticed that there was a drop in the amount of vertical break on a pitch-by-pitch basis. Pitches were lingering higher in the strike zone, sitting ducks for hitters to slam into the outfield and beyond.
While this explanation is the first concrete answer we’ve had on this subject, Wilson is cautious to attribute the whole home run spike to this analysis. Neither it seems, are MLB teams who have yet to take Wilson up on his offer to run analysis for their organizations — although Wilson says that many teams are developing metrics like his in-house.
Baseball Fans Want More Baseball Stats
Home runs have actually been down this season, Wilson says, which might indicate that there’s been a change in the way the game is actually played. Perhaps pitching is simply getting better; with the flood of new pitching-linked statistics, we may soon be able to find out for sure.
QOP, it seems, isn’t the only new metric attempting to add granularity to pitch analysis. In fact, it’s one of many stats vying to be the leader of a new type of baseball-related statistical research called “pitch quantification.” A potential competitor is a metric proposed by a team at the University of Canada that uses the amount of bases conceded as a result of a single pitch to help identify which pitches are working and which ones aren’t.
“What the value that I feel we are adding is that, there’s a single number on every pitch,” Wilson says of the QOP. “But if you move down to a pitch to pitch level, that’s a level of granularity that typical baseball statistics don’t allow access to.”