A Different Way of Looking at Home Run Rate by Ben Clemens July 26, 2019 Recently, I’ve been pondering the strange way I think about HR/FB ratio. On one hand, it’s a way to explain away a hot or cold stretch from a hitter. When Joc Pederson got off to a blazing start this year, I looked at his HR/FB, a spicy 33.3% through the end of April, and told myself it was a small sample size phenomena. That’s the first way I use HR/FB for hitters — as a sanity check. At the same time, HR/FB is something we’ve all used to explain someone’s power. Joey Gallo is powerful, obviously. How do we know that? Well, he hits the ball really hard, which gets expressed by more of his fly balls turning into home runs. Gallo had a 47.6% HR/FB at the end of April, and even though I didn’t expect that to continue, I was willing to accept high numbers for Gallo’s HR/FB much more easily than I was for Pederson. This leaves HR/FB in a weird spot. It’s a number we use to see if players are getting lucky or unlucky relative to average, but it’s also a number we use to look for underlying skill. Problems arise when it’s unclear what is noise and what is signal. Is David Fletcher unlucky to have a 5.4% HR/FB? Surely not — he’s a contact hitter. Is Jose Ramirez unlucky to have a 6.5% HR/FB? I assume so, but I only assume so because he hit 39 home runs with a 16.9% HR/FB last year. What if last year was the outlier, not this one? Another way to think about this conundrum is that HR/FB contains an inherent contradiction we have to work around mentally. Putting fly balls as the bottom of the ratio implies that all fly balls are created equal, and that’s clearly untrue. Gallo is unloading on the ball, crushing many of the fly balls he hits into orbit. Fletcher, meanwhile, sports one of the lowest average exit velocities in the game. Even though a home run counts the same for each, the population of fly balls is tremendously different. How do we handle this contradiction? The glib answer is that it’s complicated but that our minds manage it. Aaron Judge has power, so a lot of his fly balls become home runs. We see his statistics and understand roughly how much sense they make. If Kolten Wong started hitting 35% of his fly balls for home runs for a few weeks, we’d intuitively grasp that there’s luck involved there. The human brain is a powerful thing. Luckily, though, we don’t have to rely solely on our brains being good at this difficult, subjective calculation. We have instrumentation that records how hard every batted ball is hit. Why confine ourselves to one mass of “fly balls” when we can differentiate, say, between a 110 mph laser beam and an 85 mph can of corn? There’s absolutely no question that exit velocity has a lot to say about how likely a fly ball (or line drive) is to become a home run. That’s true in a common-sense way, without even looking at the data. The data, though, lets us quantify it. Take a look at the percentage of balls hit between 10 and 50 degrees of launch angle (what Baseball Savant defines as line drives and fly balls) that become home runs at various exit velocities: That’s mostly unsurprising, though it’s interesting to see how much a few miles per hour can change a ball’s fate. Aside from a few sample-size hiccups at the top end of the scale, we see a monotonic increase, which makes good sense. Hitting the ball harder isn’t likely to produce fewer home runs. Before we do a little alchemy and produce expected HR/FB for players based on their exit velocity, we need to unify our definitions. What do I mean by that? Well, the “fly balls” in the standard HR/FB includes everything classified as a fly ball, including pop ups, but doesn’t include line drives. Our exit velocity chart, meanwhile, includes fly balls and line drives, but not pop ups. So, first things first — let’s make a new ratio. Instead of HR/FB, we’re going to look at HR/Potential HR — the number of home runs a player hit divided by their non-pop-up fly balls plus their line drives — in other words, the balls that could potentially become home runs. Until we see a home run hit with a launch angle higher than 50 degrees, there’s no reason to keep pop ups in the mix. Our new metric is going to be lower in general, since line drives are less likely to become home runs than fly balls, but it should still look roughly the same as overall HR/FB when it comes to the best and the brightest. The hardest hitters, after all, also hit line drives hard enough that they often become home runs. Let’s take a look at that list, minimum 50 line drives and fly balls: Top HR/Potential HR Rates, 2019 Name HR/Potential HR Joey Gallo 26.2% Miguel Sano 25.8% Mitch Garver 24.1% Christian Yelich 23.2% Peter Alonso 22.9% Hunter Renfroe 21.9% Franmil Reyes 21.8% Derek Dietrich 21.8% Eloy Jimenez 21.0% Willson Contreras 20.7% With that out of the way, it’s time for the trickier part. I took every line drive and fly ball hit this year and assigned it a home run probability based on exit velocity. Then, I worked out each player’s expected HR/FB based on those home run probabilities. If a player mostly hits the ball really hard, in other words, they get credit for it in the form of higher home run probabilities. If a player is hitting a ton of lazy fly balls, those debit their expected HR/Potential HR. Let’s take a look at this list, with the same 50 potential home run minimum: Top xHR/Potential HR Rates, 2019 Name xHR/FB% Using EV Joey Gallo 24.5% Aaron Judge 20.8% Nelson Cruz 19.7% Miguel Sano 19.4% Wil Myers 18.8% Josh Donaldson 18.8% Marcell Ozuna 18.5% Shohei Ohtani 18.5% Franmil Reyes 18.0% Jorge Alfaro 17.9% Now that’s a satisfying list. Want a stat that shows power? You can’t do better than having Gallo, Judge, and Nelson Cruz top the list. How about the other side, though — the hitters who we should expect to have minuscule home run totals? Lowest xHR/Potential HR Rates Name xHR/FB% Using EV Billy Hamilton 0.9% Dee Gordon 2.4% Nicky Lopez 2.6% Eric Sogard 2.7% David Fletcher 2.9% Tony Wolters 3.0% Jose Peraza 3.3% Luis Arraez 3.3% Hanser Alberto 3.4% Tony Kemp 3.4% Again, it looks like we’re onto something. Billy Hamilton’s strengths aren’t on the home run front. Dee Gordon and Tony Wolters aren’t threats to leave the park. It looks, for the most part, like our stat does what we want it to do, separating the wheat from the chaff when it comes to power. With an expected home run rate in hand, we can understand the two questions HR/FB rate wants to answer more clearly. Who are the most powerful batters? The ones with the highest xHR/Potential HR (yeah, that stat name is going to need some work). Joey Gallo isn’t just lucking into his home runs — he’s hitting the ball so hard that he deserves them. Billy Hamilton isn’t just getting unlucky — he’s not generating the kind of contact that leads to round-trippers. How about figuring out which batters are getting lucky, the guys who are punching a little above their weight? Well, we can compare their expected rate to their observed rate and look for outliers. Take Derek Dietrich, for instance. He’s become a power threat this year, mashing a career-high 19 home runs on a 26.8% HR/FB rate. His HR/Potential HR is a career high, as well — 21.8%, good for eighth-best in baseball. His expected HR/Potential HR, though, is a mere 12.5%, 155th out of 487 batters. He’s the biggest out-performer, in fact, in all of baseball. Here is the top 10: HR/Potential HR Outperformers Name xHR/FB% Using EV Actual HR% Gap Derek Dietrich 12.5% 21.8% 9.3% Mitch Garver 15.6% 24.1% 8.4% Mark Canha 12.8% 20.3% 7.5% Hunter Renfroe 15.0% 21.9% 6.9% Eloy Jimenez 14.3% 21.0% 6.7% Eugenio Suarez 12.3% 18.9% 6.6% Willson Contreras 14.1% 20.7% 6.6% Peter Alonso 16.4% 22.9% 6.5% Miguel Sano 19.4% 25.8% 6.4% Jordan Luplow 9.2% 15.4% 6.2% This doesn’t mean, to be clear, that all of these guys lack home run power. Miguel Sano appears high on this list, but he’s also fourth in expected home run rate. He’s both lucky and good. Christian Yelich is 12th on this list, and also 12th in expected home run rate. Just because you’re skillful doesn’t mean you can’t also benefit from good fortune. The other side of the ledger is more fun to look at — a who’s who of people who seem to be underperforming this year but have the contact quality to improve: HR/Potential HR Underperformers Name xHR/FB% Using EV Actual HR% Gap Kendrys Morales 12.6% 3.1% -9.5% Mike Zunino 15.6% 7.8% -7.8% Stephen Vogt 13.7% 6.2% -7.5% Luis Rengifo 10.5% 3.8% -6.7% Miguel Cabrera 10.2% 3.8% -6.4% Byron Buxton 13.4% 7.8% -5.6% Rafael Devers 17.2% 11.7% -5.5% Josh Harrison 6.8% 1.6% -5.2% Nicholas Castellanos 10.9% 6.0% -4.8% Austin Hedges 12.4% 7.7% -4.7% Comerica Park figures heavily on this list — Niko Goodrum just missed at 11th, and Nicholas Castellanos has already railed against the hilarious outfield dimensions. Reports of the demise of Mike Zunino’s power may have been exaggerated, a fact I swear I didn’t make up just to make Meg like this article more. Rafael Devers is already having a banner year, and he’s actually gotten a little unlucky! There are some excellent names who just missed, as well — Bryce Harper, Mookie Betts, and Lorenzo Cain can all feel jinxed to not have more dingers. Is this statistic perfect? Most definitely not. It doesn’t take launch angle into account, for example, and a 25 degree ball is far more likely to become a home run than a 15 degree launch angle ball. I left that out for one main reason. Adding as much information as we possibly can is nice, but it can lead to unwanted complexity. Are the buckets smooth? How do we interpolate likely home runs across a huge spectrum of launch angles and exit velocities? We certainly won’t have enough data to create an outcome matrix without doing a lot of smoothing. I see reasons you might want the more complete stat — a more xwOBA-style accounting of everything, in other words. Still, though, I think it’s pretty useful to consider exit velocity alone. Simply adding this one component adds a ton of information, without delving too far into opacity. Maybe more can be done with this, but for now, I’m content. Now, Mike Zunino, on the other hand — it sure looks like he has a right to be annoyed.