The Most Unlikely Home Run

It seems like a simple question to ask. Which recent home run was the least likely?

You could flippantly answer — the one Erick Aybar hit this year, or the one Melky Cabrera hit this year — and because they’ve got the lowest isolated slugging percentages with at least one homer hit, you would be right. But that doesn’t control for the quality of the pitcher. Aybar hit his off of Rick Porcello, who is having some issues with the home run right now.

A slightly more sophisticated approach might have you scan down the list of the worst isolated powers in the game right now, and then cross-reference those names with the pitchers that allowed those home runs. If you do that, you’ll eventually settle on Alexei Ramirez, who hit his first homer of the year off of Johnny Cueto earlier this year.

Eh. I mean, that’s down broadstreet, and though it’s nice velocity, Johnny Cueto’s four-seamer is not his best asset. And really, home run suppression is not his best foot forward. He’s given up almost a homer per game, eventually Alexei Ramirez was going to run into one.

Let’s instead focus on the pitch itself first. There are 49 pitches that have been thrown 300 times since last year and haven’t been taken yard. At the top of the list is Wade Davis, who has thrown 952 four-seamers without giving up a home run the last two years. David Phelps has thrown 768 sliders and not one has left the building. Andrew Miller has thrown 653 sliders without losing the battle on that level. All very impressive.

But nobody has taken those pitches yard, yet, so let’s focus instead on the pitches that have. If you look at pitches that have been thrown at least 300 times and have given up at least one homer, and then convert the number of homers those pitches have allowed into a percentage, and then sort that leaderboard, you get these pitches:

Start at the top of the list with… Jake Diekman. Hey, it’s a good sinker. Above-average whiffs, and though the ground-ball rate is average, he hasn’t given up many homers on it. Well, the one he did give up was to… Joey Votto. Not so unlikely, maybe. Lance Lynn? Gave up one stinking homer, but it was to Mark Trumbo, who has a little power. Zach Walters hit one off Jake Petricka, and that’s a flawed batter whose carrying tool is power, and a great pitch off a flawed pitcher. Carlos Beltran took Jesse Hahn’s sinker deep. Jonathan Schoop took that awesome Dellin Betances knuckle curve deep, but Schoop has some power.

Tyson Ross hasn’t allowed many home runs off his four-seamer. He doesn’t throw it as much as his sinker, but he throws it enough. One home run, he allowed. It was to David Murphy and his .159 ISO, in Cleveland last year. That’s the best combo of low ISO and homer-stingy pitch, but does it tickle our itch?

Deeper down the list, Jon Lester’s curveball catches your eye. One homer allowed, and it was Jeff Baker that hit it. But, man, Baker has some power (.169 career ISO), and actually he likes curve balls (his second best pitch type values come from curves), and he’s right-handed, so he enjoyed the platoon advantage. Not that crazy.

Scooter Gennett homered off of Edinson Volquez’s knuckle curve, and was alone, but Volquez has given up homers to Jhonny Peralta and Colby Rasmus off pitches classified as curve balls, so maybe not so impressive. Lonnie Chisenhall hit a homer off of a 95.7 mph four-seam from Kelvin Herrera. Jed Lowrie hit a curve ball from Felix Hernandez that left his hand at 79.2 mph. These things seem to lack the singular nature we’re looking for.

Let’s open it up to the beginning of 2011. What are the stingiest pitches of the last five years? These five pitches only allowed one homer over the sample.

Clayton Kershaw (LHP), 5/11/14, Curve (.06% HR), 0-2 count, Brandon Hicks (RHB, career .156 ISO)
Tim Stauffer (RHP), 5/31/15, Sinker (.08% HR), 2-1 count, Josh Donaldson (RHB, career .199 ISO)
Jake Petricka (RHP), 8/26/14, Sinker (.09% HR), 3-2 count, Zach Walters (Switch, career .216 ISO)
Brandon Cumpton (RHP), 5/31/14, Sinker (.09% HR), 1-0 count, Hanley Ramirez (RHB, career .200 ISO)
Alex Wood (LHP), 4/29/14, Knuckle Curve (.1% HR), 1-0 count, Giancarlo Stanton (RHB, career .276 ISO)

Plenty has been written about that Kershaw curve ball to Hicks, and so far it’s the leader in the clubhouse. But the consensus is that Kershaw allowed another homer to Allen Craig on the curveball, once, and that the pitch was misclassified. (And now Kris Bryant has also homered off the curve, pushing Kershaw’s yellow hammer out of the top nine by this metric.)

If so, that seems to suggest that the Zach Walters home run was the least likely. But, with the door open for some two-homer pitches, a more unlikely combination of hitter and pitcher pops out like a sore thumb.

Andrew Miller’s slider.

Andrew Miller’s slider has been taken deep twice in the last five years. He’s thrown 1396 of them, so those homers represent .14% of all the sliders he’s thrown. Domonic Brown once took him deep in 2013, and that’s weird because Brown is a lefty. But that was also a great year for the Phillie, as he went deep 27 times and had a .222 ISO and life was good.

The other batter that once took Andrew Miller‘s slider deep was also a lefty, but one with less punch. This corner infielder has a career .129 ISO, which drops to .099 against lefties. Given the demonstrated likelihood of this lefty batter hitting a home run off a lefty, and Miller’s demonstrated likelihood of allowing a homer on a slider — both accrued in a combined sample of about six thousand pitches, though admittedly unregressed — there was a 1 in 50,000-pitch chance that this batter would take Andrew Miller’s slide piece deep.

And yet, on June 10th, 2013, on a 1-1 count, James Loney got the barrel on this Andrew Miller slider and took it yard. And that’s probably your least likely homer of the last five years.