The nastiest pitches in baseball end with ... a sword

Even the great Mookie Betts, believe it or not, can be made to look uncomfortable on a baseball field.

Last season on Aug. 31, Braves ace Spencer Strider threw Betts a wicked 86 mph slider that dove into the ground halfway toward the plate. Betts tried his best to hold up, but he was too late. Even though he hardly took a full cut – the head of the bat on his follow-through reached only the height of his shoulders as he tried to stop the swing entirely – he went around for strike two.

The Atlanta broadcast called it a "weak swing.” On the Dodgers side, Orel Hershiser said that Betts had been "embarrassed," though in the context of preferring that outcome to putting a weakly hit ball into play. It was, essentially, the definition of a great pitcher fooling a great hitter.

Betts eventually got the last laugh, ending the plate appearance several pitches later with his 37th home run of the season. But for that moment, at least, Strider had thrown a pitch so nasty that he did more than collect a strike and advance the count in his favor. He had made one of the most elite hitters and best athletes in the sport look, for lack of a better term, bad.

There have been weak swings for decades, of course. Uncomfortable hacks. Awkward cuts. But in the baseball world of 2024, there’s a far more entertaining and descriptive way to talk about them. They are swords, as popularized by Rob Friedman (a.k.a., Pitching Ninja) since 2017. A reference to the quote "don't chop at it, it's not a sword" from the 2006 film “The Benchwarmers,” a sword is when a pitcher fools a hitter so badly that he forces a non-competitive swing, one where a batter either regrets his choice or can’t stop himself from taking a hack that looks so ugly it ends up going viral on social media.

Or, as Pete Alonso described it in 2021: "It's not a bad swing. It's an embarrassing swing."

It is, generally, a feel thing. It’s a sword when Friedman says it is, something like the Justice Potter test of pitching bad-assery, and for the last few years that’s all it’s been. Yet even the prolific Friedman can’t watch every pitch of every game, so we figured we’d offer some help, via the bat-tracking ability added to Statcast in 2023. (This is merely the tip of the iceberg, an entertaining teaser, as more robust metrics are built out for the upcoming season.) As this new data arrived, it became clear that figuring out what a sword was would be doable, and even more notably, fun.

If a sword is the ability to make hitters look foolish, then no one did that more often last season than Dylan Cease. That's thanks largely to his deadly slider, which had been baseball’s most valuable pitch in 2022 and which last year was responsible for 44 of these 55 swords – like this one when Christian Yelich’s 52 mph swing all but expresses the emotion of absolutely knowing when you’ve been defeated. (“That was .. a bit of a golf swing,” noted then-White Sox broadcaster Jason Benetti.)

How is a sword defined? It’s basically trying to put parameters on a feeling, trying to put borders around what is at its core just: Wow, did that batter look terrible there. (Or: Wow, did that pitcher throw such a nasty pitch that he made the batter look terrible there, as so much of Friedman’s work is pitching-positive.) That, then, is exactly what we did, with MLB data scientist Clay Nunnally leading the charge based on feedback from Friedman and others. In order to be a sword, a swing must:

• Be a swinging strike (no fouls)

• Cross the front face of home plate

• Be an “incomplete swing,” defined as a swing where the head of the bat crosses a line set at 5 inches ahead of the front of the plate but does not return back through it

• Have bat speed that is in the 10th percentile or lower for that player, to make clear that the swing was not fully unleashed, and no more than 20 mph in the final tracked frames, to avoid “full swings”

As we said: it’s putting data to a vibe, math to a mood, science to the art of making your opponent regret getting out of bed that morning.

So this? This is a sword.

So is this.

And this, against Juan Soto, of all people? That’s a sword.

If you really want to see what Statcast sees, the sword that Strider laid on Betts looks like this, thanks to a work-in-progress visual from MLB.com's Dana Bennett. The animation below makes it clear both how far away Betts' bat was from the ball, as well as exactly at what point he realizes he's doomed and attempts to stop the speed on his swing.

All told, in 2023 (including the postseason), there were over 6,200 swords, or about 2.5 per game. The Rangers (245) threw the most, led, perhaps surprisingly, by Jon Gray’s 29. (Here he is getting none other than Seattle superstar Julio Rodríguez.) The pitch-to-contact Cardinals had the fewest, at 152.

A sword doesn’t count for more than any other swinging strike. You won’t find it appearing with a more important run value than a non-sword strike. But it certainly slows a bat down – the average sword swing is 47 mph, well below the average competitive swing of 71 mph – and it might speak to nastiness, to deception, to the ability to make a professional hitter look like, well, this:

It has to mean something that Cease got a sword every 3.2 innings, while “less nasty” pitchers at the other end of the list got one every 20 innings (Miles Mikolas, Zack Greinke) or even once every 37 innings (Martín Pérez). There’s not one way to succeed, of course; very few would say that Cease had a better season than Seattle’s George Kirby, yet Kirby managed a sword merely once every 13.6 innings, or less than one per game.

It might not be surprising that the most swords come on nasty breaking balls from pitchers who tend to throw hard. Strider’s sword on Betts was one of 45 he got on his slider, the most on any pitch type in baseball, just ahead of Cease’s 44.

But Strider and Cease are both starters, so on a per-nine-inning basis, the most effective (minimum 50 innings) swordsman of the year was new Royal Will Smith, who had 3.9 per nine, just ahead of Matt Brash (3.7) and Trevor Richards (3.3), who does things like this:

On the batting side, well, we prefer to focus on the pitchers, because this is clearly a one-way street: What is good for the pitcher here has to be bad for the batter. But since you want to know, Yankees rookie Everson Pereira had the highest sword-per-plate-appearance rate (minimum 100 plate appearances), at 10.7, which befits a batter who had a .151/.233/.194 line.

That said, the five batters with the most swords were all productive or better – Christian Walker, Brent Rooker, Teoscar Hernández, Ryan Noda and Betts – which speaks a little to playing every day and a lot to how swing-and-miss doesn’t prevent great seasons by hitters.

While Pereira had 11 swords in 103 plate appearances, contact king Luis Arraez had just three in 574 times up, which shows you exactly how difficult – nearly impossible – it is to fool him. It’s the lowest rate of anyone with at least 500 plate appearances, and entertainingly, one of those was thrown by Shohei Ohtani, who is of course a two-way swordsman, causing 26 on the mound while being responsible for seven as a batter.

After Arraez came Tommy Edman, also with just three, though in 528 plate appearances, and then Vladimir Guerrero Jr., who even in the midst of a disappointing season was hard to fool this way, with just five in 682 times up.

J.D. Davis summed up a sword pretty well in the same SNY video as Alonso two years ago, when they were Mets teammates: "It's when someone makes a bad swing at a really good pitcher's pitch." Davis had a good year for the Giants in 2023, but he took 30 swords as a hitter.

One of those came in April, when he faced certain Hall of Famer Clayton Kershaw. The legendary lefty buried a slider that looked like a strike until it wasn't. Davis half-heartedly went -- a sword -- and all he could do was shake his head. He'd been had.

That's what a sword is, really. It's when a batter knows he got beat, at times before the pitch even gets to the plate. When a batter reacts like Davis did, it might not count for more in the box score, but it definitely counts for more. Just ask the players.