Shelby Miller Is Risen, and He’s Hungry for Outs

Shelby Miller is so old… (“How old is he?”)… he got R.A. Dickey to ground out in his first major league inning. He’s so old he threw more than 200 innings for the Braves when they were bad. He’s so old he threw more than 200 innings in a single season, full stop.

I guess 34 isn’t that old, but Miller has lived and died a hundred times during his career in professional baseball, and if the first eight appearances of his second go-around with the Diamondbacks are any indication (10 innings, 10 strikeouts, only four total baserunners), he’s back to life again.

There was a time, about 13 or 14 years ago, when Miller was the hot pitching prospect in baseball. In 2013, his age-22 season, Miller made 31 starts for the Cardinals, won 15 of them, and posted a 3.06 ERA. If he hadn’t come of age in a bizarrely loaded rookie class, he could well have gone home with hardware.

That 2013 NL rookie class merits a brief digression, because behind the race between José Fernández and Yasiel Puig for the top spot, there was Hyun Jin Ryu, Julio Teheran, Nolan Arenado, Gerrit Cole, Marcell Ozuna, Anthony Rendon, and Christian Yelich — and that’s just in the National League. The AL rookie class, by contrast, was highly forgettable. Wil Myers won Rookie of the Year, ahead of Jose Iglesias, and Dan Straily got a first-place vote. JB Shuck of all people got a second-place vote. The top six rookies that year, according to WAR, were in the NL.

Anyway, that’s as good as it got for the young Miller, who got traded for Jason Heyward at the end of 2014, then traded again to Arizona for a package that included Ender Inciarte and Dansby Swanson 13 months later. From there, he’s had close to a decade marred by elbow injuries, ineffectiveness, and pitch tinkering as he’s struggled to overcome those first two things.

Miller’s only start since 2019 was a one-inning opener stint for the Dodgers in September 2023. He hasn’t thrown 60 or more innings in a season since 2016, and he’s posted an ERA under 4.00 just once and an ERA under 6.00 just three times since 2015, regardless of workload.

Since 2018, Miller has been granted free agency or released 10 times by nine different organizations, but things have been on a gradual upswing over the past three years. Miller was outstanding in his aforementioned Dodgers stint, with a 1.71 ERA in 42 innings across 36 appearances. But his surface numbers weren’t backed up by the peripherals, which included an 11.7% walk rate, a .165 BABIP, and a strand rate of almost 90%.

The Tigers took a look in 2024, and Miller stayed healthy all year, apart from a three-week IL stint for ulnar nerve irritation starting in mid-May. But despite logging his highest innings total since 2016 (55 2/3) and posting a 4.53 ERA for a team that leaned heavily on its bullpen heading into a surprise playoff run, Miller was waived and released in the last week of the regular season.

This is now Year Nine of the Shelby Miller Reconstruction Project. Honestly, if you came into this season still believing in Miller, that’s not optimism or waiting for a post-hype breakout. If you believe in something without hard evidence for nine years, that’s basically a religion.

So is this 10-inning scoreless streak a miracle, or is there a scientific explanation?

Let’s tick off the low-hanging fruit: Miller’s throwing harder. Not by a ton; his fastball is averaging 94.4 mph instead of 93.4, but it’s noticeable. This time last year, Miller’s fastball was in the 92s, which simply does not cut it for a right-handed single-inning reliever.

More interesting: Miller has changed his breaking ball from a slider to a sweeper. The old slider sat in the high 80s, sometimes bumping into the low 90s, and while opponents didn’t exactly tee off on it — they posted a .285 wOBA against Miller’s slider last year — it had pretty lackluster movement, especially in the vertical plane.

Into the trash that goes. I think. Baseball Savant has Miller down for nine sweepers and two sliders this regular season, but I’m all but certain that at least one of those sliders was a classification error. Miller has thrown 184 breaking pitches since the start of the 2024 regular season; nine of the 10 slowest pitches, and eight of the nine highest spin rates, came on this year’s sweeper. The one interloper, in both cases, is a pitch that got classified as a slider on April 20 of this year.

The sweeper, in addition to spinning faster, moves horizontally instead of vertically; basically, Miller traded about six miles an hour for 11 inches of horizontal break.

That’s forced him to use his breaking ball differently. When Miller was throwing a hard slider with more up-and-down movement, he could throw it to batters on either side of the plate, and for strikes. This grander breaking ball is more of a chase pitch.

Crucially, it’s only a chase pitch against same-handed batters.

Here’s the one breaking ball Miller’s thrown to a lefty this year. It’s the one slider I don’t think was a coding error.

I don’t know how lefties didn’t tee off on that pitch last year; it looks like a total cement mixer. More to the point, this hard slider gave Miller almost no contrast to his fastball and splitter. All 850 pitches Miller threw in 2024 were contained within a velocity band of just 12.8 mph; 845 of those came in between 84.5 mph and 96.0 mph — a range of just 11.5 mph. Miller threw 12 sliders that were faster than his slowest four-seamer of the year, and more than 100 that were slower than his fastest splitter.

It’s just not really clear to me what Miller’s slider was doing for him last year. Now, instead of throwing that — and his splitter — for strikes, Miller is working his secondaries low and out of the zone. And now that he’s throwing pitches outside the strike zone more, batters are obliging him by chasing. Miller’s chase rate is fifth among qualified relievers as of this writing.

What stands out about all this to me is how basic all these changes are. Spin your breaking ball harder, change speeds, make hitters chase, throw an off-speed pitch to opposite-handed hitters and a breaking ball to same-handed ones but don’t cross the streams. All of this is stuff I learned by reading a children’s biography of Greg Maddux when I was 10, but it’s all stuff Miller either wasn’t doing — or wasn’t doing well — in 2024.

Now, he’s striking out more batters, allowing fewer baserunners, producing weaker contact. Even if you expect some regression, he’s a high-leverage-type reliever now. It’s a miracle. Hallelujah.