How Duffey Used a Slider and Knucklecurve to Cement a Spot in Twins Bullpen

Mandatory Credit: Kim Klement-USA TODAY Sports

A couple of offseasons ago, we ran a series similar to what we’re doing here, entitled “Minnesota Twins 40-Man Report Cards.” The only key difference to be aware of is since it’s happening now instead of during the offseason, we’re going to make it forward-looking — that is, with the current roster as constructed, looking back on their 2019 season.

Previous editions:

So in some cases, it’ll be looking at players who might not have necessarily spent all or even any of their 2019 season in the organization.

Let’s dive right in, starting with the pitchers:

  • Player: Tyler Duffey
  • 2019 team(s): Minnesota Twins
  • Pertinent Numbers: 57.2 IP, 2.50 ERA/3.06 FIP, 1.2 fWAR/0.9 bWAR (with Minnesota); 13.2 IP, 1.32 ERA/1.72 FIP (with Rochester)
2019 Lowdown

What a difference a year makes.

The Twins opened last season with a bullpen of Taylor Rogers, Trevor May, Blake Parker, Trevor Hildenberger, Ryne Harper and Adalberto Mejia (!).

That’s right — not even spring training hype man Fernando Romero made the Opening Day roster. Nor did Duffey, who posted a 7.04 ERA in 7.2 spring innings as he was working through some things.

It didn’t take long for Duffey to get the call. On April 16 (exactly one year ago, as luck would have it) he returned from Rochester — and a little over a week later he was sent back. Duffey spent three weeks with the Red Wings, was recalled in mid-May and never returned to the minors.

And since he’s now out of options, he won’t again. But that’s pretty much a moot point.

Shortly after Duffey’s last recall, he gave up three earned runs in low-leverage work on May 16 against the Mariners. From that point forward, Duffey had a 2.19 ERA, 71 strikeouts and 12 walks in 49.1 innings (13.0 K/9) and a slash-line against of .191/.251/.301.

It took Duffey a while to carve out a significant role in Rocco Baldelli‘s bullpen, however. Baseball-Reference has a statistic called “adjusted leverage index” and it uses a scale to determine how dicey a situation a reliever is brought into. In this case, 1.0 is considered average pressure; anything above is a more difficult situation and anything below is, well, you know, not.

Only one of Duffey’s first 12 outings had a higher-than-usual aLI. In that game, Baldelli used five relievers behind Kyle Gibson in a 4-3 win over the Los Angeles Angels. Duffey threw three pitches, gave up one hit and hit the showers.

Duffey was truly thrown into the deep end against the Detroit Tigers on June 7. With an aLI of 2.87, this actually wound up being the fourth-toughest spot he worked all season long.

It was self-inflicted damage, but Duffey successfully navigated it to preserve a 5-3 lead for his first hold of the season. After getting Miguel Cabrera to ground to short, Brandon Dixon and Ronny Rodriguez followed with singles and a two-out wild pitch put the tying run in scoring position. Duffey bore down and got Dawel Lugo to ground back to the mound, and the threat was effectively diffused.

Still, just two of Duffey’s first 15 outings came in high-leverage spots. It wasn’t until mid-June that he worked back-to-back big spots. It wasn’t until July 20-23 that he worked three of them in a row.

But starting with that three-game span — one in Oakland and two in the thrilling Yankees series — Duffey was absolutely incredible.


  • 1.35 ERA
  • 45-8 K/BB in 26.2 innings
  • 15.2 K/9 (!)
  • .143/.218/.209 line against
  • 19 percent swinging-strike rate

The upshot is this: After working just seven high-leverage spots in his first 28 outings (25 percent), 21 of his final 30 were (70 percent) in spots with an aLI of at least 1.00. Five of the nine that weren’t had an aLI of .75 or higher.

In other words, Baldelli read the room perfectly and Duffey responded in epic fashion.

And for a guy who came into the season with a career ERA of 5.46 in 287.0 innings, Duffey showed that pitching development isn’t linear and that there’s really rarely a reason to give up on someone with a live arm or a good pitch mix.

Speaking of pitch mix, the big board at Target Field and PITCHf/x both said Duffey threw a slider in 2019. It appears that has been amended as Fangraphs — which houses PITCHf/x data — no longer says this, but Duffey basically told Zone Coverage that it was the same grip as his knucklecurve, just thrown at a different speed to get different movement. The shape stays the same, but the break is shorter when it’s thrown harder.

Duffey has also taken to the concept of tunneling, or essentially throwing pitches through the same “tunnel” with movement happening late to deceive batters.

This is a good illustration of the concept:

For Duffey, that means throwing his fastball up in the zone and his knucklecurve through that tunnel where it breaks low in the zone — or out of it completely if the fastball tunnels a little lower in the zone. Duffey admitted in early April he’d heard the concept but never really had it explained to him, and while it wasn’t hard for him to buy into it, it took a little while for him to actually take to it.

But as Duffey said in that same discussion, the struggles were a necessary evil (from April 23 piece):

“For me, yes, because I was a sinkerball guy forever,” Duffey said of the adjustment this spring to working up in the zone. “I was down in the zone in the 4-spot with the fastball down and in — down and away to a righty — and I was leaving it in the bottom corner of the zone like (Kyle) Gibson. And now it was like ‘Hey, I gotta change my sights.’”

If Duffey can sustain his late-season success from 2019 going into whenever the 2020 season starts is anyone’s guess, but if he can merely repeat his 2019 on the whole, he’ll be a huge part of Baldelli’s bullpen moving forward.

And that’s a huge step from a year ago.

Jeremy Maschino‘s Take

While Duffey enjoyed a breakout season last year, he is primed to continue that success into this year due to the analytical statistics on his pitches.

Last season, Duffey threw four pitches: four-seam, slider, knucklecurve and sinker. But, as I look deeper into Duffey, it seems like he has all but dropped in his sinker as well as his slider and knucklecurve are actually just two different curveballs.

Duffey’s four-seam is his most frequently thrown pitch by far. Essentially every other pitch by Duffey is a four-seam. It’s easy to justify this though as he gathers a whiff rate (swing and misses divided by total swings) of 30.3 percent and a hard-hit rate (batted balls with an exit velocity of 95+ MPH divided by total batted balls in play) of only 28.6 percent.

But, those rates could be even better.

Duffey throws his four-seam on average at 93.9 MPH with a spin rate of 2,305 RPM, but only at a spin efficiency of 91.3 percent. This shows us the four-seam is a good pitch, but he still has the potential to increase it to a great pitch. If he could increase his spin efficiency to the high-90 percent range, he will see his four-seam generate more carry, leading to even more swings and misses in the top half of the strike zone.

As alluded to earlier, Duffey’s “slider” is more like a curveball, just thrown slightly harder than usual though.

The “slider” comes in at an average of 83.6 MPH, with a spin rate of 2,550 RPM, a spin rate efficiency of 65.9 percent and a tilt of 7:45. This leads to 12 total inches of spin induced movement, with most of the movement breaking downward. Most sliders from overhand throwers like Duffey will have a spin efficiency of 40 percent or lower, have tilts around 9:00 and generate 7.0 inches or less of total spin induced movement.

However, this is not a knock on Duffey’s “slider” at all. It gives him a much different pitch than what hitters are used to seeing. Because of that, he gets a whiff rate of 41.7 percent on it, making the “slider” a very effective pitch for him.

Duffey’s knucklecurve is a fairly standard curveball and is probably thrown identical to his “slider” but softer and with less intent. He throws it at an average of 75.9 MPH, an 18 MPH difference from his fastball and an 8.0 MPH difference from his “slider.” Because the velocity is lower — and his intent as well — we see that the spin rate drops from 2,550 RPM to 2,385 RPM.

However, we also see an increase in spin efficiency from 65.9 percent to 83.1 percent. For this to be a really successful curveball though, I would like to see him bump his spin efficiency to around 90 percent or higher.

This would allow him to differentiate his movement profile from his “slider” and it would create a pitch that would tunnel well with his four-seam. Given that his four-seam has a tilt of 12:45 and his knuckle curve has a tilt of 7:15, this creates a difference of around 180 degrees in terms of how the ball spins, creating a spin mirror.

Increasing his spin efficiency would only enhance the spin mirroring between the two and leading to increased effectiveness in both pitches.

Final 2019 Grade: After a slow start, Duffey was absolutely bonkers to end the season. I had a few semesters like that in college. Anyway, it’s a solid A- for The Doof.

Jeremy Maschino is an independent pitching analyst and researcher who specializes in data-driven baseball. Follow him on Twitter here.

Predicting the Next Six Months For the Twins
By Chris Schad - Oct 21, 2021
The Twins Can’t Let Taylor Rogers Become Another One That Got Away
By Lou Hennessy - Oct 20, 2021

How Do the Twins Stack Up In the AL Central Right Now?

Mandatory Credit: Kim Klement-USA TODAY Sports

The American League Central Division has become one of the worst divisions in baseball recently. Last year, only the Chicago White Sox finished with a winning record. […]

Continue Reading