The Minnesota Twins, Zach Duke and Best Practices

When Duke transitioned into a reliever, he noticed that the majority of the bullpenners were chucking mid-90s heat. Duke never averaged a season over 90 miles per hour in his entire career. He needed something that would help him get hitters out, he needed deception. It was while with the Brewers, Duke began occasionally dropping his arm slot and alternating from a three-quarter and sidearm look.

Sometimes hitters got his normal look…

...And sometimes they got his drop-down slot.

It launched his best statistical season of his career. Duke’s ground ball rate shot up to near 60 percent and he struck out over 30 percent of hitters faced. That offseason, the Chicago White Sox signed him to a three-year, $15 million deal. Duke continued to post those solid numbers (with an unfortunate spike in home runs allowed with Chicago but, hey, it’s windy there) and he was ultimately traded to St. Louis where he tore his UCL and required Tommy John surgery in October 2016.

Written off for the 2017 season, Duke made a hard-charge comeback from the surgery and was back in the Cardinals’ bullpen by the end of July of this past year -- a little over ten months removed from his UCL rebuild. The 266 recovery days is the shortest turnaround in the Tommy John era.

If Duke is healthy -- and his medical tests by the Minnesota Twins prior to signing a one-year deal suggests that he is -- there is no reason to think that he cannot replicate a similar level of production as he provided the Brewers, White Sox and Cardinals.

That being said, there is a way Duke might actually be *better* going forward because of this one simple fact: He has been infinity better when throwing from the lower arm slot.

The numbers, courtesy of Statcast, between Duke’s regular three-quarter slot and his sidearm slot are stark. Since 2015, Statcast has tracked 2,351 pitches from him -- 1,249 have been thrown from the three-quarter slot (53%) and 1,023 (44%) were thrown from the sidearm slot. Now, take a look at the difference in results:

That’s huge. Everything from the lower arm slot is getting better results: More ground balls, more strikeouts, less solid contact. Imagine instead of saying “Three-Quarter” and “Sidearm” that chart said “Pitcher A” and “Pitcher B”. Wouldn’t you want to have “Pitcher B” on the mound for you? Given the data, it makes one wonder why Duke wouldn’t transition to the lower arm slot full-time.

Some would argue that Duke’s multiple and varied arm slots actually enhances his lower arm slot -- something like throwing a fastball from the three-quarter slot sets up his slider from down below. The logic is that with the variation, hitters will have to look for a wider release window for the pitch. However, hitters appear to be able to pick up the ball fairly effortlessly from the three-quarter slot so it is safe to assume that Duke’s lower arm slot is not enhancing his three-quarter slot (or, if it is, marginally so).

In many ways, asking Duke to favor the sidearm slot is akin to the development teams have had over the last few seasons of encouraging pitchers to lean heavily on their breaking balls. According to Fangraphs, Yankees pitchers threw a league-low 44% fastballs last season. Not far behind them were the Indians, Rays and Astros -- all forward-thinking, data-driven front offices. Hitters made more contact against fastballs. They were having more success against that pitch. Why not have your pitchers throw their best pitch more often?

Same thing should be asked about Duke. Why not focus on pitching the way which provided the most success?

Duke’s deception isn’t that he gives them different looks. His deception is releasing the ball from an angle that hitters don’t see that often. In 2017, there were 180 left-handed pitchers who released the ball within the similar vertical window as Duke did from three-quarter slot. There were just 43 who threw from the same vertical window as he did from the sidearm slot -- and most of those pitchers did it less than a handful of times. In the first group hitters saw 154,000 pitches come from that angle. From the latter, they saw just 6,200.

The more you dig into the numbers, the clearer it is that Duke should be throwing from the lower slot at a higher rate than he has been in the past. Save the three-quarter slot to mess with a hitter’s vision.

This Twins have beefed up their front office, including adding Josh Kalk (a pitching data pioneer in the PitchFX era) to the analytics team, and have plenty of sharp individuals who have reviewed this exact data and possibly came to the same conclusion. If the organization’s analytics arm actually identified this potential, ran it by the decision-makers, and they were able to secure a low-cost reliever who has a decent upside with a minor modification, that would be a super savvy maneuver never seen in these parts.

Either way, even if they don’t get Duke to reduce his reliance on the three-quarter arm slot, the Twins still have themselves a decent bullpen arm at a substantially low cost.