Season Preview: Ryan Yarbrough

Given the chaos of the last season, it’s a little easy to have lost track of but Ryan Yarbrough no longer needs cover. The pitcher who made his name in a role without a name (“headliner?” “bulk guy?” “follower?”) has started 18 of his last 20 appearances and has a 4.02 ERA in the 20 appearances.

How did we get here?

It all started with his now infamous start in Seattle in August of 2019. Yarbrough had been semi-alternating traditional starts with “headliner” outings, five of his 14 appearances since May had come as traditional starts and the other nine a mix of “headliners” and shorter appearances.

For Seattle, a team with a pair of lefties in the top spots of the batting order, it made sense to have Yarbs get the traditional start. And boy did he make the most of it. The lefty cruised through the Mariners lineup three times over, even getting a chance to face and retire those first two lefties a fourth time each. He sat one out away from a complete game shutout. Of course, that’s when Kevin Cash sauntered to the mound and requested the ball from his starter, much to the shock of everyone in Safeco, who saw a pitcher who hadn’t allowed a base runner since the fifth, with no one on base, and just one measly out away from his first career shutout.

But that’s not how the Rays work. And their fans knew it, and Yarbrough knew it. He walked off the mound; Cash handed the ball to Emilio Pagan; and the Rays closer got the final out on two pitches, clinching the 1-0 win for Tampa Bay.

In many ways this moment was a precursor to Game 6 of the World Series, but in a somewhat more interesting sideplot, it was also the moment the Rays seemed to hand over full trust to Yarbrough.

As noted, the lefty has started 18 of his 20 appearances since in the more traditional manner, and he has thrived in that role. Now, this is still a pitcher who barely tops out at 90 mph with his fastball, and doesn’t have stuff that screams Best Rays Starter. But if I had to make a prediction for which Rays pitcher will lead the team in rWAR in 2021, I’d go with Yarbrough and feel good about it [this is Jim speaking; not Ian]. So how does he do it?

Soft Contact Inducer

There’s something about players at the extremes that just make them more fun to study. Take a look at his 2020 MLB percentile rankings from Baseball Savant:

Of the 106 pitchers to throw at least 150.0 IP since the start of 2019, only Tommy Milone and Kyle Hendricks have a softer average fastball velocity (per FanGraphs).

This is a pitcher who thrives on an ability to induce soft contact.

When Yarbrough broke on to the scene, it was easy to write off that ability to induce soft contact as a potential red flag for future regression. Over small sample sizes, extreme performances are rarely indicative of a repeatable extreme skill.

But yet there are pitchers who have shown a repeatable skill at inducing soft contact.

Here’s a chart of all the pitchers to qualify for the innings pitched count for at least three seasons from 2015-2019. Their exit velocity percentiles are laid out in color scale for quick glancing since the chart is a bit overwhelming. Basically, green is good, red is bad, and you’ll notice lots of pitchers staying in their lane in that regard. (This chart will be a nightmare on mobile, but we’ll spell out the takeaways below.)

Gerrit Cole has alternated strong and weak ability to limit average exit velocity over the past four years, and 2020 proved funky for a lot of the typical stalwarts (Jacob deGrom and Max Scherzer most notably), which isn't unexpected given the strange nature of last year.

But the aforementioned Hendricks has been at the top of the league every season since 2015. DeGrom, Scherzer, Zack Greinkie and Zach Wheeler have been good-to-great, and no worse than average for long periods of time.

Overall, the relationship between one-year average exit velocity and the next year’s average exit velocity is weak but positive. The chart below includes all pitchers to have pitched at least 100 innings in a given season, two seasons in a row, which buys us Yarbrough’s 2018 and 2019.

Data from FanGraphs

The only other pitcher to have limited hard contact as well as Yarbrough two times in a row pulled off this feat twice. Our data set doesn’t include 2020, but if it were to, Yarbrough would have joined CC Sabathia with another dot in the bottom left, as he lead baseball in average exit velocity once again.

And looking at the dataset more holistically, there is still some evidence that a displayed ability at giving up extremely soft contact is more predictive of the quality of future contact than is any other part of the set.

I made a very simple prediction by regressing Year 1 of the set against 64% of the Year 1 league average, and then calculated the root mean squared error (RMSE) of those predictions for the whole set (113 pitchers), for the subset of pitchers whose Year 1 exit velocity was at least a standard deviation below average (19 pitchers), and for the subset of pitchers whose Year 1 exit velocity was at least a standard deviation above average (20 pitchers).

The hard contact group had the RMSE, at 1.24, the entire group (and the middle group) had a RMSE of 1.20, while the soft contact group had the lowest error at 1.17.

Put plainly, there’s a lot of noise in average exit velocity for pitchers, but there’s some evidence that there’s more signal when a pitcher posts an extremely average low exit velocity than there is when he does literally anything else.

And when a pitcher, like Yarbrough, posts the lowest average EV over and over, like Sabathia, or Hendricks, or now Ryan Yarbrough, it’s time to pay attention and probably to believe.

What Difference Might That Make?

On the surface, allowing a lower average exit velocity is a good thing. However, not all average exit velo is created equal. A pitcher who could truly hold hitters to an 84.8 mph batted ball each and every time (Yarbrough’s career average exit velo) would have a massive advantage over your average pitcher. However, a pitcher who simply alternated 60 mph batted balls with 110 mph batted balls (which would come out to the same 85 mph exit velo number) would obviously not be nearly as valuable.

That being said, extremely low average exit velocities do seem to benefit pitchers. If we cycle back to the main comp we’ve made a couple times now, Kyle Hendricks, is a player with a career 3.76 xFIP but a career 3.12 ERA. That’s a pattern Yarbrough has followed his young season so far. Yarbs has a career 4.33 xFIP but a career 3.94 ERA.

Johnny Cueto is another famously “tricky” pitcher who was able to beat his xFIP throughout his career, with an ERA a half run below his xFIP throughout his career.

Concluding Thoughts on 2021

Why does this matter?

Well, most projections systems trust peripherals (that are expressed in metrics like FIP or xFIP) more than actual runs allowed, because for most pitchers stripping out the more noisy components of a pitchers performance is a better predictor of future success than is ERA. But Yarbrough isn’t like most pitchers.

I feel pretty comfortable saying that Yarbrough should beat his projected ERA for 2021 by around a half a run as long as he continues his ability to limit hard contact; which, at this point, seems like a likely proposition.

With that in mind, most projection systems have Yarbrough sitting around a 4.30, but I’d basically call the under on that a lock, and set the bar closer to 3.80. From there, the errors bars could certainly bring that number back up to around 4.30, but they could also bring him down into the 3.30 zone, and that’s before we even take into account potential growth from the fourth-year pro. Finally, given the aforementioned fact that Yarbs has managed to lower his average exit velocity allowed each and every season, I have incredibly high hopes (but also expectations) for Yarbs this season.

Count me in for a 3.45 ERA, 150 innings pitched, and the title of Best Rays Pitcher by the end of the season. (All this hype belongs to Jim, not Ian, just so you all know who to come after when Yarbs implodes halfway through the season.)