I support Kyle Lewis.
He was an absolute steal in the No. 11 overall selection in the 2016 Draft. Outside of Mercer University, Lewis did not have the pedigree of most of the top varsity rookies, but his dominance both at the University of Mercer. school and in the wood bat competition made him a solid choice with a bright future.
Then he blew his knee just over a month later, tearing his ACL and lateral and medial meniscus, significantly delaying what appeared to be a meteoric rise in the big leagues. He only reached the greats in 2019 and maintained his rookie status in the 2020 campaign shortened by Covid.
Last night, the Seattle Mariners outfielder was unanimously named American League Rookie of the Year, beating White Sox finalists Luis Robert and Astros Cristian Javier. I’m here to tell you that the voters got it wrong and Lewis is clearly the third choice among this group.
What did voters see? They allegedly take the numbers into consideration these days, and they’ve certainly looked at WAR (Wins Above Replacement). Fangraphs WAR ranked Lewis (1.7 WAR) and Robert (1.5) far ahead of Javier (0.5). Baseball Reference WAR was much closer: Robert (1.6 WAR) followed by Lewis (1.4) and Javier (1.3).
If you’ve ever read my work here, you know I see it a bit differently. What do the batt-based metrics tell us that these players “should have” played this season?
First, let’s look at the pitcher who made his way into this conversation with the two young center fielders.
Very quietly, Cristian Javier has had a hell of a year for an Astros club which found itself without Gerrit Cole (via free agency) and Justin Verlander (via injury). Thanks in part to Javier and his compatriot Framber Valdez, the Astros have made their way to the playoffs.
Valdez and Javier are two polar opposites. Valdez is a plumber with the lowest average allowable throw angle (1.1) degrees among MLB ERA qualifiers. Javier fell within six innings of qualifying for the ERA title and posted a significantly higher average launch angle (24.6 degrees) than the more qualified starter (Trevor Bauer, 20.7 degrees).
Javier’s low Fangraphs WAR total is due to his FIP of 4.94, which was well above his ERA of 3.48. Using granular data on batted balls, Javier “should have” posted an even better “Tru” ERA of 3.22. Fangraphs sees a flying ball launcher and penalizes him for being one. Javier allowed fairly weak contact with the flying ball (80 adjusted score for contact with the flying ball, on a scale where 100 equals the average). Oh, and his average allowed ground exit speed of 78.4 MPH was slower than any AL ERA qualifier. His adjusted contact score of 77 would have placed him 2nd in the AL Contact Manager of the Year race, had he pitched just six more innings. He deserves credit for this aspect of his performance.
His K / BB profile was average for the league, in fact a little better. His “Tru” ERA of 73 translates to 7.2 shots above average. Only 7 AL ERA qualifiers exceeded this mark; in fact, he just crushed his teammate Valdez (6.2), who finished tied for 8th.
Next, let’s take a look at the two center fielder, Lewis and Robert. They operated in tandem offensively throughout the season, starting strong and ending in deep collapses. Robert finished with a traditional hitting line of 0.233-0.302-0.436, and Lewis finished at 0.262-0.364-0.437. What do the batted ball metrics tell us?
Robert was about who his “real” numbers told us he was with the bat. He underperformed a bit on flying balls (148 unadjusted vs. 174 adjusted contact score) and Grounders (143 vs. 97) and slightly outperformed on liners (120 vs. 107). Adding the K’s and the BB’s, Robert’s projected line of .221-.291-.445 offers almost exactly the same value as his actual line. Overall, it’s worth (1.3) above average batting strokes, or 1.3 below average.
Lewis, on the other hand, was extremely lucky on in-play balls, especially on fly balls (151 unadjusted flyball contact score vs. adjusted 116) and Grounders (179 against 80). He sort of beat .286 on the Grounders despite an extreme tendency to shoot the Grounders. His liner (92.0 mph) and Grounder (82.7 mph) average authority were both significantly below the league average. There’s a tremendous amount of luck in Lewis’s ‘real’ 2020 numbers, thanks to some bloops and dribblers who landed hits – he ‘should have’ hit .222-.332-.394, or (0.3) hits above average.
To fairly compare Robert against Lewis, we need to take both base run and defense into consideration. Per Fangraphs, they both added 1.7 points net with their baserunning, but Robert has a huge advantage in defensive skill, from 3.4 to (1.0) points net. And you only need to see the two play once to see the difference – Robert is a true center fielder with corner outfield power, and Lewis is a corner outfielder with a pop in the middle of the field.
Count it, and Robert is recording 3.8 points above average, and Lewis is essentially a league average player at +0.4.
The numbers tell us Javier is the clear pick for Rookie of the Year. Now pitchers will sometimes take a little longer to reach their true level of talent compared to batters, so I could argue that Robert’s 3.8 rating is more “real” than Javier’s 7.2, but I could live with either of the players winning the material. Unfortunately, the writers went in another direction.
That said, I am happy to see Kyle Lewis healthy and established as an MLB contributor. Best wishes to him as the Mariners continue to rebuild.
Tomorrow we’ll take a similar look at the NL ROY race, won by Brewers reliever Devin Williams.