Sophomore Slumps Aren’t a Thing

Even in an age in which baseball – and most sports to an extent – has become an extremely data-driven enterprise, the stew of conventional wisdom, mythology, and storylines could still feed a pretty large family. That’s not to say that this is a bad thing; even an old, jaded stat nerd like me gets excited to enjoy such a stew from time to time. But at the end of the day, an analyst has to focus on what’s true and what is not, and very few bits of baseball orthodoxy are more persistent than that of the sophomore slump. Coined for underperforming second-year high school or college athletes, the meaning in baseball is roughly parallel it: After a successful rookie season, a player finds it difficult to maintain the performance from their debut and are weighed down by the greatly increased expectations. As an analyst, the inevitable follow-up question is whether the sophomore slump is actually real.

While I entered this article with some rather developed skepticism, there’s no denying that high-performing rookies do occasionally have pretty wretched follow-up campaigns. Every longtime baseball fan can probably rattle off a dozen or so names instantly after reading the title of the article. For me, visions of Joe Charboneau, Pat Listach, Mark Fidrych, Jerome Walton, and Chris Coghlan dance in my head. And the list goes on and on. However, a second-year skid doesn’t mean there’s a special effect that causes it. The fact of the matter is that you should expect a lot of regression toward the mean for any player in baseball who can be optioned freely to the minors. The way baseball’s minor league system works accentuates the selection bias; underperforming rookies are typically demoted while the ones crushing reasonable expectations get to stay.

Looking at the sophomore slumpers, the story is typically more complicated than the cautionary tale. ZiPS has minor league translations going back to 1950 at this point, and while Super Joe (Charboneau) hit very well in the season before his debut (.352/.422/.597 for Double-A Chattanooga), at 24, he wasn’t young for the level, and ZiPS takes enough air out of that line to drop his translated OPS below .800. ZiPS thought he’d be an OK lefty-masher, but not much more than that.

Charboneau had a solid offensive rookie season, winning the AL Rookie of the Year award, but in his case, the fates didn’t really give him a fair opportunity to repeat that season. He injured his back in spring training and played through the injury, as was the style of the time. Across a couple of stints in the majors after his rookie breakout, he combined to bat .210/.247/.362 over 147 at-bats, and he was never healthy or trusted enough to make good. He didn’t hit again in the minors, either, with the only exception a walk-heavy .791 OPS as a 29-year-old in A-Ball (!).

As quick as Charboneau’s fall from grace was, it was far from the biggest rookie WAR drop-off. Using the definition of rookie in our leaderboards, which doesn’t know about roster service time days but is suitable for the approach of identifying rookies rather than specific Rookie of the Year eligibility, here are the biggest sophomore slides by WAR since 1901.

Some of these players recovered to have solid major league careers and some of these slumps resulted from serious injury, such as Kerry Wood’s, but for some of the players, that was the end of the road for them in the big leagues. As for Super Joe, his skid was the 100th worst in history among hitters!

So, how do we extract a sophomore-slump effect from simple sophomore slumps? At this point, I’ve been running projections for two decades, so I have a decent-sized database of projections calculated contemporaneously (as opposed to backfilling before ZiPS existed). I certainly haven’t told ZiPS to give a special penalty to solid rookies having bad follow-up campaigns, so I went back and looked at the projections vs. realities for every hitter with a two-WAR rookie season and every pitcher who eclipsed 1.5 WAR. (Rookie pitchers tend to have more trouble grabbing playing time.) That gave me 166 hitters and 207 pitchers. Let’s start with the hitters.

The 26 players in the top bucket averaged 5.1 WAR in their rookie seasons and 3.7 WAR in their sophomore seasons. That’s a pretty significant drop-off, but they were projected for an even steeper decline. The next group — 44 players who accumulated 3-4 WAR as rookies — underperformed its projection by about two runs per player, while the 96 rookies who finished with 2-3 WAR slightly overperformed their projections, but it was very close. As for the entire sample of 166 hitters, ZiPS projected a decline from an average 3.1 WAR as rookies to 2.3 in their sophomore seasons. Their actual average in their second year was… 2.3 WAR. Let’s look at the pitchers.

This is the same story, with the decline for pitchers being about as predictable as it was for hitters: ZiPS underestimated their second-year WAR by about 0.08 wins on average.

That’s not the end of it, however. I wanted to see if ZiPS has projected a similar decline for players who were coming off their second through fifth seasons, because that would determine whether ZiPS was capturing a sophomore-slump effect or if this was just a more general regression to the mean for players with less major league experience.

In sum, ZiPS didn’t knock more performance off high-performing rookies than it did for sophomores, juniors, seniors, and guys who stayed a fifth year because they had to drop too many 8 a.m. classes that they slept through. That’s because the sophomore-slump effect doesn’t exist.

So yes, projections will likely project fewer WAR next season from this year’s standout rookies, such as Jackson Merrill, Jackson Chourio, and Masyn Winn. But that dip is likely to be the result of the typical regression toward the mean that any high performer with a limited track record is expected to experience.