Are The Red Sox Doomers Wrong?

Boston has been tough to predict... though other teams are tougher.

Clapping back at critics who’ve been predicting doom for the Red Sox ever since the team failed to re-sign Xander Bogaerts (among other cheap and/or weird offseason decisions), Boston president and CEO Sam Kennedy said his team was optimistic for 2023 — and questioned anyone predicting it would suffer another down year.

"As long as I've been around, [the predictions] are usually wrong. I don't put a lot of stock in them," Kennedy said. "We'll let the players do the talking."

Certainly the preseason forecasts are not especially bullish on Boston’s chances to be much more than a .500 team this season. According to my usual weighted cocktail of predictions from Clay Davenport, Baseball Prospectus, FanGraphs and Elo ratings, Boston is expected to win 81.9 games, which ranks fourth in the AL East and 16th in MLB overall.

Nor is there very much of a dispute between the systems about how good these Red Sox will be. The standard deviation of win forecasts between the component systems ranks 10th-lowest among all teams, meaning Boston’s composite prediction is actually among the most agreed-upon forecasts heading into the season.

But if Kennedy is right, such agreement might just mean all of the systems are wrong. And that argument isn’t without its merits if you consider recent history. As I’ve written about frequently, Boston has been the ultimate boom-or-bust team for more than a decade now, swinging wildly between good and bad (or bad and good) seasons like few clubs in baseball history. So any preseason prediction about how the Red Sox will do might carry the caveat that even Boston itself rarely seems to know the answer.

Still, does that mean the Red Sox are notably hard for forecast systems to predict, or that they’ve defied the projections more than anyone else? Not quite. Using the FiveThirtyEight preseason forecasts, I went back to 2016 and collected every team’s actual wins and predicted wins going into each season; then I calculated the root mean-squared error (RMSE) around each team’s win predictions. Over the course of those seven seasons, the Red Sox checked in with a RMSE of 10.3 — ranking them among the 10 hardest-to-predict teams in the league, but nowhere close to challenging the famously fickle San Francisco Giants for No. 1.

Get the data here.

So Kennedy is correct that the Red Sox haven’t exactly been the easiest team for statistical forecasts to peg in recent seasons… but they haven’t really been all that much tougher than the typical team. The truth is that John Sterling was right — you can’t predict baseball (at least not with a high degree of certainty) when it comes to just about any team. And while this could mean all those computers calling for a rough Red Sox season are just making an error, Boston’s brass didn’t exactly do itself any favors in boosting the odds this past offseason.