Nobody knows how the 2023 baseball season will turn out. Nobody.
The rosters for all 30 clubs were in flux as the calendar flipped into the new year, and they’re certain to remain unstable all spring. Free agents will keep hopping from team to team as January, February, and March drag on. Trading activity will continue unabated as the weather gets warmer. Phenoms will pop up in spring-training games in Florida and Arizona. Marginal players will be waived, and equally marginal replacements will be signed.
It’s a fool’s errand to predict 2023’s standings in the midst of such uncertainty, yet a plethora of preseason magazines, sports websites, columnists, and beat writers will issue their fearless forecasts prior to opening day. And I’m joining the parade.
Subscribe — free — to Baseball’s Best (and Worst)
A new installment will arrive in your email each Tuesday and Friday morning
I’ve developed a computer program to compare each team’s record during the past three seasons (2020-2022) against the corresponding marks for 1,258 clubs that played during the era of free agency, which began in 1976. (A couple of notes: Expansion teams obviously couldn’t be added to this mix until they played three seasons. And I cut off the comparisons after 2020, so that all of the clubs involved would have at least two years of future data.)
My aim was to find parallels from history that might offer hints about the future, to gain insights into each present-day team by studying what eventually happened to its close matches from the past.
The 2020-2022 team scores for a current club were compared to the corresponding scores for every team in the comparison pool. (Click here for a quick refresher on team scores.) The absolute differences between each set of scores were added on a 3-2-1 basis, putting the strongest emphasis on the most recent pair. The smaller the sum, the greater the similarities between the two clubs.
I’ll spare you most of the math, but here’s a quick look at how the formula works.
My first example pits the 2020-2022 Milwaukee Brewers against the 1987-1989 Boston Red Sox. Their team scores were remarkably similar, especially in their most recent years: 51.203 for the 2022 Brewers, 51.960 for the 1989 Red Sox.
The absolute difference between those two teams was 0.757 points. That was the most recent pair in this example, so it was tripled to 2.271. Then we stepped back a year to compare the 2021 Brewers and the 1988 Red Sox, just 1.321 points apart, doubled to 2.642. Then to the 2020 Brewers and the 1987 Red Sox, just 0.783 from each other. The sum of these absolute differences was 5.696, a very close match.
Cooperstown at the Crossroads
Read about the Hall of Fame’s checkered history (and uncertain future)
Here’s another example, this time showing a wide gap between a pair of clubs. The 2020-2022 Brewers are involved again, this time with the 2003-2005 Kansas City Royals. Their sum of absolute differences was 212.790 points, an enormous gap that tells us these versions of the Brewers and Royals had virtually nothing in common. It proved to be one of the eight worst matches for Milwaukee in the entire pool of 1,258 comparisons.
But our initial example, involving the Brewers and Red Sox, had a much different outcome. It emerged as Milwaukee’s very best match of all. No other club from the period of free agency came closer to the 2020-2022 Brewers than the Red Sox from 1987 to 1989.
I’ve compiled a list of the 50 closest matches (CMs) for each current club, culling the top 4 percent of all 1,258 candidates, the 4 percent with the smallest sums of absolute differences.
I then examined the subsequent performance of each of the 50 CMs. We’ve already evaluated the team scores for the 1987-1989 Red Sox, for example, but how did Boston perform in its next season, 1990? The same question was repeated for all 50 look-alikes.
And that’s how my predictions for 2023 were generated. The next-year records for each current club’s 50 closest matches were tabulated, generating separate breakdowns of playoff qualifiers, league champions, and World Series winners.
The Red Sox, by the way, went 88-74 in 1990 and made it to the playoffs, which is a positive sign for Milwaukee’s 2023 prospects. But the second-best match for the 2020-2022 Brewers — the 1998-2000 Cincinnati Reds — stumbled to a 66-96 mark in 2001, giving cause for concern.
Thirty-three of Milwaukee’s 50 closest matches played better than .500 ball the following season. Twenty-two made the playoffs, four won league championships (the Orioles in 1979, Red Sox in 1986, Giants in 1989, and Astros in 2017), and one (the Astros) won a World Series. Those are good omens for the Brewers, but their fans should keep in mind that the 2001 Reds, 2009 Diamondbacks (70-92), 2009 Indians (65-97), and 2017 Mets (70-92) also fit Milwaukee’s template for 2023.
You can find my complete predictions for the coming season in my new book, Baseball’s Best (and Worst) 2023 Yearbook. I’ll also touch on the highlights in this newsletter over the next six Tuesdays, starting with the American League East on January 17 and wrapping up with the National League West on February 21.
How accurate is my prediction system? The only honest answer is that I’m not sure. I suspect it’s as reliable as the guesswork or direct extrapolations favored by many prognosticators. But we won’t know for certain until October.
Baseball’s Best (and Worst) 2023 Yearbook
A complete rundown of 2022 stats — and a look ahead at the season to come