Could the 2023 World Series Have Been Predicted?

Take my hand and let me lead you down a rabbit hole. (Or let go before it’s too late!) The 2023 World Series pitted the Arizona Diamondbacks against the Texas Rangers. Could this unlikely matchup actually have been predicted?

 

The Superficial Take

The Diamondbacks finished the 2023 regular season with a record of 84-78, i.e. a winning percentage of .519. Sure, the 2006 St. Louis Cardinals won the World Series with a winning percentage of .516, but the Diamondbacks barely made the playoffs and had no business representing a National League (NL) that saw one team win 104 games (Atlanta) and another win 100 (Los Angeles).

The Rangers ended the regular season with a record of 90-72, i.e. a winning percentage of .556. Their presence was not as distressing, but in an American League (AL) with a 101-game winner (Baltimore), a 99-game winner (Tampa Bay), and the defending champion (Houston), it was also far from predictable.

 

A Little Digging

Let’s start with the Rangers this time.

Run differential (runs scored minus runs allowed) isn’t the most refined statistic, but over the course of a 162-game season, it does tell us something about a team.

It turns out that the Rangers had a regular-season run differential (DIFF) of +165, second best in the AL and fourth best in all of baseball.

The closely related Simple Rating System (SRS), which takes into account per-game run differential and adjusts for strength of schedule, gave the Rangers a 0.9, also good for second in the AL and fourth overall.

So why, then, did the Rangers win only 90 games if they excelled in two categories that are supposed to correlate with winning percentage?

Luck, or lack thereof.

Their expected record based on runs scored and runs allowed was 96-66 (.593).

As for the Diamondbacks, we’re still left scratching our heads.

With a DIFF of -15 and SRS of -0.1, the Snakes were about as middling as it gets. (Yes, the 1987 Minnesota Twins did win the World Series with a DIFF of -20, but that’s anecdotal.)

In fact, Arizona’s expected record taking this data into account was 80-82 (.494), hardly championship material.

 

The Rabbit Hole

Logic would have it that an average team with average players would be unlikely to suddenly exceed expectations.

There’s only one issue—Arizona didn’t have average players.

As defined by Wins Above Replacement (WAR)—click here if you need a refresher—the Diamondbacks had elite talent. Three of the top 25 NL players by WAR played for the Snakes, about double what would have been expected by random distribution alone. And when talent of that caliber performs at the right time, the results can be uplifting.

Entering the World Series, Corbin Carroll (WAR 6.0) had a postseason on-base percentage (OBP) of .396 and went 3 for 4 with 2 RBIs and 2 runs scored in game 7 of the National League Championship Series (NLCS). In the same timeframe, Ketel Marte (WAR 4.2) had a postseason OBP of .382 and at least one hit in every postseason game, earning the NLCS MVP Award.

The outcome: A Fall Classic.

Not that we need to dig this deep for the Rangers (whose team stats were more than sufficient to explain performance), but let’s roll.

Two of the top three and three of the top 15 AL players by WAR played in Arlington (TX).

Entering the World Series, Corey Seager (WAR 6.1) had a postseason on-base plus slugging (OPS) of 1.127 and went 3 for 5 with a key home run in game 7 of the American League Championship Series (ALCS). Over the same period, Adolis García (WAR 4.8) had a postseason OPS of 1.102 with 7 home runs and an ALCS MVP Award.

The translation: Halloween baseball.

 

So could the 2023 World Series matchup have been predicted? Uh, probably not.

Addendum: The Rangers won the best-of-seven series 4-1, earning their first ever World Series title.