When I was a kid, our elementary school used to have these book fairs, run by Scholastic book publishers. The company set up a display in the lobby outside the library, their various books for sale, and each spring I would spend a dollar or two on the latest edition of "All-Pro Baseball Stars." I guess the book fairs are still around, although the last edition of "All-Pro Baseball Stars" appears to be 1984.

It was a wonderful little paperback, 92 or 96 pages long, with two-page bios of the best players from the previous season, plus some statistics and standings and so on. Bruce Weber was the author, and he began the series in 1976. Before that, there was a similar series simply called "Baseball Stars of 1967" or whatever the year, with Ray Robinson as writer.

Thinking of those books, I got to wondering: Who are the all-pro baseball stars of 2021? Of course, that description doesn't really fit; all-pro is more of a football term (there was also an "All-Pro Football Stars" series). Maybe the idea is simply this: Who are the superstars of 2021? And how many of them are there? And as we begin the season, how many of them will get bumped off the list by the end of the season?

I like the idea of "superstar" as inexact science. You have to earn your way to the superstar description -- and one bad season shouldn't necessarily knock you off. A list of superstars is mostly about the best players in the game, but it goes beyond that, right? Winning matters, maybe popularity counts. It's not just about projecting the best players of 2021 -- which is what our MLB Rank of the top 100 players does -- but it's also not about career value. Albert Pujols and Miguel Cabrera are not currently superstars.

I concluded -- disagree if you wish -- that there are 24 superstars in the majors in any given season, rationalizing that number this way: If we were to play the ultimate All-Star Game, we'd need 18 position players. Then we'd need three pitchers per team, six in total. That gives us 24 superstars. My list will be position agnostic, however, so if we end up with five shortstops and no catchers, that's OK.

I said inexact science, but I don't want this to be completely subjective, so I came up with a formula to rank the players:

1. Start with three-year total WAR (I used FanGraphs). This is the "earn" way part. One great season might not be enough to crack the top 24.

2. Bonus points for MVP voting results from the previous three seasons, awarded as such:

2020: 5-4-3-2-1
2019: 4-3-2-1
2018: 3-2-1

This gives more credit to recent seasons, so a 2020 MVP season is valued more than a 2018 MVP season.

3. Bonus points for Cy Young voting results from the previous three seasons, awarded as such:

2020: 5-4-3
2019: 4-3-2
2018: 3-2-1

(No double-dipping for pitchers, so they can't get points for finishing high in both Cy Young and MVP voting.)

4. Two points if you were the World Series MVP in any of the three previous seasons. Call it a bonus for clutch performance on the biggest stage.

5. One point if you were an LCS MVP in any of the three previous seasons. Similar idea: Postseason results matter.

6. One point for each World Series champion you played on in the three previous seasons. The goal is to win.

7. One point if you had one of the top five selling jerseys the previous year (in 2020, the top five jerseys belonged to Mookie Betts, Aaron Judge, Bryce Harper, Fernando Tatis Jr. and Juan Soto, so they each get an additional point). A way to reward popularity.