Building a White Sox 25-man roster from the last 25 years

Last week, Baseball Prospectus’ Russell Carleton threw out a fun little exercise on Twitter that’s a little more complicated than it looks:

He tossed that out there right as SoxFest was about to start, so I had to put it on the backburner. But on a February Friday before spring training, with no new rumors to track and all dead horses further exhausted? Now’s the time to take a crack at it.

Before I started to look at it, I already had four roster locks in mind. I did double-check to make sure I wasn’t missing any kind of hidden greatness, but I didn’t find anything to sway me from cementing ...

The no-brainers

1994 Frank Thomas, 1B: The second of his MVP seasons, Thomas was hitting .353/.487/.729 over 113 games before the strike ended the season. He would have obliterated the team record for homers, as he was on pace for 55. I could’ve picked from several other Thomas seasons to create opportunities for other players’ best attempts, but since Thomas is the franchise’s defining hitter, I’m going with his masterpiece.

1998 Albert Belle, LF: Belle ended up setting the team records for homers (49), doubles (48) and RBIs (152), all of which still stand today. The crazy thing is that his first half was merely OK by his standards.

2006 Jim Thome, DH: With Thomas locked in as a first baseman, it gave Thome a clear path to the roster for his first season on the South Side. He hit .288/.416/.598 with 42 homers, and while Jermaine Dye hit 44 the same year, it was easier to find a right fielder in other seasons.

1999 Keith Foulke, RP: Before he was a closer, Foulke occupied the traditional fireman role, throwing 105 innings over 67 appearances. That kind of workload would’ve been impressive even if he hadn’t struck out 123 while posting a 2.22 ERA, and it was going to stand out on a team with kids still learning how to play.

With these guys in mind, here’s how I filled the rest of the roster around them.

*Used WARP for catchers, and extrapolated 1994 Thomas’ strike-shortened total to a 162-season.

Lineup

With Thomas, Belle and Thome in place, it left only six starting spots to fill. It also blocked off some of the best positions -- Belle in particular made it harder for 1992 Tim Raines and 2008 Carlos Quentin to crack the roster. Fortunately, startling consistency from Ventura, Durham and Ordonez made it easier to accommodate better seasons/shorter careers elsewhere.

Bench

Between Karkovice, Ramirez and Eaton, this is a great defensive bench that provides complements to the starting roster. You could even flip Ramirez with Valentin if you wanted, although I feel compelled to start the mustachioed one at shortstop since Royce Clayton should’ve never knocked him off the position.

Lance Johnson was also a consideration for Eaton’s spot, depending on how much you wanted to prioritize him over the best of Thomas, Ventura and McDowell. (Actually, Johnson might’ve started if he made this roster, considering his best is probably better than Rowand’s). It made more sense to use Eaton as a great corner outfielder when he’s behind subpar defenders in Belle and Ordonez.

It might be cheating to use Tony Philips, who hit .310/.440/.403 over just 36 games before he was traded to Anaheim in 1997. His numbers were in line with his previous several seasons, and more importantly, he started the year with the Sox. That’s why I feel comfortable using him and Crain in this exercise, while I couldn’t comfortably consider 2000 Charles Johnson to bolster the backstops.

Konerko is around to mash some lefties. Jose Abreu could’ve also fit this role, depending on how I shifted around Sale and Quintana.

Starting pitching

Buehrle’s another guy whose consistency made it possible to accommodate others. His 2005 season was his fifth-best according to bWAR, but it works. Likewise, you can pick whichever Sale you want, and then use the Quintana from the vacant season. I went with the per-start dominance Sale of 2014, rather than the strikeout-record version of 2015.

I could’ve flipped around McDowell and Karkovice between 1992 and 1993. McDowell won the Cy in the latter year, which is still the last time a Sox pitcher won one. He didn’t deserve it, and he probably wasn’t the best pitcher on the staff, but Fernandez is still represented on the staff, and in his peak form.

The hardest exclusion is 2003 Esteban Loaiza, who went 21-9 with a 2.90 ERA over 226¹⁄₃ innings (7.2 bWAR), but there’s still a bullpen left to build.

Bullpen

Theoretically, I could’ve just as easily put Loaiza in the rotation and Quintana in the bullpen, but that struck me as cheating since he’s made just three relief appearances in his MLB career.

Instead, I went with a traditional six-man setup. Foulke was a lock, as well as Jenks’ best season in a 90-loss year. Crain was damn near invincible during a 99-loss season until his shoulder gave out, and I’m pretending it wouldn’t happen on this team since overuse is less of a risk. Dotel is the strange one, but 2008’s best players were blocked by other positions, so I went with a guy who racked up strikeouts.

Thornton is the franchise’s leader in appearances, and his best year lined up with kind of a down year for everybody else in 2009. Because the Sox have struggled to find a lefty during or after Thornton’s time, I went back and grabbed Marte at the height of his powers, which unfortunately left Loaiza as a casualty.

How would you have navigated some of these dilemmas?