Stolen Base Regression for Juan Soto & Josh Naylor
Syndication: Arizona Republic
Last season, Josh Naylor and Juan Soto defied all expectations and each stole at least 30 bases. In the previous three seasons combined, they stole just 25 or fewer bases. With such a major jump, I wanted to see how many bases players stole after such a major increase. The answer is way more than I would have guessed.
To find the player pool, I set the following variables since 1980.
In the three seasons before the jump:
- 30 or fewer combined stolen bases
- At least a combined 1200 PA
In the season of the jump:
- 25 SB
- Age 25 or older
Only 12 hitters met the criteria, with Shin-Soo Choo 추신수 not playing the next season. I found the raw number change in stolen bases, along with the overall percentage drop. Here are the players:
| Name | Season | Age | SB | Previous 3 PA | Previous 3 SB | Next PA | Next SB | Difference |
|---|---|---|---|---|---|---|---|---|
| Johnnie LeMaster | 1983 | 30 | 39 | 1270 | 16 | 493 | 17 | -22 |
| Shawn Green | 1998 | 26 | 35 | 1341 | 20 | 696 | 20 | -15 |
| Ivan Calderon | 1990 | 29 | 32 | 1584 | 21 | 537 | 31 | -1 |
| Aaron Boone | 2002 | 30 | 32 | 1278 | 29 | 656 | 23 | -9 |
| Jose Canseco | 1998 | 34 | 29 | 1328 | 15 | 502 | 3 | -26 |
| Carney Lansford | 1987 | 31 | 27 | 1723 | 27 | 607 | 29 | 2 |
| Matt Lawton | 1999 | 28 | 26 | 1496 | 27 | 664 | 23 | -3 |
| Gary Ward | 1985 | 32 | 26 | 1951 | 28 | 418 | 12 | -14 |
| Ivan Rodriguez | 1999 | 28 | 25 | 1950 | 21 | 389 | 5 | -20 |
| Orlando Cabrera | 2002 | 28 | 25 | 1545 | 25 | 691 | 24 | -1 |
| Marcus Semien | 2022 | 32 | 25 | 1707 | 29 | 753 | 14 | -11 |
| Average | 29.8 | 29.2 | 1561 | 23.5 | 582 | 18.3 | -10.9 | |
| Median | 30.0 | 27.0 | 1545 | 25.0 | 607 | 20.0 | -11.0 | |
| % Change | 74% |
The jump in stolen bases rarely happens, with the last player to do it being Marcus Semien, who barely made the cutoffs. The next two were in 2002. Coming in, I would have expected at least a 50% drop, maybe more. Both could have some built-in stolen base upside.
By just going off these numbers, here are the expected number of stolen bases for Soto and Naylor with their current Steamer600 stolen base projections.
Name: Raw Drop, % Drop, Steamer600
Soto: 27, 28.1, 20
Naylor: 19, 22.2, 17
It looks like Soto would have more upside, but both should at least meet current expectations.
Jeff, one of the authors of the fantasy baseball guide,The Process, writes for RotoGraphs, The Hardball Times, Rotowire, Baseball America, and BaseballHQ. He has been nominated for two SABR Analytics Research Award for Contemporary Analysis and won it in 2013 in tandem with Bill Petti. He has won four FSWA Awards including on for his Mining the News series. He's won Tout Wars three times, LABR twice, and got his first NFBC Main Event win in 2021. Follow him on Twitter @jeffwzimmerman.
I wonder if percent jump would show something similar. Ohtani missed the cutoff of 30 or fewer SBs the previous 3 seasons, but his jump to 59 SBs last year was similarly dramatic. And he regressed back to 20 SBs. He may be a special case for stealing less when pitching, but a quick scan of ‘fluke’ steal seasons shows regression back to previous norms for some (Brock, Willy Taveras), but a new level for others (Otis Nixon–maybe the poster child for elderly basestealing).