When the Indians acquired both Michael Bourn and Drew Stubbs this past offseason, I think many of us expected a lot more than we ended up receiving in the stolen base department.  Towards the end of the year, Bourn made the claim that switching leagues was a major reason for his decline in stolen bases in 2013.  Bourn’s point was that he was trying to steal bases against many pitchers who were unfamiliar to him.  Thus, he was unfamiliar with the pick-off moves and timings of many of the American League pitchers.  I decided to investigate this a little bit to determine if there was any legitimacy to Bourn’s claim of changing leagues affecting stolen bases (at least in the first year after the change).

I took a look at all base-stealers that changed leagues since the year 2000.  I defined “base stealer” as a player that had at least 20 SB the year before changing leagues.  Thus, the before SB total is listed in the “Before” column, and the total the following season after changing leagues is denoted in the “After” column.  The “Year” column indicates the year the player obtained the “After” SB total amount.  Obviously the “Difference” column subtracts the after column from the before column.  In addition, all players must have had at least 300 AB in both seasons to qualify, and in the case that players played in both leagues over the course of a season (if they were traded), counted them as qualifying for the league in which they had the majority of their AB.

Here is some of the data that I obtained:

Player Year Before After Difference

Percent Change

Bonifacio

2013

30

28

-2

-7%

Bourn

2013

42

23

-19

-45%

Revere

2013

40

22

-18

-45%

Victorino

2013

39

21

-18

-46%

Choo

2013

21

20

-1

-5%

Upton

2013

31

12

-19

-61%

Stubbs

2013

30

17

-13

-43%

Reyes

2013

40

15

-25

-63%

Pierre

2012

27

37

10

37%

Pierre

2010

30

68

38

127%

Abreu

2007

30

25

-5

-17%

Lofton

2007

32

23

-9

-28%

Soriano

2006

30

41

11

37%

Podsednik

2005

70

59

-11

-16%

Beltran

2005

42

17

-25

-60%

Womack

2005

26

27

1

4%

Young

2004

28

14

-14

-50%

Lofton

2003

29

30

1

3%

Sanchez

2003

37

52

15

41%

Durham

2003

26

7

-19

-73%

Cedeno

2002

55

25

-30

-55%

Henderson

2002

25

8

-17

-68%

Macias

2002

21

8

-13

-62%

Cedeno

2001

25

55

30

120%

Henderson

2001

36

25

-11

-31%

Cameron

2000

38

24

-14

-37%

Mondesi

2000

36

22

-14

-39%

Henderson

2000

37

36

-1

-3%

Green

2000

20

24

4

20%

Goodwin

2000

39

55

16

41%

 

Here are some statistics about the data.  The average difference between the two groups is -5.73 and the median difference is -6.5.  For the stats geeks out there, we obtain a p-value of .061 after doing a paired-sample t-test.  Thus, our test is close to being significant at the .05 level, and it is significant at the .10 level.  There does seem to be  at least somewhat of an effect on switching leagues on players that are considered to be pretty good base stealers.

I also added the column percentage change to the study as well to get a more balanced study across all observations.  This way someone that steals 70 bases their “before” year and 11 fewer their “after” year is not analyzed on the same scale as someone who stole 20 bases their “before” and 11 less their “after” year.  We get a mean percentage change of -14%, with a median of -29%.  This is important as it tells us that far more players were negatively affected by the league change than players that improved their SB total after the league change.  It looks like Bourn does have a valid point to some degree about the learning curve of base stealers when changing leagues.   Now does this completely explain why Bourn had a 45% and Stubbs had a 43% decrease in stolen bases in 2013?  Probably not.  However, it does seem to serve as one reasonable partial explanation at the very least and it does back up Bourn’s point.

9 Comments

  • Gvl Steve says:

    Thanks for all the work you put into that. I always figured that changing leagues would be a hard adjustment, but I was surprised at how much the statistics bore that out. Now the next question is, how much do those players rebound their 2nd year in the new league? I would expect that two years of aging would have a negative effect on stolen base totals generally, but if the league switch is at least partly to blame, we should see some recovery the 2nd year.

    • Ryan Pinheiro says:

      Great question and good points Steve. I’m actually planning on writing a Part II (guess I should’ve called this Part I!) to this post and taking a look at what happens the second year after the league change. Should be up sometime next week most likely. Thanks again for the comments!

  • Doug says:

    As a control, you might want to look at players who changed teams while staying in the same league. I think I read an article that suggested players often underperform in their first year with a new team.

  • Xavery says:

    With the DH in play, there aren’t very many calls from the bench to steal. Most times, it’s the players discretion in some situations though. More hit and run calls than steals. I have a hard time thinking that most pick off moves are all that different. Good base stealers are good no matter who is on the mound. Ask Vince Coleman.

    • Cale says:

      Good point. Is there a difference going from NL to AL as opposed from AL to NL? Just looking at the big changes from 2013, they were all NL to AL changes except Revere.

      • Ryan Pinheiro says:

        I’ll try to look at difference in switching between AL to NL and vice versa in one of my next posts. Thanks guys.

  • Ryan McCrystal says:

    Awesome stuff Ryan… I’m left with a couple of questions to take this even deeper…

    To what extend does age factor in? Most players don’t switch teams and leagues until they’re in or past their prime. Many of the guys on the list are in their 30s at the time of the league change. Any way to control for this with your data? I wonder what the average dropoff is for all players past the age of 30.

    If possible, could you run the same data with players grouped as 30 and under, and over 30? Or maybe 32 and under? I’m sure there’s a breaking point for a league-wide dropoff regardless of league changes.

    Also, I wonder if SB per game would alter the results. Some of the largest changes appear to be influenced simply by opportunities. For example, in Cedeno’s big leap, he only improved by 25% in terms of SB per game, rather than 120% in total SB.

    • Ryan Pinheiro says:

      I like your ideas Ryan. I’m planning on writing a part II (and maybe even a part III) to this article based on subsequent questions myself and readers have. I could definitely group the players into age groups and run the data again. It’d probably be beneficial for me to go back a few years farther to ensure a decent sample size in each group. I like your idea of SB/game as well to put everyone on the same playing field. I’ll switch over to that in my next posts. Thanks Ryan!

  • Peter says:

    Let’s not forget, particularly in Stubbs case, you have to get on first to steal second. What about comparing the On Base % in the above chart?