A Glance at the 1913 National League Caught Stealing Data

by Kevin Harlow
December 19, 2004

The Sporting News published caught stealing data for the 1913 National League on December 11, 1913. This makes it the earliest (nearly) complete CS dataset for the AL or NL. The data included the players last name, club, SB, CS, and SB% listed in order of decreasing SB%. Although the source was less than entirely legible, I believe the data presented below is an accurate presentation since I used the previously known SB data, the SB% equation, and the decreasing order of SB% provided as constraints. However, the dataset is not without problems. Maranville is listed as 27 SB (instead of 25 SB) and Smith with 25 SB (instead of 22 SB). There are a few players with 1 and 2 SB in the 1913 NL that are not included. Some players with 0 SB are listed with their CS but it's unknown if that is all of them. Also, "Seymour, Boston" is listed twice in the source. In the data below I've shown the second entry with 1 SB and 2 CS as "UNKNOWN". This could be Seymour with his SB and CS transposed (he had 2 SB and 1 CS) or it could be Jack Quinn or Charlie Deal who played for Boston and had only 1 SB but was not otherwise listed.

The overall SB% is 56.185%. For each player I calculated Stolen Base Runs (SBR) using SB=0.3R and CS=-0.41R. That's a breakeven point of about 57.75%, slightly higher than the actual total SB%. This breakeven point was chosen based on the low 4.15 runs per game, managers being slightly more agressive than they should be, and less efficient defense than today against the stolen base attempt (possibly more errors and more out of defensive position in stolen base situations).

Some misc. comments:
1) Is Sherry Magee the 74% base stealer of 1913 or the 55% base stealer of 1915, the only year that CS data previously existed for him? He was often in the top 6 in SB and 3B so I'd lean toward the higher SB%. I view Magee as at least a borderline Hall of Fame player. I think being a good baserunner (at least a good base stealer) should put him over the top. Kind of like Strawberry with 2100 more PA.
2) Similarly, who's the real Larry Doyle? 73% in 1913 and 55% in 1915. Again, he had incredibly number of 3B to (2B+3B) while at the Polo Ground III.
3) Lots of poor base stealers were "running" too much.

Anyway, here's a table of the data followed by a second table of those players with stolen bases in the NL in 1913 that weren't listed.

1913 NL CS Data from Sporting News December 11, 1913

Players with Stolen Bases in 1913 NL Not Listed

Go to Kevin Harlow's home page.