MLB Teammate Analysis - Academic and Mathematics Research

Academic and Mathematics Research

The application of Social Network Analysis (SNA) is not completely new to professional and other sports analysis. Academic researchers and others have applied the concepts of SNA to better understand several different sports. However, almost all prior applications of SNA to sports have focused on "micro-level" analytics as opposed to the "macro-level" analytics we use.

Micro-level SNA looks at connections made between players within a single game or a small number of games. For example, one might look at the network created when soccer players pass the ball during one possession in a soccer game, when basketball players pass the ball during one possession in a basketball game, or when baseball players throw a baseball during a single baseball play. SNA can be a very useful tool in understanding the flow of play in any of these or other sports.

Macro-level SNA examines the connections made by athletes across many years. Our own analysis examined Major League Baseball starting rosters for every regular season game from 1914 through 2015 (that's right - 101 YEARS of game data!). The connections we examined were created when two players were listed together as teammates on the starting roster of a regular-season MLB game. We believe these connections are very important in understanding how skills are attained or passed from one player to another. That is, it is generally human nature to adopt the skills we see in others when we complete a task with them. If we can complete the task with many other individuals and not just the same individuals, we increase the probability that we will work with someone who has skills that we do not have, and we can then incorporate their skills into our own skillset when we next work on the task. This is why it is most important to know not just the total number of other players a professional athlete has played with, but how many different players they have played with over time.

Some past academic research papers that show applications of SNA to sports include:

Beckman, P. and Chi, J. “More Highly Connected Baseball Players Have Better Offensive Performance”, Baseball Research Journal, Fall, 2011. [Macro-level SNA applied to U.S. professional baseball.]

Fewell, J., Armbruster, D., Ingraham, J., Petersen, A., and Waters, J. “Basketball Teams as Strategic Networks”, PLOS ONE, Nov. 6, 2012. [Micro-level SNA applied to U.S. professional basketball.]

Grund, T. “Network Structure and Team Performance: The Case of English Premier League Soccer Teams.” Social Networks 34.4 (2012): 682-690. [Micro-level SNA applied to U.K. professional football (soccer).]

Piette, J., Pham, L., and Anand, S. “Evaluating Basketball Player Performance via Statistical Network Modeling”, MIT Sloan Sports Analytics Conference, 2011, March 4-5, 2011, Boston, MA, USA. [Macro-level SNA applied to U.S. professional basketball.]

Reifman, A. “Network Analysis of Basketball Passing Patterns II”, NetSci2006, May 22-25, 2006, Bloomington, IN, USA. [Micro-level SNA applied to U.S. college basketball.]

Skinner, B. “The Price of Anarchy in Basketball.” Journal of Quantitative Analysis in Sports, 6.1 (2010). [Micro-level SNA applied to U.S. professional basketball.]