A Novel Approach to Analyzing NBA Player Positions
Fans of the National Basketball Association (NBA) are well aware that there are 5 player positions in basketball: point guard, shooting guard, small forward, power forward, and center. These are the traditional positions, but in the modern NBA, the lines between positions have become less clear-cut. Thus, this has led to studies to redefine basketball positions in the NBA by taking various players and finding common qualities that cluster them together to form an overall network.
After extensive data analysis, 13 possible positions were presented. The nodes in the generated network represent individual players and the edges may represent the existence of a shared feature, such as 3PT-Rebounder, Paint Protector, Shooting-Ball Handler, Role Player, All-NBA 1st Team, etc. It can then be deduced that the term “center”, for instance, is merely a label for players over a certain height. Instead, there are actually multiple types of big men in the NBA: ones that can handle the ball and score, ones that protect the paint, ones that score and rebound, etc. Thus, a player like Dirk Nowitzki would be categorized as a Scoring Rebounder, while Joakim Noah is a Paint Protector. Thus, such players are nodes in separate components in the overall NBA network.
It is of particular interest to analyze the “One-of-A-Kind” players such as Lebron James in this network. A player with such various talents and abilities has edge connections to many components like Scoring Rebounder, Offensive Ball Handler, All-NBA 1st Team, Paint Protector, etc. Thus, these unique players act as bridges between distinct connected components of the network, and are difficult to place in a particular category. Consequently, we can label Lebron James as a hybrid point-forward, who defies the boundaries between positions and is capable of defining a new position on his own. It is important to note, however, that the network does not satisfy the Strong Triadic Closure Property. Consider three players, Lebron James, Chris Paul, and Deandre Jordan. While Lebron has edges that are strong ties to both Chris and Deandre, since Lebron and Chris both share ball handling and scoring abilities and Lebron and Deandre can both rebound and block shots very well, there is no edge between Deandre and Chris since they do not share a common attribute.
All in all, this information about the NBA player network can be valuable for teams when making decisions about whom to trade and draft, instead of relying merely on the traditional positions they require.
Source: http://flowingdata.com/2012/03/21/redefining-nba-basketball-positions/