Predicting the outcome of an NBA game has always stumped fans and followers of the game, but studies and theories are often examined, even on the ASU campus.
A team of researchers at ASU is working on a project that studies the network dynamics of teams in the NBA, and has found the secret behind which teams are the most successful and the safest ones to put your bets on.
Jennifer Fewell, an ASU associate professor of life sciences, decided to come up with a study to combine her knowledge of social networks with her love of basketball. Fewell has been a Suns fan since Charles Barkley took the Suns to the NBA Championships against the Chicago Bulls in 1993.
In spring 2009, Fewell attended a presentation by Dieter Armbruster, an ASU professor of mathematics, about examining networks. Fewell and Armbruster paired up and quickly set to work on creating the basis for the NBA analysis project.
With the help of two students, John Ingraham and Alex Petersen, the group analyzed 16 2009-2010 NBA playoff games, and documented every ball possession in the games.
It was an immense amount of work. Each NBA game averaged between 70 to 100 ball possessions.
One of the research students, 21-year-old John Ingraham, a mathematics and biochemistry major, was responsible for writing the data collection software used to input data when watching the games.
“I've been interested in basketball although not a dedicated fan,” Ingraham said. “Having to collect the type of data we did gave me more appreciation for the league and how the game is played.”
The research team coded the different players as “nodes” and ball movements as “links.” They kept track of every ball movement in all the games, and all movements combined made up the "network" for each team.
After calculating how many links the network had they could analyze the differences between the networks of different teams.
The teams that had played together longer did better in the playoffs, Fewell said.
“The teams that are successful are not the ones that are newly put together,” Fewell said. “They know how to play together. They have to have talent, but they also have to be able to coordinate that talent.”
From analyzing the Suns in the 2009-2010 playoffs, Fewell learned that Amare Stoudemire tended to shoot the ball instead of passing it, which allowed people to double- and triple-team him, since they knew he wasn’t going to pass it to other people.
“Grant Hill was really important to distributing the ball throughout the team,” Fewell said. “He’s involved in almost all of the successful networks.”
Another thing the group learned was that the more players the teams have involved in passing and playmaking, the more unpredictable their play is and successful they are.
"What we've found from the games we watched is that the team that is more unpredictable will win over the team that is more predictable," Armbruster said.
Teams that fell into this unpredictable category included the Los Angeles Lakers, the San Antonio Spurs and the Boston Celtics, Armbruster said.
In the Suns’ case, Steve Nash dishes the ball to out to most players, and they more often return it to Nash than other players, so the Suns are more predictable than other teams.
Fewell and Armbruster are now working to analyze the NCAA March Madness games.
“In particular we’re very excited about trying to find out if we can quantify the style of difference between NBA basketball and NCAA basketball,” Armbruster said.
The research group hopes its networking project will extend to reach groups other than basketball.
“The research can apply to any team that networks, such as teams that work together in an industry or military group,” Armbruster said.
Reach the reporter at nicole.klauss@asu.edu
