Statistical Analysis Affirms Larranaga’s Coaching Style
Posted: December 12, 2002 at 1:00 am, Last Updated: November 30, -0001 at 12:00 am
By Robin Herron
Men’s basketball head coach Jim Larranaga typically does a postseason analysis every spring, and every spring he considers whether he should change anything in his approach to coaching. This year, he decided to rely on some hard data, and he turned to the School of Information Technology and Engineering for help.
Department of Applied and Engineering Statistics professors Clifton Sutton and Donald Gantz, who staff the department’s Statistical Consulting Center, met with Larranaga last spring to discuss analyzing the team’s basketball data.
“We learned that Larranaga was a former math major and had a good ability to think quantitatively,” says Sutton. “There were oodles of things he wanted to look at, but we decided it would be better to have a particular focus.” They chose to look at defense.
“A main objective was to see if the data from last season’s games could be used to suggest whether a particular defensive style that he had been using seemed better than an alternative style which would lead to better rebounding on the defensive end of the floor, but would also lead to the opponent having a higher field goal shooting percentage,” Sutton says.
Because of scheduling conflicts, Gantz and Sutton were unable to work on the project together, so Sutton did the final analysis. Using last season’s box score data (field goals, missed field goals, rebounds, turnovers, blocks, assists, fouls, free throws, missed free throws, and steals), he developed probability models and determined the relationship between shooting, rebounding, and the overall chance of scoring before giving up the ball to the defense.
In August, Sutton presented the coach with a binder full of explanations, data summaries, and graphs, and they discussed the results for more than two hours. Then, last month,
Sutton gave the same presentation at a Statistics Colloquium Series seminar, which was attended not only by statisticians but also members of the coaching staff.
The verdict? “My model indicates that the style of defense that Larranaga had been teaching for years seemed to be much better than an alternative style which would give up a higher shooting percentage in return for increased rebounding strength,” Sutton says.
That’s good news for George Mason and Larranaga, but not surprising, since he has finished 18 seasons as a college head coach (5 at George Mason), and guided the Patriots to two Colonial Athletic Association league championships and two NCAA Tournament bids in the last 3 seasons. He has also written a book, Coaching Basketball’s Scramble Defense.
“In the book The Art of War, Sun Tzu says clearly that to win you must first know thyself and thy enemy,” states Larranaga. “Before a smart general sends his troops into battle, he wants as much knowledge as possible about his army and his adversary’s army. For a general, the information you get from good intelligence is essential to making wise decisions.
“The same is true for a coach. But, our information comes from analyzing critical statistics. Dr. Sutton has analyzed, dissected, and presented his findings after reviewing our stats in a clear and helpful fashion for my staff and myself. It will prove to be extremely beneficial. Utilizing last year’s stats will help us stay focused on the things we need to do to win this season,” Larranaga says.
The response to Sutton’s work and his presentation has been such that he decided to do a game-by-game analysis this season, adding refinements as he goes along.
Sutton estimates he spent about 135 hours on the initial data analysis and developing his “Sutton System,” but says it would have taken him much longer if he had not already given a lot of thought to applying probability and statistics to basketball. A fan of Washington’s professional women’s basketball team, the Mystics, Sutton was convinced that their 2001 coach, who didn’t return for 2002, wasn’t using his players correctly. “That got me to giving serious thought about how to assess players’ contributions–how to combine shooting, rebounding, and other aspects of basketball into an overall measure of value.”
Basketball is not the only sport that Sutton has considered in a statistical light. Several years ago he and colleague Carl Harris, now deceased, began analyzing old Negro Leagues baseball data to show that some of those players should have been selected for the All-Century Team. “Possibly, in addition to lack of knowledge about these players, some people didn’t vote for them because statistical comparisons are difficult. I plan to use statistical methods to relate Negro League statistics to those from the National and American Leagues,” Sutton says. “If anybody’s interested, I could use some help.”
For more information on the Sutton System, click here.