From CS 294-10 Visualization Sp10

Revision as of 09:16, 11 May 2010 by Stephen Chu (Talk | contribs)
(diff) ← Older revision | Current revision (diff) | Newer revision → (diff)
Jump to: navigation, search



Group Members

  • Stephen Chu


The use of statistical analysis in professional sports has become increasingly popular. While baseball teams rely heavily on analytics, some basketball and most football teams are skeptical about its value. The Boston Red Sox and the Dallas Mavericks attribute their successes partly to the competitive advantage gained from their analytics teams. At the 2010 MIT Sloan Sports Analytics Conference, Bill Polian, president of the Indianapolis Colts, said he doubts that sports analytics will ever have a significant impact in the football industry because the game is so team-oriented and there hasn't been a way to account for the vast variety of techniques and strategies. Furthermore, he believes that the data taken from one season is too small of a sample size to ever be considered significant. Mark Cuban, on the other hand, voiced his belief that the more successful teams in the NBA invest in analytics.

I plan to focus on the NBA dataset and create a useful interactive visualization tool that should give teams further insight about their players' abilities, team's chemistry, and specific lineups they use against opponents. There is a lack of information visualization tools for NBA fans and teams, even those who focus heavily on statistics. Most fans go to ESPN.com or NBA.com as vendors for data, but these sources mainly provide raw datasheets, making it hard to see all the important relationships between teammates and their opponents. Even the "expert analysts" rarely use visualizations when presenting their arguments supported by numbers. NBA team owner Mark Cuban claims that hiring an analytics team has given his Mavericks an edge, but other teams such as the Clippers fail to see the importance of analyzing statistics closely. perhaps leading to the Clippers' continued lack of success. I believe an interactive visualization tool will be a way to demonstrate the usefulness of sports analytics while uncovering the important hidden relationships between the data.

Initial Problem Presentation

Final Deliverables

The shot chart prototype was built with Processing. Open processing.exe, then open shot_chart_40.pde and run. Please email me if you have any questions about running the code or if the links don't work.

Screenshots of executable

Image: Screenshot2.PNG

The shot chart configuration above displays the shooting percentages by floor position when the selected players play at the same time. The left and right side of the court contain Lakers shots and Celtics shots, respectively. The lower left shows the shooting percentages of each player and team. The lower right shows team performance stats that provide some indication of how well the teams match up when the selected players are on the court together. A value of -1 means that there were no recorded statistics for the corresponding category.

Image: Screenshot4.PNG

The shot chart aims to visualize how different combinations of players perform together. The image above shows the change in shooting percentages when Trevor Ariza takes the place of Vladimir Radmanovic.

Note about the dataset: The data is taken from Lakers-Celtics games of the 2007-2008 season. However, not all of the data from these games have been included. I'll address data collection problems in the paper.

[add comment]
Personal tools