Trevor Masters & Danny Persia
Bringing Extinct Sponges to Life: Stromogrow – A New Program for Modeling Stromatoporoid Growth
by Trevor Masters
Most geoscience students readily learn how organisms are shaped by their environment. However, few actually have an opportunity to explore the direct relationship between environmental pressures and an organism's response to them. For example, phototropic plants grow towards a bright light in order to have greater access to this vital resource. But under what type and intensity of light does this occur? How far can the plant grow in any given direction? How do other forces, such as wind and gravity, act in opposition to phototropic growth? By exploring the direct relationship between an environmental pressure and an organism’s response, answers to these kinds of questions can become evident more easily. StromoGrow explores these types of direct relationships between a stromatoporoid and its environment, and therefore functions as both a teaching and a learning tool. Our goal was to produce a stromatoporoid-‐generating program for use in the classroom, to help students understand how stromatoporoids grew, and assist them in visualizing their three-‐dimensional growth when given a two-‐dimensional cross-‐section. We wanted the tool to accomplish the following goals: 1) to model stromatoporoid growth in three-‐dimensions; 2) to make the program easy and intuitive to use; 3) make it fast enough to grow large organisms in a reasonable amount of time; 4) to make a fully interactive organism that could be easily rotated and inspected; and, 5) make it highly configurable to handle a variety of conditions and settings.
Modeling NBA Player Value from Box Score Data
by Danny Persia
Recent finding in basketball analytics reveal the importance of five-‐man units in modeling player interactions on the court. In a synergistic environment, how much is a shot worth? How much should we value an assist? A turnover? A block? In this session we’ll explore how box score data can be used to evaluate NBA player performance. We introduce an adjusted plus-‐minus statistic, generated for NBA players across five seasons, and regress against linear and non-‐linear variations of the box score. We consider models to predict future performance, with an eye toward determining which player types play well together, producing synergistic effects. Can a reasonable definition of player value be determined?
Refreshments will be served at 3:30 p.m. Talk begins at 3:45