Optimizing Safe Seating for Live Events at a Division 1 University

Monday, February 22, 2021

As COVID-19 impacted the United States in the summer of 2020, universities struggled to plan for the return of live sports events at their stadiums and arenas. Ticketing and analytics groups at many major universities manually prepared seating plans for the fall football season based on assumptions of limited capacity in a safe environment, without knowing whether fans would be permitted or to what extent. If attendance were to be permitted, it was unclear how a venue could best host different sizes of pods of fans sitting together (most fans attend events with family, friends or classmates). Many universities opted to wait for a blanket software solution from a ticket service provider. The football season, which generates more than $4 billion in annual revenue for the top 65 universities, approached amid extreme health and regulatory uncertainty, and divergent local strategies.

The Need for Rigor, Speed and Adaptability

Executives at one Division I university with a Power Five football program and elite academic standing had been working for months to assess safe stadium seating scenarios, with only limited success. Their spreadsheet-based method could not handle enough scenarios, and it was impossible to promptly and rigorously evaluate policies for the varying stadium sections and season ticket holder populations. The university’s ticketing platform provider did not offer a solution, and the executives were skeptical of an off-the-shelf solution that another provider was reportedly developing.

The executives sought a software tool that would help them account for the complexity of safely accommodating pods of fans of varying sizes, different types of fans (students, non-students, visiting team fans, etc.) and tiers of season ticket holders, and other institutional priorities. The tool would help them plan before and adapt during the season as the COVID-19 environment changed. They retained Princeton Consultants to build and deliver an application for staff to analyze various policies for social distancing, and to assign groups to specific seats or rows using a mathematical optimization system. In Spring 2020, Princeton Consultants was the first firm to publish the description of an optimization-powered safe seating solution, based on its research and algorithmic development.

The Challenging Path to Optimization

The Princeton Consultants and university teams worked together to establish a set of seating rules, such as the maximum number of groups per row and the number of empty rows between filled seats. They then specified optimization parameters, such as maximizing the total number of fans, solution time requested, fixed seats vs. general admission, and proprietary university rules.

The Princeton Consultants team imported the stadium seating manifest and ticketing data. For non-students in attendance, ticket holders were categorized into three tiers. The distribution of pod sizes, which ranged from 2-6 fans, differed among these categories. Additional attendee groups, such as visiting fans, the student band and ADA sections, required different policies.

The physical characteristics of the more than 70 sections were challenging to model. There are bleachers as well as seats with backs, the widths of which vary. Some lower-deck sections are symmetrical on both sides, but corner sections are unique: in one section Seat 10 is next to Seat 20. There are aisles or portals in the middle of many sections, requiring further customization. With regard to social distancing, the university executives requested oval spacing instead of rectangular (seat or row-driven) spacing between fans.

Solution Success

In less than one week, based on its previously developed algorithms, Princeton Consultants built an application that solves in less than one minute for a section and about one hour for the entire stadium. Rules are easily modified to test scenarios, and they can be added to the application so university leaders can assess changing conditions and policies.

The results are mathematically optimized: quantifiably better than the results from manually generated analyses. If the university were to try an off-the-shelf software solution, it is doubtful it would perform comparably. As evidence, Princeton Consultants analyzed the safe seating plan of an NFL team that used a general software solution and hosted fans in September (a consultant is a season ticket holder), and conservatively estimated an additional 1,000 fans could have safely attended games with the Princeton Consultants approach and tool.

The university executives are now planning for live outdoor and indoor events in 2021. Uncertainty continues, making the optimization-powered application a valuable support tool for important decisions.

To learn more, visit our webpage, https://princetonoptimization.com/optimizingsafeseating/.