Project Info
Advanced Magic Number Calculation for Major League Baseball
Alexandra M Newman
anewman@mines.edu
Project Goals and Description:
Standings in the major professional athletics leagues are published daily in newspapers and updated in real time on sports websites. These outlets typically report when teams are eliminated from the playoffs (i.e., have fallen so far behind that they cannot qualify even if they win all of their remaining games) and provide so-called “magic numbers” that represent how close the leading teams are to the opposite state, that of clinching a playoff spot. Fan interest in the standings helps drive traffic to commercial websites, determines which games to televise nationally, and influences what merchandise to offer and when to offer it. Teams and their managers have interest in this information because they can start selling tickets for home playoff games as soon as they have clinched a spot. Accurate assessment of teams’ playoff chances can also help coaches plan playing time for injured or rookie athletes. Sports media primarily rely on heuristics to make such assessments, which can lead them to announce eliminations or clinches late or even erroneously.
The Remote Interactive Optimization (RIOT) Sports website, riotsports.net, is powered by a collection of integer-programming optimization models that consider the teams' schedules and the leagues' tie-breaking rules to yield more accurate and informative magic numbers for Major League Baseball, the National Basketball Association, and, more recently, the Women's National Basketball Association, and the Korean Baseball League. Using these models, RIOT can announce when a team has clinched or been eliminated from a playoff spot days before it is reported in popular media outlets. These models are run nightly on a behind-the-scenes server to generate up-to-date information regarding enhanced and traditional magic numbers. The site, uses publicly available data to compute traditional magic numbers (i.e., playoff elimination, playoff clinch, first place elimination and first place clinch), which are updated every day based on current standings during the regular season.
More Information:
Grand Challenge: Not applicable.
| [1] | I. Adler, A. Erera, D. Hochbaum and E. Olinick, "Baseball, optimization, and the world wide web," Interfaces, vol. 32, no. 2, p. 12–22, 2002. |
| [2] | A. Fahey, "Warriors Clinch Sixth Consecutive NBA Playoff Berth," 12 March 2018. [Online]. Available: https://www.nba.com/warriors/news/playoffs-2018-clinched. |
| [3] | M. Husted, E. Olinick and A. Newman, "Improving sports media’s crystal ball for National Basketball," INFORMS Journal on Applied Analytics, vol. 51, no. 2, pp. 91-165, 2021. |
| [4] | L. Robinson, "Baseball playoff eliminations: An application of linear programming," Operations Research Letters, vol. 10, no. 2, p. 67 – 74, 1991. |
| [5] | C. Ribeiro and S. Urrutia, "An application of integer programming to playoff elimination in football championships," International Transactions in Operational Research, vol. 4, no. 12, pp. 375-386, 2005. |
Primary Contacts:
Alexandra Newman, anewman@mines.edu
Student Preparation
Qualifications
Computer science major, or math major with good coding skills
TIME COMMITMENT (HRS/WK)
5
SKILLS/TECHNIQUES GAINED
Roles:
*Collect data
*Write scrips
*Help build optimization models
*Run optimization models
*Analyze results
*Disseminate results to research sponsors and/or on websites
Learning objectives:
*Work on multi-disciplinary project
*Understand how to write and interpret optimization models
*Be able to disseminate results and communicate them
MENTORING PLAN
*Meet with professor at least once a week
*Meet with PhD students multiple time per week
*Provide opportunities for students to present results both within Mines and externally
Preferred Student Status
Sophomore
Junior
Senior