Predicting performance in 4 x 200-m freestyle swimming relay events

Area of Science:

• Sports Science
• Performance Analysis
• Swimming Analytics

Background:

• Understanding performance differences between individual and relay swimming events is crucial for strategic team assembly.
• Predictive modeling can offer valuable insights into optimizing relay event outcomes.

Purpose of the Study:

• To predict and understand performance variations in swimming between individual and relay events.
• To develop a predictive model for the 4x200-m freestyle relay event to inform team selection and strategy.

Main Methods:

• Analysis of 716 4x200-m freestyle relay finals from 2010-2018.
• Inclusion of individual 200-m freestyle season best times for each swimmer.
• Application of linear regression and machine learning techniques.

Main Results:

• The lowest-ranked swimmer and American swimmers showed faster 200-m times in relay events compared to individual events.
• Random forest models demonstrated high sensitivity in predicting medal outcomes (gold: 100%, non-medal: 93%).
• Team finishing position strongly correlated with differential times, world rankings, and swimmer order.

Conclusions:

• Team strength (differential times, world rankings) and strategic swimmer order are key determinants of relay success.
• Predictive analytics can assist coaches in swimmer selection and strategic lineup decisions for relay events.
• This approach can enhance the likelihood of success in competitive swimming relays.