In Olympic cycling, the difference between gold and missing the podium is measured in fractions of a second. For the USA Cycling Women’s Team Pursuit squad, the 2023 World Championships in Glasgow highlighted a stark reality: Finishing 6th place with a time of 4:12.684 and missing the qualification round for the bronze medal final by 0.159 seconds. With the Paris 2024 Olympics approaching, the challenge was clear: reduce nearly seven seconds off their performance in just one year, an improvement considered nearly impossible in elite sports. With limited funding compared to powerhouse nations like Great Britain and Germany, USA Cycling relied on operations research (O.R.), machine learning (ML), data analytics, race simulation, and targeted athlete development to bridge the gap. Armed with cutting-edge analytics and modeling, real-time performance tracking, and aerodynamic innovations, this initiative aimed at optimizing every aspect of the team’s race strategy. Athlete injuries early in 2024 cemented the need for a comprehensive solution as key athletes came into the games below their 2023 performance peaks, dropping the team to an 8th-place World Ranking. Through the power of data-driven decision-making, USA Cycling achieved what few thought was possible: a stunning eight-second reduction in time to capture Olympic gold, with a time of 4:04.32.
The Women’s Team Pursuit event is a highly strategic 4-kilometer (2.5-mile) race where success depends on precise pacing, strategic rotation of riders at the front, and the well-timed drop of a fatigued cyclist. It is a well-coordinated race in which four riders must race on an indoor velodrome, with the team’s time recorded when the third rider crosses the finish line. Facing resource constraints, Team USA entered the new Olympic cycle, lagging top teams, such as Great Britain, New Zealand, and Italy, by nearly seven seconds, a margin traditionally considered insurmountable within a single season. Unlike wealthier federations with year-round velodrome access and unlimited wind-tunnel testing, USA Cycling had to ensure data backed every training session, aerodynamic refinement, and strategic decision to yield the highest return on investment.
The primary catalyst of Project 4:05 is a time determination model, developed using historical Olympic and World Championship results while accounting for environmental factors such as air density and altitude. The model identified 4:07 as the likely threshold for a podium finish and 4:05 as the gold-contending mark, setting a clear and data-driven performance target. The team implemented a mixed- integer programming (MIP) optimization model to determine the most efficient race execution. The model factored in each athlete’s Critical Power (CP20), Functional Reserve Capacity (FRC), and aerodynamic drag coefficients to calculate the optimal order and duration of rider pulls at the front. It also identified the best moment to drop a cyclist without disrupting draft efficiency. The MIP was validated by reconstructing past races, consistently predicting finish times within a 1% margin of actual results. This gave the coaches confidence to adopt the strategy in real-world competitions.
Competing for the Edelman award places you and your company among the upper echelons of those within the O.R. and Analytics fields today.
Because aerodynamic drag becomes the dominant force, slowing cyclists at speeds above 60 km/h (40 mph), the team partnered with Vorteq Sports to develop aerodynamic simulations that guided incremental but critical performance improvements. Each athlete underwent high-precision body scans to design customized skinsuits tailored to her posture and riding style. Limited wind-tunnel time required a highly targeted approach, so the MIP’s analysis guided decisions on which gear or rider positional tweaks would produce the most significant impact. Real-world validation came through on-track testing using aerodynamic sensors, ensuring theoretical gains translated into improved times. These refinements drove the reduction of each rider’s drag coefficient and boosted overall efficiency.
Real-time Z-score performance benchmarking further enhanced training efficiency. By continuously comparing each rider’s power output, endurance capacities, and aerodynamics against “gold-capable” thresholds, coaches could make targeted and dynamic adjustments through daily feedback loops. Whether focusing on anaerobic power intervals, refining pacing strategy, or tweaking riding positions, every decision was grounded in the continuous flow of performance data.
The impact of Project 4:05 became undeniable in Paris. The team shattered previous U.S. records with 4:05.20 time in the qualifying round, validating the optimization strategy by nearly hitting the projected gold-medal threshold. Before the semifinal against Great Britain, a last-minute adjustment based on the real-time air density readings led to a 4:04.629 finish to secure a spot in the final. In the final against New Zealand, the highest-ranked team from internal projects for gold, the riders executed their MIP-optimized race plan to perfection by meticulously following the optimal rotation schedule and drop strategy. Crossing the finish line in 4:04.32, team USA not only captured an Olympic gold but also proved that data-driven decision-making can fundamentally alter the competitive landscape in elite sports. The team cut eight seconds from their 2023 World Championship time, an unheard-of transformation in professional track cycling while operating on less than 50% of the funding available to other top-tier competitors.
The impact of Project 4:05 reverberated well beyond a single event. Project 4:05 is now being studied as a template for other Olympic disciplines with the U.S. Olympic & Paralympic Committee to replicate the methodology, embracing a long-term vision for integrating analytics and operations research across the board. Project 4:05 is an enduring blueprint for how evidence-based decision-making can spark dramatic gains in elite sports.