Author(s): LOISEL, C., SEIFERT, L., Institution: UNIVERSITY OF ROUEN NORMANDY, Country: FRANCE, Abstract-ID: 1113

INTRODUCTION:
Swimming race management is typically analyzed by examining distinct race segments such as the start, clean swimming, turn, and finish. However, most analyses of the clean swimming segment rely on lap averages, which overlook meaningful cycle-to-cycle variability. Pacing in swimming primarily concerns how swimmers distribute time between laps, as the presence of walls divides the race into successive segments; accordingly, lap time management was the first focus of race analysis (McGibbon et al., 2018). A major task constraint is to minimize speed loss when approaching the wall and to maximize speed when pushing off the wall (Simbaña-Escobar et al., 2018). The aim of this study was to utilize cycle-to-cycle data to investigate intra- and inter-lap variability in stroking parameters, swimming speed (S), stroke rate (SR), and stroke length (SL), in order to identify individual race management profiles in 100 m freestyle races and explain the relationships between race management strategies, performance outcomes, race type, competition level, and level of adversity.
METHODS:
Cycle-to-cycle data on swimming speed (S) in m·s⁻¹, stroke rate (SR) in Hz or cycles·min⁻¹, and stroke length (SL) in m were extracted from six major competitions using a semi-automatic race analysis system, including three French National Championships, the 2022 and 2023 World Championships, and the 2024 Paris Olympic Games. Latent Class Mixed Models (LCMM) (Proust-Lima, Philipps, & Liquet, 2015) will be used to identify distinct classes of speed trajectories for each lap of the 100 m freestyle race. Once these trajectory-based groups are defined, a second clustering step using k-means will be performed to characterize the variability of stroking parameters (S, SR, SL) within groups sharing similar speed management patterns.
RESULTS:
Latent Class Mixed Models (LCMM) identified several distinct speed trajectory classes across the 100 m freestyle races. Among these, one class exhibited stable speed with minimal variation throughout the laps (adopted by the largest number of swimmers), while another displayed J-shaped pacing characterized by a fast start, stable intermediate section, and fast finish. Higher average speeds combined with lower coefficients of variation in stroke length (SL) and stroke rate (SR) were associated with superior WA points and better overall performance. Conversely, high variability in SL and SR correlated with greater speed variability both within laps and across the entire race, indicating lower performance.
CONCLUSION:
Elite swimmers do not adhere to a single “ideal” race management strategy and several race management profiles can lead to high performance (Simbaña-Escobar et al., 2018). These insights may inform individualized race analysis and coaching strategies targeting both speed trajectories and stroking parameter variability.