Reproducibility of pacing profiles in elite swimmers

To analyze the reproducibility of pacing in elite swimmers during competitions and to compare heats and finals within 1 event.

Finals and heats of 158 male swimmers (age 22.8 ± 2.9 y) from 29 nations were analyzed in 2 competitions (downloaded from swimrankings.net). Of these, 134 were listed in the world's top 50 in 2010; the remaining 24 were finalists of the Pan Pacific Games or European Championships. The level of both competitions for the analysis had to be at least national championships (7.7 ± 5.4 wk apart). Standard error of measurement expressed as percentage of the subject's mean score (CV) with 90% confidence limits (CL) for each 50-m split time and for total times were calculated. In addition, mixed general modeling was used to determine standard deviations between and within swimmers.

CV for total time in finals ranged between 0.8% and 1.3% (CL 0.6-2.2%). Regarding split times, 200-m freestyle showed a consistent pacing over all split times (CV 0.9-1.6%). During butterfly, backstroke, and 400-m freestyle, CVs were low in the first 3 and 7 sections, respectively (CV 0.9-1.7%), with greater variability in the last section (1.9-2.2%). In breaststroke, values were higher in all sections (CV 1.2-2.3%). Within-subject SDs for changes between laps were between 0.9% and 2.6% in all finals. Split-time variability for finals and heats ranged between 0.9% and 2.5% (CL 0.3-4.9%).

Pacing profiles are consistent between different competitions. Variability of pacing seems to be a result of the within-subject variation rather than a result of different competitions.