Aim: The aim of this study was to examine the separate and combined effects of several factors deemed important in determining peaking in the Athens Olympic swimming competition.
Methods: Data on 301 Olympic swimmers, who competed in 424 events, were used to analyze the relative swimming performance gain (RSPG%: the relative differences between the entry swimming results obtained in national trials or another competition, and in the Olympic competition expressed as a percentage). RSPG% was described with respect to 24 National, stroke-type, swimming distance, swimmer's rank, gender, mode of selection and duration of the final stage preparation (FSP) prior the Olympics. Analysis of variance (ANOVA) and cluster linear regression were performed to estimate the effect of these factors on RSPG%.
Results: The average RSPG% gain equaled 0.58% (standard deviation: 1.13%) indicating performance decline, embracing 68.2% of all the swimming events. However, two Olympians' categories, medalists and swimmers ranked 4-8, broke their personal best on average by 0.35% and 0.12%, respectively. One-way ANOVA considering nations with ''rigid selection'' vs ''liberal selection'', revealed significant superiority (P=0.04) of the swimmers who were selected rigorously over swimmers selected liberally.
Conclusion: The results indicate the marked tendency of performance decline in all the swimming events and strokes during the FSP. This general tendency was not representative of swimmers who reached the finals. Furthermore, a highly significant superiority of peaking in the teams was revealed, which required ''rigid selection'' rather than ''liberal selection''.