The Performance Effect of Centralizing a Nation's Elite Swim Program

Many national sporting organizations recruit talented athletes to well-resourced centralized training squads to improve their performance. Purpose: To develop a method to monitor performance progression of swimming squads and to use this method to assess the progression of New Zealand's centralized elite swimming squad. Methods: Best annual long-course competition times of all New Zealand swimmers with at least 3 y of performances in an event between 2002 and 2013 were downloaded from takeyourmarks.com (similar to 281,000 times from similar to 8500 swimmers). A mixed linear model accounting for event, age, club, year, and elite-squad membership produced estimates of mean annual performance for 175 swim clubs and mean estimates of the deviation of swimmers' performances from their individual quadratic trajectories after they joined the elite squad. Effects were evaluated using magnitude-based inferences, with a smallest important improvement in swim time of -0.24%. Results: Before 2009, effects of elite-squad membership were mostly unclear and trivial to small in magnitude. Thereafter, both sexes showed clear additional performance enhancements, increasing from large in 2009 (males -1.4% +/- 0.8%, females -1.5% +/- 0.8%; mean +/- 90% confidence limits) to extremely large in 2013 (males -6.8% +/- 1.7%, females -9.8% +/- 2.9%). Some clubs also showed clear performance trends during the 11-y period. Conclusions: Our method of quantifying deviations from individual trends in competition performance with a mixed model showed that Swimming New Zealand's centralization strategy took several years to produce substantial performance effects. The method may also be useful for evaluating performance-enhancement strategies introduced at national or club level in other sports.