Critical-Speed Modeling Using Familiar Performance Tests Can Accurately Predict 1500-m Rowing-Ergometer Performance Capability.

Abstract

Purpose: To assess the accuracy of a linear critical-speed model for predicting maximal rowing-ergometer performance.

Methods: Twenty (5 female) highly trained rowers completed 5 ergometer time trials (TTs). The linear distance-time relationship between the 2000-m and the fastest of two 500-m TTs was used to predict 1500-m performance. Two time-to-exhaustion (TTE) trials, with intensity targets set to the predicted 1500-m performance capacity, were subsequently performed. The accuracy of the critical-speed model was assessed by comparing predicted and actual distances in each athlete's longest TTE trial, using absolute and percentage differences, standard error of estimate, and coefficient of variation.

Results: The mean difference in 500-m TT completion time was 1.2 (1.0) seconds (1.3% [1.1%]). Athletes maintained 98.6% (0.9%) of target power output during the TTE trials, covering an average distance of 1587 (169) m. The predicted distance was 1589 (162) m. The mean difference between predicted and actual distances covered was 7.6 (6.6) m (0.4% [0.4%]), with a near-perfect association (R2 = .99). The standard error of estimate was 8.3 m, and the coefficient of variation was 0.3% between the modeled and actual TTE trial distances.

Conclusions: The linear critical-speed model, based on 2 familiar and reliable performance tests, accurately predicts maximum rowing performance. These outcomes allow continued focus on 2000-m performance preparation while indirectly tracking 1500-m performance progression. In addition, this model allows for understanding of the influence that improvement in either 2000- or 500-m TT times has on predicted 1500-m performance.