Running Performance can be Predicted by Log v-Log t-Models based on Non-Exhaustive Tests and Rating of Perceived Exertion
Individual running performances at different distances can be described by log v-log t-models. The study tests, if rating of perceived exertion while submaximal graded exercise (GXT) and constant load test (CLT) can be used for parameter estimation of log-log-models so that running performance can be predicted. Under laboratory and field conditions 13 and 12 subjects performed a submaximal GXT and CLT, where RPE was measured using RPE15-Scale (Borg, 2004) to calculate 3000m running-performance. Finally, maximal 3000m track runs were done for model verification. Predicted and measured running speed were highly correlated (r = .92 vs. r = .96, p < .001). In the laboratory study predicted speeds were significant lower than measured speeds, whereas by field tests predicted values didn´t differ from measured ones. It is concluded that the presented approach offers an accurate, economic and non-exhaustive possibility of performance prediction in running.