A method to ensure the accuracy of estimates of anaerobic capacity derived using the critical power concept

Estimates of anaerobic work capacity (AWC) were determined in 26 subjects using the critical power concept. AWC estimates were derived using three mathematically equivalent models: nonlinear power-time, linear power-1/time, and linear work-time. We hypothesized that large variability between estimates generated using the three models might reflect systematic error in the data and that large standard errors of estimate (SEE) might reflect random error. Therefore, subjects were grouped according to the variability in the three AWC estimates and according to the magnitude of the SEE of the parameter estimates. Then AWC estimates were compared to a criterion measure of anaerobic capacity, namely maximal oxygen deficit. When variability in the AWC estimates was low (< 10%), all provided accurate measures of oxygen deficit; also, when SEE were low, estimates were accurate. When variability was high, mean AWC estimates derived using linear models differed from oxygen deficit by over 15%; when SEE were high, mean AWC estimates derived using all models differed from oxygen deficit by over 10%. It is concluded that the accuracy of estimates of anaerobic capacity derived using the critical power concept can best be ensured by accepting values only when all three models provide the same value and/or when the SEE of the AWC estimate is small.