Akaike weights (<i>w</i>_<i>i</i>) representing a measure of strength of evidence for probability of best fit of the two regression lines (breakpoint model) and the single regression line, exponential and 3^rd order polynomial curve (continuous no-breakpoint models) (mean <i>w</i>_<i>i</i> ± SD (95% CI) [# of participants with better fit of the respective model compared to the two regression lines]) fitted to the gas exchange data of 15 elite Para ice hockey players following a protocol with stepwise increases in workload every 5 min while upper-body poling.

Akaike weights (wi) representing a measure of strength of evidence for probability of best fit of the two regression lines (breakpoint model) and the single regression line, exponential and 3rd order polynomial curve (continuous no-breakpoint models) (mean wi ± SD (95% CI) [# of participants with better fit of the respective model compared to the two regression lines]) fitted to the gas exchange data of 15 elite Para ice hockey players following a protocol with stepwise increases in workload every 5 min while upper-body poling.