An improved methodology for estimating critical power from mean maximal power output data

Area of Science:

• Sports Science
• Exercise Physiology
• Performance Analytics

Background:

• Critical Power (CP) and W' are key physiological metrics for endurance performance.
• Estimating CP and W' from race data is valuable for training optimization.
• Previous methods for estimating CP and W' from mean maximal power output (MMP) have limitations.

Purpose of the Study:

• To determine if CP and W' can be reliably estimated from MMP data collected during professional cycling races.
• To compare the accuracy of CP and W' estimations using fixed MMP durations versus a novel filtering method.
• To validate these estimations against laboratory-derived CP and W' values.

Main Methods:

• Collected MMP data from 13 U23 professional cyclists over 90 days of racing.
• Determined CP and W' in the field (CP_Test, W'_Test).
• Estimated CP and W' from MMP using fixed durations (CP_Fixed, W'_Fixed) and a novel filtering technique (CP_Filtered, W'_Filtered).

Main Results:

• CP_Fixed and CP_Filtered were not significantly different from CP_Test (p > 0.05).
• W'_Fixed and W'_Filtered were not significantly different from W'_Test (p > 0.05).
• Both CP_Fixed and CP_Filtered showed significant correlations with CP_Test (r = 0.872 and 0.922, respectively; p < 0.0001), with CP_Filtered providing a better estimate.

Conclusions:

• CP can be validly estimated from MMP data during cycling races using both fixed and filtered methods.
• The novel filtering method (CP_Filtered) provides a more accurate estimation of CP compared to fixed MMP durations.
• While W' estimations showed no significant difference, their correlation with laboratory values was not significant, indicating a need for further research in W' estimation from race data.