Reliability of force-velocity parameters obtained from linear and curvilinear regressions for the bench press and squat exercises.

This study aimed to compare the goodness of fit and the reliability of different regression models for fitting the force-velocity relationship (FV) of bench press (BP) and squat (SQ). Additionally, the reliability of the position on FV of the velocity (V1RM) and the force performed with the 1RM (F1RM) was explored. Nine rugby players and 12 judokas participated in this study. The FV of BP and SQ were obtained twice by a protocol performed until the 1RM. Individual FV were fitted by linear (LM), quadratic polynomial (PM), and exponential models (EM). Adjusted coefficients of determination of LM and PM (medians higher than 0.919) were higher than for EM. The reliability was higher for LM in comparison with PM. The reliability of V1RM was not acceptable (CV% = 19 and 18% for BP and SQ). High reliability was observed for F1RM (CV% = 3 and 2% for BP and SQ) and for the ratio between F1RM and the force-axis intercept of FV (CV% = 2 and 4% for BP and SQ). The reliability of the relative values of F1RM around 92 and 87% of F0 for BP and SQ suggests the use of these values for monitoring resistance training programmes.