For more than a century, many concepts and several theories and principles pertaining to the goals, organization, methodology and evaluation of the effects of resistance training (RT) have been developed and discussed between coaches and scientists. This cumulative body of knowledge and practices has contributed substantially to the evolution of RT methodology. However, a detailed and rigorous examination of the existing literature reveals many inconsistencies that, unless resolved, could seriously hinder further progress in our field. The purpose of this review is to constructively expose, analyze and discuss a set of anomalies present in the current RT methodology, including: (a) the often inappropriate and misleading terminology used, (b) the need to clarify the aims of RT, (c) the very concept of maximal strength, (d) the control and monitoring of the resistance exercise dose, (e) the existing programming models and (f) the evaluation of training effects. A thorough and unbiased examination of these deficiencies could well lead to the adoption of a revised paradigm for RT. This new paradigm must guarantee a precise knowledge of the loads being applied, the effort they involve and their effects. To the best of our knowledge, currently this can only be achieved by monitoring repetition velocity during training. The main contribution of a velocity-based RT approach is that it provides the necessary information to know the actual training loads that induce a specific effect in each athlete. The correct adoption of this revised paradigm will provide coaches and strength and conditioning professionals with accurate and objective information concerning the applied load (relative load, level of effort and training effect). This knowledge is essential to make rational and informed decisions and to improve the training methodology itself.

This study aimed to compare the effects of 2 resistance training (RT) programs with different velocity losses (VLs) allowed in each set: 10% (VL10%) versus 30% (VL30%) on neuromuscular performance and hormonal response. Twenty-five young healthy males were randomly assigned into 2 groups: VL10% (n = 12) or VL30% (n = 13). Subjects followed a velocity-based RT program for 8 weeks (2 sessions per week) using only the full-squat (SQ) exercise at 70%–85% 1-repetition maximum (1RM). Repetition velocity was recorded in all training sessions. A 20-m running sprint, countermovement jump (CMJ), 1RM, muscle endurance, and electromyogram (EMG) during SQ exercise and resting hormonal concentrations were assessed before and after the RT program. Both groups showed similar improvements in muscle strength and endurance variables (VL10%: 7.0%–74.8%; VL30%: 4.2%–73.2%). The VL10% resulted in greater percentage increments in CMJ (9.2% vs. 5.4%) and sprint performance (–1.5% vs. 0.4%) than VL30%, despite VL10% performing less than half of the repetitions than VL30% during RT. In addition, only VL10% showed slight increments in EMG variables, whereas no significant changes in resting hormonal concentrations were observed. Therefore, our results suggest that velocity losses in the set as low as 10% are enough to achieve significant improvements in neuromuscular performance, which means greater efficiency during RT. Novelty The VL10% group showed similar or even greater percentage of changes in physical performance compared with VL30%. No significant changes in resting hormonal concentrations were observed for any training group. Curvilinear relationships between percentage VL in the set and changes in strength and CMJ performance were observed.

The aim of this study was to compare the time course of recovery following four different resistance exercise protocols in terms of loading magnitude (60% vs. 80% 1RM—one-repetition maximum) and velocity loss in the set (20% vs. 40%). Seventeen males performed four different protocols in full squat exercise, which were as follows: (1) 60% 1RM with a velocity loss of 20% (60-20), (2) 60% 1RM with a velocity loss of 40% (60-40), (3) 80% 1RM with a velocity loss of 20% (80-20), and (4) 80% 1RM with a velocity loss of 40% (80-40). Movement velocity against the load that elicited a 1 m·s⁻¹ velocity at baseline measurements (V₁-load), countermovement jump (CMJ) height, and sprint time at 20 m (T20) were assessed at Pre, Post, 6 h-Post, 24 h-Post, and 48 h-Post. Impairments in V₁-load were significantly higher for 60-40 than other protocols at Post (p < 0.05). The 60-20 and 80-40 protocols exhibited significant performance impairments for V₁-load at 6 h-Post and 24 h-Post, respectively (p < 0.05). CMJ height remained decreased for 60-20 and 60-40 until 24 h-Post (p < 0.001–0.05). Regarding T20, the 80-40 protocol resulted in higher performance than 60-40 at 24 h-Post and the 80-20 protocol induced a greater performance than 60-40 protocol at 48 h-Post (p < 0.05). A higher velocity loss during the set (40%) and a lower relative load (60% 1RM) resulted in greater fatigue and slower rate of recovery than lower velocity loss (20%) and higher relative load (80% 1RM).

The aim of this study was to compare the effects of two velocity-based resistance training (RT) programs using moderate loads (45–60% 1RM) but different magnitudes of velocity loss (VL) limits (10% vs. 30%) on the changes in physical performance in young soccer players. Twenty young soccer players were randomly allocated into two groups: VL10% (n = 10) and VL30% (n = 10). All participants were assessed before and after the 8-week RT program (twice a week) involving the following tests: 20 m running sprint (T20), countermovement jump (CMJ), kicking a ball (KB), and progressive loading test in the full squat (SQ) exercise. The RT program was conducted using only the SQ exercise and movement velocity was monitored in all repetitions. Significant ‘time × group’ interaction (p < 0.05) was observed for sprint performance, KB and 1RM in the SQ exercise in favor of VL10%. No significant changes between groups at post-test were observed. The VL10% resulted in significant (p < 0.05–0.001) intra-group changes in all variables analyzed, except for KB, whereas VL30% only showed significant (p < 0.05) performance increments in a sprint test and 1RM in the SQ exercise. The percentage of change and the intra-group’s effect size were of greater magnitude for VL10% in all variables analyzed compared to VL30%. In conclusion, our results suggest that, for non-trained young soccer players, squat training with low to moderate relative loads and 10%VL is sufficient to elicit significant increases in muscle strength and sport-specific actions compared to 30%VL in the set.

This study aimed to analyze the intra-device agreement of a new linear position transducer (Vitruve, VT) and the inter-device agreement with a previously validated linear velocity transducer (T-Force System, TF) in different range of velocities. A group of 50 healthy, physically active men performed a progressive loading test during a bench press (BP) and full-squat (SQ) exercise with a simultaneous recording of two VT and one TF devices. The mean propulsive velocity (MPV) and peak of velocity (PV) were recorded for subsequent analysis. A set of statistics was used to determine the degree of agreement (Intraclass correlation coefficient [ICC], Lin’s concordance correlation coefficient [CCC], mean square deviation [MSD], and variance of the difference between measurements [VMD]) and the error magnitude (standard error of measurement [SEM], smallest detectable change [SDC], and maximum errors [ME]) between devices. The established velocity ranges were as follows: >1.20 m·s⁻¹; 1.20–0.95 m·s⁻¹; 0.95–0.70 m·s⁻¹; 0.70–0.45 m·s⁻¹; ≤0.45 m·s⁻¹ for BP; and >1.50 m·s⁻¹; 1.50–1.25 m·s⁻¹; 1.25–1.00 m·s⁻¹; 1.00–0.75 m·s⁻¹; and ≤0.75 m·s⁻¹ for SQ. For the MPV, the VT system showed high intra- and inter-device agreement and moderate error magnitude with pooled data in both exercises. However, the level of agreement decreased (ICC: 0.790–0.996; CCC: 0.663–0.992) and the error increased (ME: 2.8–13.4% 1RM; SEM: 0.035–0.01 m·s⁻¹) as the velocity range increased. For the PV, the magnitude of error was very high in both exercises. In conclusion, our results suggest that the VT system should only be used at MPVs below 0.45 m·s⁻¹ for BP and 0.75 m·s⁻¹ for SQ in order to obtain an accurate and reliable measurement, preferably using the MPV variable instead of the PV. Therefore, it appears that the VT system may not be appropriate for objectively monitoring resistance training and assessing strength performance along the entire spectrum of load-velocity curve.