Loading fundamentals 09: Movement velocity and intent

Short answer. Trying to accelerate the bar as fast as possible improved strength and speed more than deliberately slower lifting at the same load. How far a set slows before you stop splits the outcomes: deeper sets favored size and cost jump and speed performance. Different sets in a session can carry different stopping points, because they are doing different jobs.

---

What it is

How fast the bar travels, and how fast you were trying to make it travel. Those are different things and only the second one is under your control at a fixed load.

This is the entry for velocity-based training and the dynamic effort approach. Both are built on the two variables below: the intent applied to each repetition, and how far into a set you allow velocity to decay before stopping.

The practical instrument is velocity loss: how much slower your last repetition is than your first within a set. It is measured with a tracking device and captures something repetition counts and load percentages do not. It is a rule for terminating working sets, and it does not transfer to short priming sets used before a session with a different goal. (→ Loading fundamentals 08: Warm-up)

The proposed mechanism

Rate of force development. Maximum force is how much you can eventually produce; rate of force development (RFD) is how quickly you get there. Most real expressions of strength happen in windows too short to reach maximum force, which is why RFD is a separate quality responding to separate training. It is trained by attempting to accelerate maximally.

Intent is the input. At a given load the bar reaches whatever speed it reaches, but the neural drive you sent differs with how hard you tried, and the drive is what the nervous system adapts to. This has been measured acutely: at matched load, volume and rest, maximal-intent squat protocols produced greater muscle activation, force and power during the exercise than deliberately half-speed protocols (Cornejo-Daza et al., 2024).

Fatigue erodes the fast qualities first. As a set proceeds, velocity decays. That decay proxies accumulating fatigue and also how much of the set was performed slowly. A set ending far slower than it started spent its final repetitions training something other than speed.

What has actually been tested

Intent has been isolated as a variable, twice. Two trials randomized participants to programs differing only in the concentric velocity they were told to produce: maximal intended velocity against deliberately half-maximal velocity, at the same loads, sets and repetitions.

In the bench press over six weeks, both groups improved and the maximal-intent group improved significantly more in every variable analyzed, including one-rep max and movement velocity (González-Badillo et al., 2014). The same comparison was run on the full squat in resistance-trained men (Pareja-Blanco et al., 2014).

This is the cleanest evidence here and it is about intent, not load. The bar was the same weight in both conditions. What differed was how hard participants tried to accelerate it, and that alone changed the adaptation.

Velocity loss modulates adaptation in a dose-response manner, and it splits the outcomes. A systematic review and meta-analysis of velocity loss thresholds found that velocity loss did not influence gains in muscular strength or endurance. Increases in velocity loss were associated with increases in hypertrophy (regression slope b = 0.006, 95% confidence interval, CI, 0.001 to 0.012) (Jukic et al., 2023). They negatively affected countermovement jump (b = -0.040, 95% CI -0.079 to -0.001) and velocity against submaximal loads (b = -0.018, 95% CI -0.029 to -0.006).

The same trade appears in individual trials. Fifty-one moderately trained men squatted at 55 to 70% of one-rep max with velocity loss capped at 0, 10, 20 or 40% across sixteen sessions. A significant group by time interaction for one-rep max favored the 20% group. Effect sizes (in standard deviations) across strength variables ran 0.50 to 1.70 at 0%, 0.53 to 2.34 at 10%, 0.81 to 2.72 at 20%, and 0.52 to 0.86 at 40%. Higher velocity loss was associated with greater cross-sectional area increases, without a significant interaction, alongside lower strength improvements (Sánchez-Valdepeñas et al., 2026).

Another four-threshold trial found no strength difference. Comparing four velocity loss thresholds, it found that strength, sprint and jump gains did not differ between groups. Velocity loss of 40% brought negative neuromuscular changes, and 20 to 40% produced the most hypertrophy (Pareja-Blanco et al., 2020).

The review's practical recommendation is explicitly mixed. Low to moderate velocity loss for strength and power adaptations, and if hypertrophy is also a goal, more of the prescribed sets at moderate velocity loss or more total sets at low to moderate.

Our reading

Velocity loss measures the same thing as proximity to failure, with a different instrument. (→ Loading fundamentals 02: Reps in reserve) Reps in reserve is a self-report with about a one-repetition error. Velocity loss measures the same fatigue objectively and adds what reps in reserve cannot: how much of the set was still fast.

That is what makes the split visible. A set taken deep is better for size and worse for the fast qualities (Jukic et al., 2023), and those are two measured outcomes moving in opposite directions in the same participants.

So the practical structure is not to choose one velocity loss threshold for a program. It is that different sets in a session can carry different thresholds because they are doing different jobs.

Working sets taken to moderate or high velocity loss develop size. Working sets stopped early preserve the fast qualities. Running every set at one setting gives up whichever outcome that setting costs.

Intent and velocity loss are separate variables and the evidence treats them separately. Intent is what you attempt on every repetition, and it has been isolated and found to matter. Velocity loss is how far into a set you go, and it has been isolated and found to split the outcomes. You can train with maximal intent and stop early, or with maximal intent and go deep, and those are different prescriptions.

What intent buys may not transfer everywhere. Velocity-based training oriented toward strength-speed has improved power production under load while showing limited transfer to unloaded jump performance. Adaptations appear specific to the conditions trained, which bounds what a loaded barbell set does for an unloaded expression.

Where the goals diverge

Pure hypertrophy. Higher velocity loss is associated with more of it, at a cost you may not care about. If you do not jump or sprint for anything, the fast qualities are not obviously worth protecting. (→ Loading fundamentals 07: Weekly volume)

Strength. Moderate velocity loss produced the largest one-rep max gains in one trial (Sánchez-Valdepeñas et al., 2026), but another found no strength difference across thresholds (Pareja-Blanco et al., 2020). An inverted U for strength rests on one trial.

Sport, jumping, sprinting. Low velocity loss, and the meta-analysis is unusually direct about it. High velocity loss actively degraded jump performance. (→ Loading fundamentals sub-hub)

Training without a velocity device. Most people are here. Reps in reserve covers fatigue but not whether a repetition was fast. (→ Loading fundamentals 02: Reps in reserve) The intent findings are reassuring: what those trials manipulated was instruction, not measurement.

The case against

Effect sizes here are small and the intervals sit close to zero. A coefficient of 0.006 with a lower bound of 0.001 (Jukic et al., 2023) is a real relationship and a slight one. We built a structural recommendation on relationships of that size.

The hypertrophy finding in the squat trial had no significant group by time interaction. The authors report the association and note the absence of the interaction, and we have carried both.

Velocity loss requires equipment most people do not have, and the thresholds do not transfer cleanly to estimated effort. Someone applying "20% velocity loss" by feel is doing something else.

An earlier version of this article applied the velocity loss findings to light explosive priming sets. That was an error: those trials prescribed velocity loss as a working-set rule at training loads, and their thresholds do not transfer to a short light set. The intent findings generalize more readily, and they too were produced at training loads.

The intent trials are small. Twenty and twenty-one participants, six weeks, one exercise each, from one research group.

And the exercise matters. Most of this work used squat and bench press. Whether velocity loss thresholds mean the same thing on a machine row is unexamined.

What is not known

Whether strength follows an inverted U at all, where it peaks if it does, and whether that differs by outcome, exercise and training status.

Whether the intent effect holds at light loads. Both intent trials used training loads rather than light ones, so the finding that intent matters is established in the range people train in and not in the range a priming set would use.

Whether the fast qualities eroded by high velocity loss training recover between sessions or accumulate across a block.

Where measurement error sits relative to the thresholds. On a short set, a velocity loss small enough to be device noise and one large enough to signal real fatigue are not far apart, and the trials above used much longer sets than a priming protocol would.

The practical warning

Maximal-intent work at light loads is not maximal-load work, and it is not free either. The demand shifts from peak tension to acceleration and deceleration, which loads connective tissue differently. Build into it rather than starting at full intent.