Reading the Concentric:Eccentric Velocity Ratio for Fatigue: What a Rising E:C Number Actually Means

Rep six of a top squat set looks identical to rep one on the raw velocity trace -- same depth, same bar path -- but the two numbers underneath it have already split apart. Concentric velocity has fallen from 0.58 m/s on rep one to 0.41 m/s on rep six, a 29 percent drop most coaches would flag immediately. What almost nobody checks is what happened on the way down: eccentric velocity on that same rep six is 0.93 m/s, barely different from rep one's 0.87 m/s. The lifter isn't only fatiguing on the way up. The gap between how fast the bar falls and how fast it comes back up has widened, and that gap -- the concentric:eccentric velocity ratio, or C:E ratio -- is often a cleaner fatigue signal than either velocity number read in isolation.

Most velocity-based training protocols track exactly one number: concentric mean velocity, compared rep to rep against a first-rep reference. That approach, built out in the velocity-loss literature, works well enough to be the industry default. But it treats a barbell rep as if only half of it existed. The eccentric phase carries its own signal, and the relationship between the two phases -- not the concentric decline by itself -- flags a specific and useful failure mode: the loss of eccentric braking control, often before anything else on the monitor changes.

The C:E ratio is simple arithmetic: eccentric mean velocity (EMV) divided by concentric mean velocity (CMV) for the same repetition. A squat rep with an EMV of 0.87 m/s and a CMV of 0.58 m/s produces a ratio of 1.5. Track that number rep by rep across a set the same way a velocity-loss protocol tracks CMV alone, and a second, largely independent fatigue signal appears.

The baseline value is not 1.0, and that surprises most people the first time they see their own numbers. In a free-tempo squat or bench press, gravity accelerates the bar on the way down while the lifter's own force output has to overcome gravity plus the bar's inertia on the way up. Descending is mechanically easier than lifting, so EMV usually exceeds CMV well before any fatigue enters the picture. A ratio of 1.4 to 1.8 on a fresh rep is completely normal for a back squat with no imposed eccentric tempo.

That baseline moves with the lift and the tempo rule in play, so it has to be read relative to itself, not against a universal number.

Every lifter should establish their own fresh-rep ratio for each lift before trying to interpret drift within a set. A lifter whose squat opens at 1.3 and one whose squat opens at 1.7 are both normal; what matters from here is how far each one moves from their own number.

Two separate physiological effects push the ratio in the same direction as a set progresses, which is part of why it tends to move faster and more visibly than concentric velocity loss alone.

The first is well documented: concentric muscle actions fatigue faster than eccentric ones at matched effort. Tesch et al. (1990) had subjects perform repeated maximal-effort knee extensions on an isokinetic dynamometer, either as concentric-only bouts or eccentric-only bouts at matched angular velocity. Peak torque fell substantially further across the repeated concentric bouts than across the repeated eccentric bouts at equivalent volume -- the concentric action simply runs out of usable force output sooner. In a barbell set, this shows up as the CMV decline that velocity-loss protocols already track.

The second effect gets almost no attention: eccentric braking is an active, fatiguable skill, not a passive fall. Slowing a loaded bar on the way down requires the same muscles to generate controlled tension against a load that is already accelerating under gravity. As fatigue erodes that braking capacity, the bar does not slow down -- it speeds up, because less muscular force is available to resist the descent. So on the same rep where CMV is falling, EMV frequently holds steady or climbs, and the ratio between them moves in both directions at once.

Chappell et al. (2005) documented a related pattern outside the weight room: after a fatiguing protocol, recreational athletes showed reduced eccentric control at the knee during landing tasks, with increased valgus loading and less controlled deceleration. The mechanism there is landing, not a barbell rep, but the underlying finding -- that fatigue degrades eccentric braking specifically, not just concentric output -- is the same phenomenon the C:E ratio is built to catch in a lifting set.

No published study has measured the C:E ratio as a single tracked metric across a fatiguing barbell set -- that gap is exactly why this interpretation leans on combining findings rather than citing one definitive source. The table below lays out what each study actually measured and where the extrapolation to a live C:E ratio starts.

Sánchez-Medina and González-Badillo (2011) remain the anchor for the concentric side of this equation. Fifty-six trained lifters performed bench press and full squat sets across a range of intensities to varying degrees of velocity loss, with mean concentric velocity recorded on every rep by a linear position transducer. Velocity loss tracked the percentage of the maximum possible reps completed almost linearly, and blood lactate and ammonia rose sharply once velocity loss passed roughly 30 to 40 percent. None of that work touched the eccentric phase, which is the piece the C:E ratio adds.

Step 1: Capture a fresh baseline. Average the ratio across the first two reps of the first working set, before any meaningful fatigue has accumulated. That number, not 1.0, is the athlete's reference point for the session.

Step 2: Set a working flag threshold. A ratio that climbs to roughly 1.4 times the fresh baseline is a reasonable default trigger for a stop-set conversation -- for a squat opening at 1.5, that means flagging around 2.1. Dead-stop lifts like conventional deadlift, which open closer to 1.0, are more sensitive to the same absolute rise, so drop the multiplier to roughly 1.25 for those.

Step 3: Hold the tempo instruction constant. If the coaching cue changes mid-set -- telling the athlete to control the descent more on rep 5 than rep 1 -- the ratio will drop on its own, and that drop has nothing to do with fatigue. The ratio is only a valid fatigue signal when the eccentric tempo instruction does not change within the set.

Step 4: Log week-to-week drift in the fresh-rep baseline itself, the same way a velocity-loss program logs first-rep CMV at a constant load. A baseline ratio creeping upward across weeks at the same load and load percentage, even before within-set fatigue is considered, points to accumulating between-session fatigue rather than anything happening inside a single set.

Step 5: Cross-check equipment placement. A ratio computed from a bar-mounted sensor and one computed from a body-mounted sensor will not agree, because bar path and body path diverge slightly through the rep. Keep the sensor position consistent session to session, or the drift in the number may be measuring a mounting change, not the athlete.

A small rise in the ratio across a working set is expected and not a concern on its own. A squat set opening at 1.5 and closing its last clean rep at 1.7 or 1.8 is ordinary within-set fatigue, the same low-grade decline that a 10 to 15 percent velocity loss represents on the concentric side alone.

The flag worth acting on is a rise that outpaces the concentric velocity loss happening in parallel. If CMV has fallen 15 percent but the ratio has jumped 50 percent above baseline, the eccentric side is doing more of the work of that change than concentric decline alone would explain -- the athlete is losing braking control faster than they are losing propulsive output, which is the pattern most associated with technical breakdown and injury risk, not simple accumulated fatigue.

The two signals read together tell a more complete story than either alone. Concentric velocity loss without much ratio movement usually just means the athlete is getting tired in the ordinary sense. Ratio movement that outpaces concentric loss means the failure mode is shifting toward control, not just output, and that calls for ending the set even if a VL30-style cutoff has not technically been reached yet.

The most common error is comparing ratios across different loads without accounting for load itself. A 60 percent 1RM set and an 85 percent 1RM set on the same lift will not open at the same ratio, because the concentric side is fighting more resistance at the heavier load even before fatigue is a factor. Compare drift within a set and within a load, not the raw number across different days at different intensities.

The second is ignoring a cueing change disguised as fatigue. A lifter who consciously slows the eccentric on the final two reps of a set -- a common instinctive safety response near failure -- will show a falling ratio exactly when fatigue is highest, which looks like the opposite of what the metric is supposed to show. Ask what the athlete was told to do, or watched themselves do, before reading a late-set ratio drop as good news.

The third is treating the ratio as a replacement for velocity loss rather than a companion to it. The two metrics catch different failure modes -- concentric decline and eccentric control loss are related but not identical processes -- and a program that drops VL tracking in favor of the ratio alone will miss straightforward propulsive fatigue that never touches the eccentric side much at all.

1. Sánchez-Medina, L., & González-Badillo, J.J. (2011). Velocity loss as an indicator of neuromuscular fatigue during resistance training. Medicine & Science in Sports & Exercise, 43(9), 1725-1734.
2. Tesch, P.A., Dudley, G.A., Duvoisin, M.R., Hather, B.M., & Harris, R.T. (1990). Force and EMG signal patterns during repeated bouts of concentric or eccentric muscle actions. Acta Physiologica Scandinavica, 138(3), 263-271.
3. Chappell, J.D., Herman, D.C., Knight, B.S., Kirkendall, D.T., Garrett, W.E., & Yu, B. (2005). Effect of fatigue on knee kinetics and kinematics in stop-jump tasks. American Journal of Sports Medicine, 33(7), 1022-1029.