Crank Length, Biomechanics, and the Limits of Reductionist Truths
As a former biomechanics researcher working in academic labs, I have deep respect for the contributions Dr. Jim Martin has made to cycling science. His work has shaped how we think about power, cadence, and mechanical efficiency, and it has brought a much-needed rigor to a field that was once dominated by anecdote.
That respect is precisely why his conclusions on crank length deserve careful scrutiny. Because while his analysis is mechanically sound, it is biologically incomplete—and that distinction matters far more to real riders than most debates in cycling science.
What Dr. Martin’s Approach Gets Right
At its core, Dr. Martin’s work—like much of classical biomechanics—is grounded in rigid-body mechanics and clean mathematical relationships. The framework assumes steady-state conditions where power is the product of torque and angular velocity, joint kinematics are prescribed or normalized, crank angular velocity is constant, and muscle force is treated implicitly rather than explicitly. Under those assumptions, the conclusions are logical and internally consistent.
Shorter cranks reduce peak joint moments. Gross efficiency does not change meaningfully at steady state. Power output can be preserved across a range of crank lengths by adjusting cadence. None of this is wrong in a physics sense. In fact, within the boundaries of the model, it’s difficult to argue otherwise. The problem isn’t the math, it’s the model.
Moving cleats rearward for better riding
The Core Limitation: Humans Are Not Motors
The most important limitation of this approach is subtle but profound: it treats the rider as a motor, not a control system. Real riders are not torque-limited in isolation. They are limited by coordination, comfort, force–velocity constraints, joint geometry, neuromuscular patterning, and fatigue management. Dr. Martin’s framework largely presumes that the rider can always re-optimize cadence and force production to suit the mechanical demands imposed by a given crank length. In practice, that assumption often fails.
Muscle force production is history-dependent. Force–length and force–velocity relationships matter. Motor patterns are learned, reinforced, and resistant to change. Joint constraints—particularly at the hip, knee, and ankle—are not trivial variables that can be normalized away. So while the equations say, “There is no measurable difference,” the body frequently responds, “Wrong.”
Where Crank Length Actually Changes the System
Changing crank length does not simply scale torque requirements. It reshapes the entire movement problem. Shorter cranks reduce hip flexion at top dead center, which can improve breathing mechanics and pelvic stability for many riders. They also increase knee extension velocity, which pushes the quadriceps into different regions of the force–velocity curve. At the ankle, changes in crank radius alter effective moment arms and how force is transferred through the foot–pedal interface.
These are not second-order effects for many riders. They directly influence how force is produced, sustained, and tolerated. When crank length is treated as interchangeable, these interactions disappear—not because they don’t exist, but because the model cannot see them.
Hip flexor demands and posture
Tangential Force Is Not Muscle Force
One of the quiet assumptions in crank-length analysis is that tangential pedal force maps cleanly back to muscle force via simple geometry. But the human body does not apply force to the pedal through a single actuator.
Force emerges from multi-joint muscle synergies, biarticular muscles that redistribute work across joints, and moment arms that change continuously throughout the pedal stroke. Altering crank length reshapes recruitment patterns, timing, and joint loading—not just average torque. Working backward from pedal forces hides the biological cost of producing them.
The Fiction of Uniform Angular Velocity
Another simplification that matters more than it appears is the assumption of uniform angular velocity. Real pedaling is not smooth. It contains torque ripple, dead spots, and phase-dependent muscle activation. Crank length alters when peak force is required relative to muscle capability and joint position. That timing effect—critical for fatigue and comfort—is invisible in steady-state average models.
This is why riders often report something that data struggles to explain: “The power is there… but it feels worse.” They’re not imagining it.
Equivalent Power Is Not Equivalent Fatigue
Even when average power output is preserved across crank lengths, the internal cost of producing that power may not be. Peak muscle forces can increase. Contraction velocities may shift into less favorable regions. Tendon loading patterns can change. Over time, these differences accumulate—not in watts, but in fatigue, discomfort, and injury risk.
Laboratory models are excellent at describing what is possible. They are far less effective at describing what is sustainable.
The More Honest Conclusion
A more accurate reading of the evidence would be this: Crank length effects are small in tightly controlled laboratory outputs. They become meaningful through comfort, repeatability, positional constraints, and long-term adaptation. Crank length interacts strongly with saddle height, setback, hip angle, breathing mechanics, and riding discipline. This is why some riders thrive on shorter cranks, others lose their sense of torque, and blanket statements like “crank length doesn’t matter” ring hollow in practice.
Dr. Martin’s analysis is mechanically correct. It is also biologically incomplete. And when we are fitting bikes for humans—not mannequins—that distinction matters.
What the Crank-Length Literature Actually Shows
Across decades of instrumented-crank and force-pedal studies—by Martin, Sanderson, Hull, Neptune, and others—a consistent pattern emerges when cadence and power are held constant. And frankly, I cannot believe this obvious finding isn’t more openly discussed. That said, when bike fitters and performance experts only read the abstract, these data do get missed…
Longer cranks increase the magnitude of torque throughout the pedal cycle. Shorter cranks reduce it. This applies not only to peak positive torque during the power phase, but also to peak negative torque during the transition and recovery phases, as well as to overall torque ripple. This finding is not interpretive. It follows directly from basic mechanics. When average power and cadence are fixed, increasing crank length requires higher average torque to achieve the same external output. Because torque is not applied uniformly across the pedal stroke, the entire torque waveform scales: peaks rise, valleys deepen, and negative torque becomes more pronounced. Shorter cranks compress the waveform toward zero. This is exactly what is observed in crank-based torque traces.
The increase in negative torque with longer cranks is often misunderstood. It is not evidence of poor technique or a failure to “pedal circles.” It arises from limb dynamics. Longer cranks increase joint excursions and distal segment velocity, which in turn increases the inertial and gravitational demands of accelerating and decelerating the leg near top and bottom dead center. Antagonist activity and braking moments rise accordingly.
Importantly, allowing cadence to vary does not eliminate this effect. Riders on longer cranks often self-select slightly lower cadence, and riders on shorter cranks slightly higher cadence, which partially offsets torque scaling. But the underlying trend remains: longer cranks produce larger torque magnitudes across the cycle, even if the difference is muted by cadence adaptation.
What the literature does not claim is equally important. Negative torque is normal. It exists even in elite riders and can coexist with high efficiency. The question is not whether negative torque is present, but how much, where in the cycle it occurs, and at what metabolic and neuromuscular cost.
This is where average-power framing can mislead. Analyses that focus on mean power, mean efficiency, or cadence compensation can conclude that crank length “doesn’t matter,” while obscuring changes in peak muscle force, joint loading, fatigue, comfort, and repeatability over time. Those internal costs do not always show up in steady-state averages—but riders feel them. Furthermore, no such research exists to explore/understand the adaptive physiology resulting from a longitudinal study on crank length.
In that sense, the literature and lived experience are not in conflict. They are answering different questions.
A Soft-Tissue Perspective: Hip Flexion, the Spine, and the Cost of Long Levers
From the perspective of a bodyworker and functional movement explorer, crank length shows up in the body long before it shows up in power data.
Longer cranks increase hip flexion at top dead center. That is mechanically obvious—but the downstream consequences are often overlooked. Increased hip flexion demands greater lengthening and control of the hip flexor complex, particularly the iliopsoas, under load and repetition.
In many riders, that demand does not stay localized at the hip.
When the hip flexors cannot accommodate repeated deep flexion cleanly—because of tissue tone, fatigue, prior injury, or simple anatomy—the nervous system looks for stability elsewhere. Very often, that stability is borrowed from the lumbar spine and pelvis.
What shows up clinically is familiar:
Increased anterior pelvic tilt under load
Lumbar extension bias near top dead center
Asymmetrical pelvic motion side to side
Guarding through the psoas and rectus femoris
Elevated tone around the sacroiliac joint
None of this requires pathology. It requires repetition. Longer cranks increase the angular excursion of the femur relative to the pelvis. They increase the braking and re-acceleration demands at the top of the stroke. Over thousands of cycles, this can translate into increased shear and rotational stress across the lumbo-pelvic complex—particularly in riders who lack reserve hip mobility or who are already managing life stress, sitting demands, or prior low-back episodes. Shorter cranks quietly change that equation.
By reducing peak hip flexion and smoothing the transition through top dead center, they reduce the tonic demand placed on the hip flexors. The pelvis is asked to stabilize rather than compensate. The lumbar spine is less frequently recruited as a control surface. The sacroiliac joints experience less repetitive torsion driven by asymmetrical limb braking.
From a soft-tissue standpoint, shorter cranks often shows up as:
Reduced tone and reactivity in the hip flexors
Improved pelvic symmetry under load
Decreased low-back irritation over longer rides
Greater tolerance to sustained efforts and fatigue
Importantly, this does not mean longer cranks cause back pain, nor that shorter cranks treat it. It means crank length influences how stress is distributed through the system—and for many riders, shorter cranks distribute that stress more favorably.
This is one of the reasons shorter cranks so often “feel better” before they ever feel faster. They reduce the amount of protective tension the body needs to generate just to keep riding. And over time, that matters.
Pelvic asymmetry and influencing mechanics
Closing: Why Shorter Cranks Keep Winning in Practice
Taken at face value, much of the crank-length literature appears neutral. Power can be preserved. Efficiency doesn’t collapse. Average outputs remain similar. From a purely mechanical standpoint, it’s tempting to conclude that crank length simply doesn’t matter. But that conclusion only holds if we care exclusively about averages.
When we zoom out to include joint kinematics, force–velocity constraints, fatigue, comfort, and repeatability over time, a pattern emerges—one that many riders and practitioners recognize immediately. Shorter cranks consistently reduce peak torque demands. They compress the torque waveform, limit excessive negative torque, and reduce extreme joint excursions at the hip and knee. They tend to place muscles in more favorable regions of their force–velocity and force–length relationships, particularly at higher cadences. For many riders, they also open hip angle, improve breathing mechanics, and reduce compensatory movement under fatigue.
None of this guarantees more power. What it often guarantees is less cost. And in endurance sport, cost matters.
The question is not whether a rider can produce the same power on longer cranks. Many can. The question is whether they can do so comfortably, repeatedly, and without accumulating unnecessary fatigue or injury risk. For a large number of riders—especially those balancing training, work, age, prior injury, or limited recovery—shorter cranks make that equation easier, not harder. This is why shorter cranks so often “work” in the real world, even when the lab suggests neutrality. They are not magic. They are simply more forgiving.
Crank length doesn’t determine performance. But it strongly influences how hard performance is to sustain. And when all else is equal, the option that lowers peak forces, expands positional freedom, and reduces biological cost tends to win—not on paper, but over months and years of riding.
That’s why, in practice, shorter cranks deserve more credit than they’re usually given. Not because they make riders stronger. But because they make strength easier to express, day after day.