Loading field map
How do the laws of physics shape athletic performance, and how can understanding those laws help athletes move faster, jump higher, or avoid injury? Sports biomechanics is the branch of sports science that asks these questions by applying mechanical principles to human movement. Over the past 140 years, the subfield has evolved through a series of methodological paradigms, each driven by new measurement technologies and computational tools. These paradigms have not simply replaced one another; they have layered on top of each other, with older approaches continuing to serve specific purposes while newer ones open up questions that earlier methods could not address.
The earliest systematic approach to sports biomechanics was the Descriptive Kinematics Paradigm. Researchers in this period focused on describing what happens during movement—the positions, angles, velocities, and accelerations of body segments—without yet measuring the forces that cause those motions. The key tools were photography and film. Étienne-Jules Marey’s chronophotography and Eadweard Muybridge’s sequential photographs of athletes in motion allowed scientists to break down complex actions like running or throwing into a series of discrete positions. This paradigm was purely observational: it could tell a coach that a sprinter’s knee angle at toe-off was 150 degrees, but it could not explain why that angle produced a certain speed. Its lasting contribution was to establish that human movement could be quantified and analyzed frame by frame, a principle that every later paradigm would inherit. The Descriptive Kinematics Paradigm coexisted with early work in exercise physiology, but it remained methodologically separate, focused on motion capture rather than on metabolic or muscular limits.
By the mid-20th century, researchers wanted to move beyond description to explanation. The Kinetic Analysis Paradigm shifted attention from kinematics (motion) to kinetics (forces). The invention of the force plate—a device that measures ground reaction forces when an athlete jumps, runs, or lands—was the technological breakthrough that defined this era. For the first time, biomechanists could calculate the net joint torques and powers that actually produce movement, using inverse dynamics. This paradigm absorbed the earlier kinematic methods: you still needed to film the motion, but now you also recorded the forces underneath it. The Kinetic Analysis Paradigm transformed coaching and injury prevention. For example, by measuring the high impact forces during landing, researchers could identify movement patterns that put athletes at risk for anterior cruciate ligament (ACL) injuries. This paradigm also created the first strong bridge to strength and conditioning science, because force-plate data could directly inform how athletes trained for power and explosiveness. However, kinetics still treated the body largely as a system of rigid segments connected by hinges; it did not yet account for the muscles and tendons that generate those forces internally.
Overlapping with the rise of kinetic analysis, the Electromyographic Analysis Paradigm opened a window into the muscles themselves. Electromyography (EMG) records the electrical activity produced by muscle fibers during contraction, allowing researchers to see which muscles are active, when they turn on, and how hard they are working. This paradigm did not replace kinetic analysis; instead, it complemented it. A force plate could tell you the net torque at the knee, but EMG could tell you whether the quadriceps or hamstrings were primarily responsible. The two methods were often used together, creating a richer picture of movement control. The EMG paradigm also connected sports biomechanics to motor learning and skill acquisition, because it revealed how muscle coordination patterns change as athletes learn new techniques or fatigue. By the 1980s, researchers could synchronize video, force plates, and EMG, producing multi-channel data streams that described a movement from three different angles simultaneously. Yet EMG had limitations: it could not measure forces inside the muscle or predict how a change in one muscle’s activation would affect the whole system.
The Musculoskeletal Modeling and Simulation Paradigm emerged as a response to those limitations. Instead of only measuring what the body does, this paradigm builds computational models of the skeleton, joints, muscles, and tendons, then simulates how they interact. A researcher can adjust a model’s muscle strength, tendon stiffness, or activation timing and see how the simulated movement changes—something impossible with live human subjects. This paradigm absorbed the data from kinematics, kinetics, and EMG as inputs for model validation, but it went further by allowing scientists to ask “what if” questions: What if an athlete’s Achilles tendon were 10% stiffer? What if a runner changed their foot-strike pattern? The models also enabled estimation of quantities that cannot be measured directly, such as individual muscle forces or joint contact pressures. This paradigm has become a leading framework in contemporary sports biomechanics because it integrates all earlier methods into a single predictive system. It is now widely used in injury prevention (simulating how ACL loading changes with landing technique), equipment design (optimizing running shoes or prosthetics), and rehabilitation (predicting how surgery or bracing alters movement). The Musculoskeletal Modeling and Simulation Paradigm also created a direct link to exercise physiology, because models can incorporate muscle energetics and fatigue, and to performance analysis, because simulations can be run on thousands of virtual trials to find optimal techniques.
The most recent paradigm, the Multiscale and Computational Biomechanics Paradigm, extends modeling beyond the whole-body level. While musculoskeletal models treat muscles as single units with lumped properties, this newer paradigm asks how phenomena at smaller scales—muscle fiber types, tendon collagen structure, bone material properties—affect whole-body performance and injury risk. It also incorporates larger scales, such as the interaction between an athlete and their environment (e.g., turf, footwear, or a ball). The key innovation is computational power: finite element analysis, agent-based models, and machine learning allow researchers to simulate movement across multiple levels simultaneously. For example, a multiscale model of sprinting might include a finite-element model of the foot–shoe interface, a musculoskeletal model of the leg, and a neural control model of muscle activation, all running together. This paradigm has transformed sports biomechanics from a purely analytical field into a predictive engineering science. It is the leading framework today because it can address questions that earlier paradigms could not: How does a change in shoe midsole foam at the millimeter scale alter an athlete’s 100-meter time? How does fatigue at the cellular level propagate up to change running gait? The Multiscale and Computational Biomechanics Paradigm also serves as a bridge to other sports science frameworks. It shares data and models with exercise physiology (muscle energetics), motor learning (neural control strategies), performance analysis (whole-team movement patterns), and strength and conditioning (training load optimization).
Today, the Musculoskeletal Modeling and Simulation Paradigm and the Multiscale and Computational Biomechanics Paradigm are the leading frameworks, but they do not stand alone. The older paradigms remain active in specific roles. Descriptive kinematics is still used in coaching and broadcast analysis, where quick visual feedback matters more than force data. Kinetic analysis with force plates remains the gold standard for assessing jump performance and balance in applied settings. EMG continues to be essential for studying muscle coordination in rehabilitation and skill acquisition. The two modeling paradigms have not rendered these older methods obsolete; rather, they have absorbed their data and expanded their scope.
What the leading frameworks agree on is that movement emerges from interacting mechanical and biological systems, and that computational models are the best way to understand those interactions. They disagree on the necessary level of detail. Some researchers argue that whole-body musculoskeletal models are sufficient for most practical questions (e.g., optimizing a golf swing), while others insist that only multiscale models that include tissue-level properties can truly predict injury or performance limits. This disagreement is productive: it drives the field to test whether added complexity improves predictions or merely adds computational cost. A second ongoing debate concerns the role of individual variability. Models are often built from averaged anatomical data, but athletes vary widely in anatomy, strength, and coordination. The field is now grappling with how to personalize models without requiring invasive measurements for every athlete.
Sports biomechanics has thus moved from watching athletes move, to measuring the forces they produce, to listening to their muscles, to simulating their bodies in silico, and finally to modeling the whole chain from molecule to playing field. Each paradigm added a layer of understanding, and the current challenge is to integrate those layers into tools that help athletes perform safely and effectively.