Jump Mat Contact Time Error From Floor Flex: Why Sprung Floors Inflate Your Numbers

A club gymnast tests a 30cm drop jump on the studio's competition floor and comes back with a contact time around 280 ms and an RSI hovering just above 1.2. Two days later, the same athlete runs the same protocol on the weight room's rubber-tiled concrete and posts 215 ms with an RSI near 1.7. Nothing about her jump changed. The coach starts second-guessing the rest day, the shoes, the warm-up - anything but the one variable that actually moved, which is the floor itself. Sprung and suspended floors are built to flex under load on purpose, for exactly the joint-friendly reasons that make them popular in dance and gymnastics facilities. That same flex sits directly underneath the switch mat during every single rep, and it changes what the mat measures without ever showing up as a number you'd think to check.

A contact mat, the kind used in most jump-testing systems since the original Bosco-style Ergojump devices, works on a simple principle: two conductive layers separated by a thin gap close a circuit the instant foot pressure presses them together, and open it again the instant that pressure releases. The system times how long the circuit stays closed to get ground contact time, and how long it stays open between reps to get flight time. That principle assumes the mat is sitting on something that does not itself move. On a poured concrete slab, that assumption holds. On a suspended wood floor, a floating gymnastics panel, or an elevated stage deck, it does not - the entire floor system deflects under body weight and springs back afterward, and the mat resting on top of it goes along for the ride.

Floor flex, in this context, means the vertical deflection of the floor assembly itself under load - typically a few millimeters to over a centimeter depending on the construction, whether it's a battened air-gap system, a rubber-puck sprung subfloor, or a cantilevered stage platform. That deflection is not a defect. Sprung floors are engineered to flex specifically to reduce impact loading on joints during repeated jumping and landing, which is exactly why gymnastics, cheer, and dance facilities install them. The problem is narrower than the floor being bad: it's that a contact mat was never designed to separate the athlete's own ground contact from the floor moving underneath both of them.

The leg behaves like a spring during a jump landing, and so does a compliant floor - the two sit in series with each other, mechanically speaking, which means the combined stiffness the body actually pushes against is always lower than the leg's stiffness alone whenever the floor underneath has any give in it at all. Lower combined stiffness, in the spring-mass models that describe jumping and hopping, means a longer contact time for the same body mass and the same vertical excursion. That's not a measurement quirk; it's the same physics that governs why a runner covers ground differently on sand than on a track.

McMahon and Greene (1979, Journal of Biomechanics) built an experimental track with tunable compliance and found that a surface roughly one-sixth as stiff as a runner's own leg-spring minimized contact time and maximized speed, predicting - and later confirming on Harvard's purpose-built indoor track - speed gains on the order of 2-3% over a rigid surface. Their model and track were tuned for running gait, not vertical drop-jump testing, and they used a custom-engineered track rather than a typical sprung gym floor or a contact-mat system, so the exact optimal-stiffness ratio doesn't transfer directly to a jump-testing setup. What does transfer is the underlying mechanism: surface compliance changes contact time by changing the effective stiffness the body is bouncing against.

Ferris and Farley (1997, Journal of Applied Physiology) hopped subjects in place on a servo-controlled surface spanning roughly a twentyfold range of stiffness and found that, given a steady rhythm and several cycles to adapt, subjects raised their own leg stiffness on the softer surfaces enough to keep contact time and center-of-mass movement within a few percent of each other across that entire range. That's a real limitation worth sitting with here: it means trained, rhythmic, repeated hopping on a fixed surface can partly self-correct. It does not mean a single drop-jump attempt, or the first few reps of an RSI test on an unfamiliar sprung floor, gets that same benefit - the adaptation the researchers measured took a consistent cadence and multiple cycles to develop, conditions a five-rep reactive-strength test on a floor an athlete rarely trains on doesn't provide.

Unlike a sensor artifact that scrambles a reading in random directions, floor flex pushes contact time the same way almost every time: longer. Flight time is largely unaffected, since it happens in the air with no interaction with the floor at all, so the RSI formula's numerator stays close to true while the denominator inflates - and RSI (jump height, or flight time, divided by contact time) drops accordingly. That directional consistency is what makes this artifact genuinely dangerous to a training log. A random noise source gets flagged eventually because the numbers bounce unpredictably. A steady bias that always points the same direction looks exactly like a real trend - a gradual RSI decline that a coach can talk themselves into reading as accumulated fatigue, poor tapering, or a stalled adaptation, when it's actually the studio's floor doing what it was engineered to do.

There's a second layer on top of the biomechanical one. A switch mat's closure depends on a physical gap being compressed shut and released. On a floor that's still flexing and rebounding under the athlete's weight, that gap can stay compressed a few extra milliseconds after true propulsive force has ended, simply because the local floor panel hasn't finished springing back - an instrumentation delay layered on top of the genuine physiological contact-time increase described above. And because sprung floors are rarely uniform, stiffness typically varies noticeably between a spot directly over a joist or spring mount and a spot at mid-panel, so the same athlete tested at two different points on what looks like one continuous floor can show meaningfully different contact times for reasons that have nothing to do with performance.

Not every floor that isn't bare concrete needs a fix, and the table below is a starting point from typical field construction types rather than a guarantee for any specific facility - the verification protocol in the next section is what actually confirms whether your floor needs the underlay, not this table alone.

The fix itself is mechanically simple: place a rigid board - roughly 18-20mm plywood or an equivalent stiff composite panel, sized at least 1m x 1m so it distributes load well beyond the mat's own footprint - directly on the sprung floor, then set the contact mat on top of the board. This decouples the mat from the floor's own spring-back by giving the athlete's foot force a stiff intermediate surface to load against instead of the compliant floor structure underneath. Most contact-mat manufacturers' installation guides specify a rigid, level, non-compliant test surface for exactly this reason, even though the reason itself rarely gets explained beyond that one line.

Run this once on any floor you suspect is suspended, floating, or engineered to flex, and repeat it whenever the test location within that floor changes.

1. Note whether the floor is a rigid slab or a suspended/sprung construction - pressing a hand down firmly near the test spot and feeling for give is a rough but usable first check.
2. Run 5 reps of your drop jump or repeat-hop protocol directly on the floor at the exact spot you normally test, logging rep-by-rep contact time and RSI, not just the average.
3. Place a rigid board (at least 1m x 1m, roughly 18-20mm thick) on that same spot, set the mat on top of it, and repeat the same 5 reps.
4. Compare both the mean and the rep-to-rep coefficient of variation for contact time between the two conditions - a mean shift of 10 ms or more paired with a CV that tightens noticeably on the board points to genuine floor-flex inflation, not athlete variability.
5. Film one rep from each condition in side-view slow motion at 240 fps or higher, counting frames from first ground contact to toe-off, to cross-check which condition's mat reading actually matches the visible contact window.
6. If a rigid board isn't practical for every session - a fixed competition floor, for instance - at minimum test in the same exact location every time, and never compare RSI trend data collected on a sprung floor directly against a baseline recorded on a rigid weight-room floor.

A club rhythmic gymnast performed a 30cm drop jump directly on her club's sprung competition floor, then repeated the same 5-rep set with a 20mm plywood board placed under the mat at the identical spot. Contact time and RSI for both conditions are below.

Mean contact time dropped from 275 ms directly on the sprung floor to 220.6 ms on the rigid board, a bias of roughly 54 ms - about a 20% reduction that had nothing to do with the athlete jumping any differently. The spread tells the rest of the story: the coefficient of variation for contact time was about 7.2% directly on the floor and fell to roughly 2.8% on the board, which is the signature of removing a noisy, location-dependent measurement layer rather than the athlete suddenly becoming more consistent. Average RSI moved from 1.17 to 1.50, a gap large enough on its own to misread as a real change in reactive strength if a coach only ever tested on that one sprung floor. A slow-motion cross-check on one rep from each condition put visually counted contact time within a few milliseconds of the rigid-board mat reading and well below the direct-floor reading, confirming the plywood underlay - not some overnight change in the athlete - was responsible for the difference.