How to Interpret Running Form Metrics on Your Watch

Your running watch spits out four or five numbers after every run—cadence, ground contact time, vertical oscillation, ground contact balance—and most runners have no idea if their numbers are good, bad, or meaningless. Here's what each one actually measures, what ranges are typical, and where the sensor's confidence runs out.

What do cadence, ground contact time, oscillation, and balance actually measure?

Cadence is your step rate: how many steps (not strides) you take per minute, abbreviated spm. It's the most reliably measured of the four, because it's just a rhythm—the watch counts foot-strike timing from an accelerometer, and accelerometers are good at counting repetitive motion.

Ground contact time (GCT) is how many milliseconds your foot spends touching the ground on each step. Vertical oscillation is how many centimeters your torso bounces up and down with each stride. Ground contact balance is the left/right split of your GCT, shown as a percentage (ideally close to 50/50).

All four come from the same wrist- or chest-mounted accelerometer and gyroscope, but they're not equally trustworthy. Cadence is a straightforward count. GCT and oscillation are estimates derived from motion patterns and manufacturer algorithms, and different brands calculate them differently—which is why the same run can show different numbers on a Garmin versus a Coros.

What cadence should you be aiming for?

A commonly cited range for recreational distance runners is 170–180 steps per minute, though this varies with height, leg length, pace, and running experience. Faster runners and those with shorter legs often land higher in or above that range; slower joggers and taller runners may sit comfortably below it.

The number itself isn't a pass/fail test. Research on step-rate manipulation (increasing cadence by 5–10% at a fixed pace) has shown it can reduce impact loading at the hip and knee in some runners, which is why coaches sometimes suggest small increases rather than large jumps. But cadence is highly individual, and forcing a number that doesn't fit your height or pace can create new problems—like overstriding in the other direction or added fatigue from choppier steps. Aim for small, gradual shifts (a few spm at a time over several weeks) rather than a one-run overhaul.

What's a normal ground contact time and left/right balance?

Typical ground contact times fall somewhere around 200–300 milliseconds, with faster runners and more efficient strides generally trending toward the lower end and slower or fatigued running trending higher. GCT naturally increases on hills, in trail conditions, and late in long runs—that's expected, not a red flag on its own.

Balance is usually reported as a percentage split, like 51%/49%. Small asymmetries are extremely common and don't necessarily mean anything is wrong; running is not a perfectly symmetrical activity, and everyone favors a side slightly. A persistent, growing imbalance—say, drifting from 50/50 toward 54/46 over several weeks—is more worth noting than a single run's split. If you want a deeper look at what a lopsided GCT balance might or might not mean, see ground contact time and balance in running, and for stride-level left/right differences specifically, stride length asymmetry between left and right walks through when that's worth a closer look.

What counts as normal vertical oscillation?

Vertical oscillation commonly falls in the range of about 6–13 centimeters per stride. Lower oscillation is often associated with more efficient running, since less energy is spent bouncing and more is spent moving forward—but this isn't a strict rule, and elite runners still show a range of values. Watches sometimes pair oscillation with a "vertical ratio" (oscillation divided by stride length) as a rough efficiency proxy, though this too is an estimate, not a lab measurement.

Context matters more than the raw number. Oscillation typically rises with fatigue, on trails, and at faster paces where you're generating more vertical force per stride. A single high reading on a hard hill workout means little; a steady upward trend across easy runs at the same pace is more informative. For a fuller breakdown of what's typical and what might warrant attention, see normal vertical oscillation numbers.

Metric Typical Range What It Reflects Sensor Confidence
Cadence ~170–180 spm Step rate/rhythm High — direct count
Ground contact time ~200–300 ms Time foot spends on ground Moderate — algorithm estimate
Vertical oscillation ~6–13 cm Vertical bounce per stride Moderate — algorithm estimate
GCT balance Close to 50/50% Left/right symmetry Moderate — algorithm estimate

How accurate are wrist and chest sensors, really?

Wrist-based accelerometers are convenient but noisy: arm swing, wrist rotation, and watch fit all introduce variability into GCT and oscillation readings. Chest-strap sensors, which sit closer to your body's center of mass, tend to produce steadier oscillation and GCT numbers than wrist units, which is why several brands recommend a chest strap specifically for running dynamics rather than the watch alone.

Even then, neither sensor type is validated against lab-grade motion capture the way research-grade 3D systems are. Consumer wearables are good at trends over time—is this week's average GCT creeping up compared to last month's—but not at precise, one-decimal-point accuracy on any single run. Treat the numbers as directional signals, not clinical measurements.

When should you add video to your watch data?

Your watch can tell you that something changed—cadence dropped, oscillation crept up, balance shifted—but it can't show you why. Video fills that gap. A side-view phone video can visually confirm things a watch only estimates, like whether you're landing with your foot too far ahead of your hip (overstriding) or whether one arm swings noticeably more than the other.

It's worth adding video when: your watch metrics shift and you want to see the mechanism, you're troubleshooting recurring soreness on one side, or you're checking foot-strike pattern, which watches don't measure directly (see is my foot strike normal for a beginner runner for what's typical there). A phone-video tool like StrideIQ can give you a cadence and general form check from that footage—useful for a quick, low-cost look between runs, though it's not a substitute for an in-person gait lab or a physical therapist's assessment when something actually hurts.

What can watch metrics not tell you?

Watch data can't diagnose an injury, confirm a joint angle, or tell you definitively that your form "causes" pain. The link between specific form metrics and injury risk is genuinely mixed in the research; cadence and GCT trends are associative signals at best, not proof of cause and effect. A number outside the "typical" range isn't automatically a problem, and a number inside it doesn't guarantee you're pain-free material.

Wrist and chest sensors also can't replace true biomechanical assessment. They don't measure joint angles, hip drop, or pelvic rotation, and their GCT/oscillation estimates carry real margin of error run to run. If you're dealing with persistent pain, a change that doesn't resolve with rest, or you're returning from injury and want individualized guidance, see a physical therapist or sports medicine physician and consider a formal gait analysis lab, which uses motion capture or force plates for measurements a wearable simply can't provide. Watch metrics are a useful trend line—not a diagnosis, and not a treatment plan.

Frequently Asked Questions

Is a lower vertical oscillation always better?

Generally, lower oscillation is associated with more efficient running because less energy goes into bouncing, but there's no single ideal number—elite runners show a range, and oscillation naturally rises with fatigue, hills, or faster paces. Compare your own trend over time rather than chasing a specific figure.

Why do my watch and a friend's watch show different ground contact times for the same run?

Different brands use different proprietary algorithms to estimate GCT from accelerometer data, and wrist versus chest placement changes the signal quality. Treat cross-device comparisons cautiously and focus on trends within one device instead.

Should I try to force my cadence into the 170-180 spm range?

Not abruptly. That range is commonly cited but varies by height, pace, and experience. If you want to adjust, shift a few steps per minute at a time over several weeks rather than jumping straight to a target number.

Can watch metrics tell me if I'm at risk of injury?

No. They show associative trends—like a rising GCT or shifting balance—but research on form metrics and injury risk is mixed, and a watch can't diagnose anything. Persistent pain warrants a physical therapist or sports medicine evaluation, not just a metrics review.

Is a chest strap more accurate than a wrist watch for running dynamics?

Chest straps sit closer to your body's center of mass and tend to produce steadier ground contact time and oscillation readings than wrist-based sensors, which is why several manufacturers recommend them specifically for running dynamics data.

Sources