What your watch reads about how you actually run — and what to do about it.
Running form gets discussed as if it's one thing, but the research splits into a few distinct questions: how you warm up before you run, how quickly your feet turn over, and what happens at ground contact. Each has its own evidence base, and each gets misapplied the same way — a nuanced finding turns into a rigid rule (stretch before you run, hit 180 cadence, minimize ground contact time). The sections below separate what the research actually supports from the oversimplified version that made it into running culture.
The clearest, least controversial finding in this pillar: static stretching on cold muscle — the classic pre-run toe-touch or quad pull — temporarily reduces strength and power output, an effect that can last up to an hour. A dynamic warm-up (leg swings, lunges, high knees, strides) does the opposite. Seven to ten minutes of it raises muscle temperature, speeds nerve conduction, and measurably improves explosive lower-body performance. The fix isn't complicated: move first, run, stretch later if you want to stretch at all.
→ Read: Warm up like a runner: dynamic moves now, save the stretching for later, Skip the Toe-Touches: Warm Up Dynamic, Stretch Static Later
If there's one form change with the most consistent research behind it, it's a modest cadence increase. A 2025 systematic review pooling 18 studies found that raising step rate by 5–10% — not sprinting, just taking slightly quicker, shorter steps at the same pace — reliably lowers vertical ground reaction force and loading rate, the two mechanical inputs most associated with overuse injury. One gait-retraining program in that review cut new-runner injury rates by 62%. Crucially, this comes at no cost to running economy: you're not burning more energy to take those extra steps.
The practical range worth knowing: if your cadence sits below roughly 165 steps per minute, that's where the room for improvement tends to live. Beyond a 5–10% bump, gains reverse — this is a dial, not a direction to keep turning.
→ Read: The Simplest Fix for Overuse Injuries: Step a Little Faster, Step a Little Faster: How a Small Cadence Bump Cuts Injury Risk, Turn Your Cadence Up a Notch and Take Stress Off Your Joints, A Small Cadence Bump Does Big Things for Your Joints
The mechanism above is solid; the mistake is turning "5–10% higher" into a fixed number everyone should chase, and 180 spm is the most common version of that mistake. That figure comes from Jack Daniels' observation that Olympic-caliber distance runners tended to land at or above 180 — but they weren't hitting 180 because it's magic, they were hitting it because they were running fast. Speed is cadence times stride length, and elite racing pace drives cadence up as a byproduct, not a cause.
Forcing your own cadence to an arbitrary number you didn't arrive at naturally tends to make your stride choppier, not more efficient. The better approach is gradual: nudge cadence up slightly on an easy run, hold pace steady, and let the change settle in over repeated sessions rather than treating any single number as pass/fail.
→ Read: Cadence isn't about running faster — it's about landing softer, Does your cadence really need to be 180? What the data says
The same "chase the elite number" trap shows up with ground contact time. Wearables report GCT and vertical oscillation as if lower is always better, but a 2019 study splitting sub-elite runners by duty factor found two distinct, equally economical strategies: runners with short ground contact and more vertical bounce, and runners with longer contact who stayed flatter. Same pace, same energy cost, different mechanics. There's no single "correct" GCT to hit.
What does matter is symmetry and stability within a run — left/right GCT imbalance, or a GCT that creeps up over the second half of a long run, correlates with form breakdown in a way that a static absolute number doesn't.
→ Read: Short ground contact time isn't always faster — and neither is low vertical bounce
These nine articles form the research base behind RunNerd's approach to form: warm up dynamically, treat cadence as a trend rather than a fixed target, and read gait metrics like ground contact time for drift and asymmetry rather than comparing them to elite benchmarks.