Why didn't my jump register?

If you clearly got air but RidePal didn’t list a jump, the event almost always failed one of the automatic checks below. These filters exist to reduce false jumps from curbs, hike-a-bike, phone rattling, or long smooth descents. Specific cutoffs (times, speeds, g-levels, sample counts) are intentionally not published—they can change between app versions.

Platform note: The behavior below matches RidePal for iOS live detection. Android currently focuses on displaying jump data when a ride includes it; if you recorded on Android and see no jumps, detection may not be running the same pipeline yet.


Quick checklist

CheckWhy it matters
Wait a momentJumps are detected from batched processing, not always immediately when you land
GPS qualityTakeoff speed comes from GPS; weak or stale GPS can invalidate a jump
Not right at recording startA short warmup at the beginning of a ride ignores early noise
Enough speedVery slow hops may be treated as non-jumps
Phone carryPocket carry is tested and usually works well; rigid bike mounts sometimes cause more missed jumps—try a pocket if you only see issues mounted
Track jumps (iOS)On the ride start screen, Track jumps (beta) must be on for that ride (default is on)

Jump tracking off for this ride (iOS)

If Track jumps (beta) is disabled when you tap start, RidePal does not run the accelerometer jump pipeline for that activity—so zero jumps is expected.

Fix: Enable the toggle before starting the next ride.


Warmup at the start of recording

The first part of a ride after jump recording starts is treated as a warmup so mounting the phone, starting the timer, or rolling from a standstill doesn’t create bogus events.

Fix: Start recording before your first jump when you care about that hit counting, or accept that an immediate hop right after start may be skipped.


Pocket vs bike mount

Jump tracking is tested and is intended to work with your phone in a pocket (shorts or jersey—snug enough that the phone moves with your body). If you use a stem or handlebar mount and see more missed or odd jumps, try the same riding with the phone in a pocket; we’ve noticed more issues with some mount setups than with typical pocket carry.


“Freefall” must be sustained

Takeoff is not one bad sample—the detector needs a coherent run of low-magnitude readings that look like air, not vibration. If the pattern keeps breaking, no takeoff is confirmed.

Fix: Stable carry; reduce rattling on the lip.


Landing and timeouts

A landing must show a clear impact signature within a bounded time. If the landing is extremely soft, damped, or the phone doesn’t see a spike, the segment may expire as “not a jump.” Very long low-magnitude stretches without a landing are also discarded.

Fix: Unusual mounts or padding can soften the signal; this is uncommon on typical MTB setups.


Airtime limits

After a landing is found, airtime must fall between a minimum and a maximum that the model considers a real jump. Anything too short looks like noise; anything too long looks like a non-jump trail segment.


Speed at takeoff (GPS)

You need a credible forward speed from GPS at takeoff. The app blends recent speed readings and rejects fixes that look stationary or too far from the takeoff moment.

If…Then…
GPS dropped or you were barely movingSpeed check can fail and the jump is rejected
Takeoff under heavy tree cover or in a structureSame risk

Fix: Let GPS settle before the line; avoid relying on stats from dead zones if you care about every jump.


Device support (iOS)

On some iPhone models or OS versions, background accelerometer recording is not available, so jump capture may not run for that device.


Duplicates and end of ride

The app avoids counting the same lip twice when processing overlaps.

Ending a ride schedules a final flush of sensor data; force-quitting the app immediately can sometimes cut off the tail of processing.


Still stuck?

If a jump reliably fails despite good GPS, a solid mount, and normal riding speeds, note rough time on trail, phone model, and mount type and contact support—that helps distinguish a bug from edge-case physics.


Related