Every hub brand claims low friction. We built the rig, tested five, and published the white paper.
Drivetrain efficiency usually gets measured while pedaling. That covers maybe half of a gravity ride. The other half you're coasting, and the drag in your rear hub is quietly taking speed off the bike the entire time.
Hub brands do not publish numbers for that. So we measured it, including our own, and wrote a white paper so you can argue with it.
How the test works
We built a precision spin-down stand. A 15 lb plate bolts to the hub's disc tabs, giving the assembly a known and dominant moment of inertia. Spin it up, let it coast, and measure how long it takes to decay from 80 RPM to 20 RPM. An inductive sensor counts screw heads passing; an Arduino logs it.
Longer decay means less drag. The physics is simple; getting it repeatable is the work.
Three configurations, each isolating something different:
Freewheeling. The plate drives the hub shell while the freehub body is held still. This is a coasting wheel: shell and bearings turning, drivetrain stationary, ratchet actively ratcheting.
All-bearing, no ratchet. The hub is spun with no relative motion between freehub and shell, which bypasses the ratchet entirely and leaves only bearing and seal friction. A loose analog for pedaling.
Complete wheel. No plate. The whole built wheel is spun with the freehub held, which captures the total system resistance a rolling wheel actually sees. Its inertia is measured first with a trifilar pendulum.
Every hub in the test on the stand, in the order they appear in the table, then the sensor, the logger, and a complete wheel on test.
What we found
Converted to power dissipated while coasting at 30 km/h, on a 29" wheel:
| Hub | Power lost at 30 km/h |
|---|---|
| e*thirteen Sidekick 2.0 | 0.179 W |
| Hope Pro 5 | 0.616 W |
| RaceFace Vault | 0.772 W |
| Industry Nine Hydra | 1.334 W |
| DT Swiss 350 DEG DF | 1.549 W |
Drag torque tells the same story from the other end. The hatched bar on each row is bearing-only drag, what the hub would give you with the ratchet taken out of the equation. On the Sidekick the two bars are almost the same length, because the ratchet is barely there while you coast. On the DT Swiss the gap between them is most of the bar.
Against the DT Swiss 350 DEG, now offered in the DEG DF hub version built to solve the same kickback problem we solve, Sidekick showed an 88.5% reduction in isolated freewheeling drag and 51% less total drag in the complete wheel test.
Why the gap is that big
Most freehubs hold their pawls or ratchet rings engaged with springs. Those springs are what make engagement instant and reliable, which is what a high-engagement hub is for. They also keep pushing while you coast. Every click is friction, and the freehub never stops rubbing.
Sidekick uses a pusher-activated ratchet. While you coast, the drive pawls retract completely, leaving only a single light pusher pawl doing the timing. There's almost nothing left to create friction. The white paper doesn't describe our internals, only measures them; the mechanism is explained in full in the anti-kickback post below.
The same retraction is why the hub is nearly silent, and why it takes pedal kickback out of the suspension.
What the test doesn't prove
Here is what this test can't tell you.
No rider on the bike. These are unloaded spin-downs. Real hubs carry radial load, which changes bearing behavior. This is a sensitive baseline, not a road simulation.
Aerodynamic drag is in there. For hub-only tests the plate profile keeps it small but not zero. For complete wheels it's comparable between wheels, and because a more efficient hub spins longer it accumulates more aero loss. The white paper's own note: that makes the complete wheel result conservative, not flattering.
Grease warms up. Consecutive runs drift as bearing grease heats. And a hub's true drag settles only after hundreds of miles, which no accelerated break-in reproduces.
We ran the test. We've an obvious interest in the outcome. That's exactly why the method, the apparatus, and the limitations are published rather than summarized. The whole white paper is on our support site, with photographs of the rig.
Is 1.37 watts worth caring about?
On its own, no. It's a rounding error next to a rider's output.
But it's constant. It applies every second you're not pedaling, which on a long descent is most of them. Over a four minute run with three minutes of coasting, that's about 250 joules you keep instead of leaving in the hub.
Marginal gains are only marginal until you stack them.
We also put Sidekick and its main competitor side by side, spec for spec, on Sidekick versus DT Swiss DEG DF. The Sidekick offers far more in its total package with an affordable price.
DT Swiss, 350 and DEG are trademarks of DT Swiss AG. Hope and Pro 5 are trademarks of Hope Technology Ltd. RaceFace and Vault are trademarks of Fox Factory Holding Corp. Industry Nine and Hydra are trademarks of Industry Nine, Inc.
