Dropped frames hurt more than no motion because they expose strain

What it is

A displacement that drops frames is not “a slightly worse clip”; it is a string of irregular hops. A hard cut is at least stable. Irregular motion shows the system straining. People read the stutter as the device failing, not as “the designer wanted a transition and missed vsync.” With no motion, the worst case is a change that is easy to miss; with motion that drops frames, a capability claim has failed.

Why it happens

Apparent motion needs roughly even time samples. Miss several vsyncs and the object’s spatial steps lurch, putting spikes in the velocity profile. Vision is more sensitive to that jitter than to a few milliseconds of mean-speed error — it resembles the struggle before a freeze. Irregular optic flow also upsets the vestibular system more than steady flow; unpredictable flow is the worse stimulus.

A hard cut never promised “I will travel there smoothly.” A dropping animation promised smoothness and delivered struggle, so the expectation violation lands on capability. Mean frame rate can still “look close to 60” because the tail was averaged away; what people see are those few stalls.

Studying it

Run the same transition at full frames, with about 20% of frames dropped, and as a hard cut. Dependent measures: smoothness ratings, capability attribution (slow device / bad app), discomfort, change detection. Prediction: the dropped-frame group is worse than the cut on capability attribution, and not necessarily better on detection. Simulate a real long-tail distribution of drops, not every fifth frame — uniform loss can still look metrically regular.

Where it stops holding

• A very short marker fade that drops one frame is hard to encode as struggle, because the path promise was weak to begin with.
• Dropped frames during the user’s own scrolling also expose strain; that is scroll performance, not decorative-transition jank, though the look is kin.
• On a high-refresh display a single dropped frame is easy to miss; a run of tail frames is not.

Applying it

• If a transition cannot hold the frame budget, default to a cut or a static marker rather than playing it stuttered.
• Decide whether the motion ships from tail frame time, not from the mean.
• Check: watch the transition on the target device. If stutter can be counted, do not ship that clip.

Related

• Same group: F7.12.1 Motion work can contend for the main thread and cause jank · F7.12.3 On low-end devices, complex motion should degrade or stop, not scale uniformly · F7.12.4 Simultaneous motions stack cost beyond a single-motion budget
• Nearby: A1.08.4 Higher refresh cannot cancel sampling jitter from content motion · F7.11.4 Forced motion feels slow in high-frequency use
• Search terms: dropped frames · jank · irregular motion