Edge and corner targets need extra margin

What it is

Touch targets at screen edges and corners are practically harder to reach than a visually identical control placed centrally, and so typically need more margin. The cause is that the physical constraints that set the minimum target-size floor — contact area and motor precision — get worse near the edge; there is nothing special about display behaviour at the edge itself.

Why it happens

Two distinct factors make edges worse. The first is motor control: a central target can be approached from any direction, giving the finger's trajectory room to correct; an edge target has no buffer on its outward side, and a corner is constrained in two directions at once, cutting away the adjustment room a fingertip has near the physical bezel and letting motor noise cross the boundary more easily. The second is sensing: capacitive touch sensors suffer field-strength falloff at the physical edge — the capacitive edge effect — because the sensor array's outermost electrodes have incomplete coverage, so contact signals near the edge are inherently more prone to distortion or lag than in the centre. This is not a motor-control problem; it is the hardware's sensing capability degrading naturally at the boundary. Both factors compound, and many systems additionally reserve a gesture zone at the screen edge (an inward swipe for back navigation or a system menu), so all three together shrink the territory an edge target can safely be hit within.

Studying it

Place an identical control at the centre, at each of the four edges, and at each of the four corners, and compare error rate, completion time, posture adjustments, and accidental system-gesture triggers across one-handed, two-handed, and different orientation conditions. A methodological point: edge performance cannot be inferred from centre-trial hit-rate data, because the edge introduces two entirely separate extra error sources — motor control and sensing distortion — that centre data contains none of; edge testing must also include conditions where a system gesture competes for the same input (an edge-swipe-to-go-back gesture active at the same time), otherwise the measured "edge hit rate" only reflects the motor-control half of the problem and hides the additional failures caused by gesture preemption.

Where it stops holding

This edge disadvantage is specific to touchscreens. A mouse or trackpad pointer in a traditional GUI actually gains an advantage from being stopped by the screen boundary — corner targets become the easiest to hit, a classic result in Fitts's-law terms, which is why menu bars sit at the top edge of the screen. Touch has no equivalent "brace against the boundary" effect, because a finger is not a pointer constrained to slide along the screen plane; once contact falls outside the physical bezel it simply is not sensed at all. Not every device has the same strength of edge field falloff — a bezel-less full-screen device and one with a raised physical bezel behave differently at the edge — so the actual margin should come from measurement on the target device rather than a generic coefficient. Adding margin to edge controls also does not mean turning every edge target into a giant button: relocating it inward to a safe area, or offering an alternative that does not depend on an edge swipe, is often more effective than simply enlarging the edge target.

Applying it

• Add extra hit buffer inward from the bezel for critical controls placed at an edge or corner, rather than positioning them flush against it visually.
• Check whether a target's position conflicts with a system-reserved edge-gesture zone, and provide an equivalent entry point that does not depend on an edge swipe rather than letting the two compete for the same screen territory.
• Test the corner range reachable by a one-handed thumb and the edge range used by two-handed index fingers separately, since their reachable arcs differ.

Related

• Same group: C2.02.1 The minimum touch-target size is set by fingertip contact area, not vision · C2.02.2 Visual size and hit size can be separated · C2.02.3 Recommended minimums differ by platform and cannot be mixed
• Nearby: C2.03 Minimum Touch-Target Spacing · C2.04 Fat-finger problem
• Search terms: capacitive edge effect · safe area · corner target acquisition