Area and bubble cursors enlarge activation without changing visual target size

What it is

An area cursor hit-tests with a pointer region rather than a single hotspot. A bubble cursor, introduced by Grossman and Balakrishnan, dynamically expands that region further, typically selecting the uniquely nearest target. Both enlarge actual activation without changing a target's visual size, helping people point at small objects.

Why it happens

A conventional cursor requires one point to fall inside a target. An area cursor treats any intersection between the pointer's outline and an object as a candidate hit; a bubble cursor expands its radius to the nearest target's boundary while avoiding a second target. Both hand part of the spatial-discrimination work from endpoint control to an algorithm, but must continuously communicate which object is currently going to be selected.

The theoretical grounding sits in Fitts's law: a target's effective width W determines pointing time. For a point cursor, effective width is just the target's visual width; an area cursor swaps that for "the minimum movement tolerance at which the pointer region intersects the target." For a sparse, isolated small target, this tolerance can be far larger than the target's own pixel width — equivalent to artificially inflating the W term in Fitts's law without changing a single visual element on screen, which shortens the theoretical pointing time. This is also why the technique is meaningless for fine drawing or continuous paths: their difficulty is not "landing inside a target" but requiring the endpoint's continuous coordinate itself to be precise, and enlarging a discrete hit tolerance does nothing for continuous control.

Studying it

Compare point, fixed-area, and dynamic-bubble cursors across target size, spacing, density, and distance. Measure time, errors, candidate switches, visual confirmation time, and learning cost. Include sparse and dense layouts; isolated small targets alone exaggerate the benefit. When comparing throughput (bits/s) across conditions using Fitts's law, recompute the index of difficulty using the area cursor's actual achieved effective width — using the target's raw visual width instead will understate its throughput gain.

Where it stops holding

An area cursor is not unconditional hit-region expansion: dense layouts create candidate ambiguity. Dynamic area can also make selection feel unstable. Drawing, text insertion, and fine dragging that need continuous precise position cannot directly substitute area hit testing.

Applying it

• Evaluate area or bubble assistance for small, sparse selection targets, and highlight the current candidate.
• Keep candidate-switching rules predictable, avoiding unannounced selection jumps near the pointer.
• How to check: measure completion time and error rate separately for sparse, dense, drag, and fine-edit tasks, and do not generalise a speed gain measured in a sparse scenario to dense layouts or continuous drawing.

Related

• Same group: C1.13.2 Benefits are substantial when targets are sparse · C1.13.3 Area and bubble cursors are fundamentally incompatible with dense layouts
• Nearby: C1.12 Target snapping and sticky targets · C1.14 Side effects of pointing assistance
• Search terms: area cursor · bubble cursor · Fitts's law effective width