Motor Control and Fitts's Law

Learning Objectives

Describe the human motor system and the stages of movement planning and execution
State Fitts's Law and explain its information-theoretic basis
Apply Fitts's Law to calculate movement time for pointing tasks
Understand the steering law and its application to constrained movement
Use motor control principles to size and position interactive elements

Introduction

After perceiving information and deciding on an action, the user must execute it. Clicking a button, tapping a touchscreen target, turning a door handle, or moving a mouse to a menu item — all require the motor system to translate intention into physical movement. The speed and accuracy of these movements are not arbitrary; they follow quantitative laws that allow designers to predict performance and optimise layouts. This chapter covers the fundamentals of human motor control relevant to design, with particular emphasis on Fitts's Law — the most widely validated quantitative model in human-computer interaction MacKenzie, 1992.

The Motor System

Human movement is controlled by a hierarchy of neural structures. The motor cortex generates movement commands; the cerebellum coordinates timing and accuracy; the basal ganglia modulate the initiation and scaling of movements; and the spinal cord executes reflexes and low-level motor patterns.

Open-Loop and Closed-Loop Control

Simple, rapid movements (a ballistic keystroke, a quick flick gesture) are primarily open-loop: they are planned in advance and executed without real-time correction. Slower, more precise movements (guiding a cursor to a small target, threading a needle) are closed-loop: the motor system continuously monitors the movement through visual and proprioceptive feedback and makes corrections during execution.

Key Principle

The distinction between open-loop and closed-loop control explains why very small targets are disproportionately slow to acquire. The initial ballistic phase of movement gets close to the target quickly, but the final correction phase — where closed-loop control guides the hand to the exact target position — takes progressively longer as the target gets smaller. This is the mechanism underlying Fitts's Law.

Reaction Time and Movement Time

Motor performance has two components: reaction time (the delay between stimulus and movement onset) and movement time (the duration of the movement itself). Reaction time is typically 150–300 milliseconds for simple reactions and increases with choice complexity (as described by Hick's Law in Chapter 4). Movement time is the primary concern of Fitts's Law.

Fitts's Law

In 1954, Paul Fitts published a paper that would become one of the most cited in human factors research Fitts, 1954. Drawing on Shannon's information theory, Fitts proposed that the difficulty of a pointing movement could be quantified as an "index of difficulty" (ID) based on the distance to the target (D) and the width of the target (W). ID = log2(2D/W) The movement time (MT) is then a linear function of this index of difficulty. MT = a + b × ID = a + b × log2(2D/W) where a and b are empirically determined constants that depend on the specific device and conditions.

Design Law

Fitts's Law states that the time to acquire a target is a logarithmic function of the ratio of distance to target width. To make targets easier (faster) to acquire: (1) make them larger, (2) move them closer to the user's starting position, or (3) do both. Conversely, small targets far from the cursor are the hardest to acquire and should be avoided for frequently used actions.

The Shannon Formulation

The formulation most commonly used in HCI research is the Shannon formulation MacKenzie, 1992, adopted by the ISO 9241-9 standard for pointing device evaluation Standardization, 2000. ID = log2(D/W + 1) This formulation avoids negative IDs for very close targets and has better empirical fit in some conditions. The practical implications are identical: larger targets and shorter distances mean faster acquisition.

Empirical Validation

Fitts's Law has been validated across an extraordinary range of conditions: mouse pointing, touchscreen tapping, pen input, eye tracking, head-mounted displays, and even physical reaching movements. The specific values of a and b change with the input device, but the logarithmic relationship between ID and movement time is remarkably consistent. Typical R² values for Fitts's Law fits exceed 0.95.

Throughput

The throughput (TP) of a pointing device is defined as the index of difficulty achieved divided by the movement time. TP = ID / MT Throughput provides a single metric for comparing pointing devices that accounts for both speed and accuracy. ISO 9241-9 specifies throughput as the primary metric for evaluating pointing devices Standardization, 2000. Typical mouse throughput is 4–5 bits/second; touchscreens achieve roughly 5–7 bits/second for direct finger input.

Applying Fitts's Law to Design

Button Size

The W term in Fitts's Law corresponds to the clickable area of a target, not its visual size. Making a button visually larger but keeping its clickable area small provides no benefit. Conversely, extending the clickable area beyond the visible boundary (a common technique for small icons) improves acquisition time.

Example

Apple's iOS Human Interface Guidelines specify minimum touch target sizes of 44 × 44 points Inc., 2024. This is not arbitrary — it reflects the practical minimum for reliable finger acquisition on a touchscreen given typical finger pad width (~10mm) and the accuracy of closed-loop correction at the distances involved in thumb-based interaction Hoober, 2013.

Target Distance

The D term in Fitts's Law is the distance from the cursor's current position (or the user's finger) to the centre of the target. Frequently used controls should be placed close to where the cursor typically rests. In desktop applications, the menu bar at the top of the screen benefits from the screen edge acting as an infinite-width wall — the user can "throw" the mouse upward without overshooting, effectively making D smaller and W infinite. This is why menu bars at screen edges are faster than menus in floating windows.

Design Law

Screen edges and corners are the easiest targets to acquire because the cursor cannot overshoot them. Placing frequently used controls at screen edges (menu bars, taskbars, scroll bars) exploits this "infinite width" effect. The four corners of the screen are the highest-value locations for pointing because they benefit from two infinite edges. Pop-up context menus (which appear at the cursor position) set D to near zero, making them among the fastest menus to use.

The Speed-Accuracy Trade-Off

Fitts's Law describes the optimal trade-off between speed and accuracy: moving faster means more errors, and increasing accuracy requires slowing down. Designers cannot control the user's speed-accuracy strategy directly, but they can influence it by:

Making targets large enough that users can be both fast and accurate
Providing feedback that informs users about their accuracy
Using confirmation mechanisms for critical actions where errors are costly

Grouping Related Controls

Fitts's Law implies that controls used in sequence should be placed close together to minimise total movement distance. A dialog box with "OK" and "Cancel" buttons far apart forces a long movement if the user changes their mind. Placing related controls in a cluster reduces the average movement distance for common workflows.

The Steering Law

The steering law, proposed by Accot and Zhai Accot, 1997, extends Fitts's Law to constrained movements — moving through a tunnel or along a path, such as navigating through nested menus or drawing along a curved line. For a straight tunnel of width W and length D. T = a + b × (D/W) Note that the relationship is linear (D/W), not logarithmic (log(D/W)). This means that constrained movements are much more sensitive to path length and width than pointing movements are to distance and target size.

Key Principle

Cascading menus — where the user must move the cursor horizontally through a narrow corridor to reach a submenu — are governed by the steering law, not Fitts's Law. The narrow corridor makes them slow and error-prone. Design alternatives include: (1) wider activation zones, (2) a brief delay before the submenu disappears (Amazon's "triangle" technique), (3) mega-menus that avoid cascading altogether, or (4) reorganising the menu structure to reduce depth.

Other Motor Control Considerations

Two-Handed Interaction

Guiard's model of bimanual interaction Guiard, 1987 describes how the two hands collaborate. The non-dominant hand sets the frame of reference (holding a phone, positioning a ruler), while the dominant hand performs fine manipulation within that frame. Designs that require the non-dominant hand to perform precise actions violate this natural division of labour.

Touch vs. Mouse vs. Keyboard

Different input devices have different motor characteristics:

Mouse: indirect pointing with high precision; benefits from screen edges; limited by the speed-accuracy trade-off of indirect control
Touchscreen: direct pointing with lower precision (finger occlusion, fat-finger problem); faster for large targets; no screen-edge benefit
Keyboard: discrete key presses with essentially binary accuracy; governed by typing speed rather than Fitts's Law; fastest for expert text entry and known shortcuts

Think About It

Touchscreens invite direct manipulation but sacrifice precision. A surgeon using a touchscreen medical device in the operating room faces the challenge of sterile glove interaction, small targets on a potentially cluttered display, and the need for high accuracy. How should the interface adapt? Larger targets? Gesture shortcuts? Voice control? Each input modality brings its own motor control constraints.

Muscle Groups and Fatigue

Different movements engage different muscle groups with different endurance characteristics. Gross arm movements (reaching) are powered by large muscles and can be sustained. Fine finger movements (typing, tapping) use smaller muscles and are more prone to fatigue. Prolonged pointing with a mouse involves sustained static contraction of shoulder and forearm muscles, contributing to repetitive strain injuries. Ergonomic design considers not just movement time but movement sustainability. Touchscreen kiosks that require sustained arm-raised interaction ("gorilla arm") cause fatigue quickly. Keyboards positioned at elbow height reduce strain. Input devices that allow variation in posture and muscle use are more sustainable for prolonged tasks.

Practical Design Guidelines

Size targets for the input device. Mouse targets can be smaller than touch targets. Minimum touch targets should be at least 44 × 44 points (approximately 7 × 7 mm).
Minimise movement distance for frequent actions. Place commonly used controls near the cursor's resting position or along common movement paths.
Exploit screen edges. Menu bars, taskbars, and scrollbars at screen edges benefit from infinite-width targeting.
Avoid narrow tunnels. Cascading menus and narrow drag paths are governed by the steering law and are inherently slow. Provide wider activation zones or alternative interaction patterns.
Group sequential controls. Controls used in sequence should be spatially proximate to minimise total movement.
Consider motor sustainability. Designs for prolonged use should minimise static postures, avoid gorilla arm, and accommodate variation in hand size and dexterity.

Key Takeaways

Human motor performance has two phases: a rapid ballistic phase and a slower corrective phase. Small targets are slow because they extend the corrective phase.
Fitts's Law (MT = a + b × log2(D/W + 1)) predicts pointing time as a logarithmic function of distance-to-width ratio. It has been validated across virtually all pointing devices.
To make targets easier to acquire: increase their size, decrease their distance from the cursor, or exploit screen edges.
The steering law governs constrained movements (navigating tunnels, cascading menus) and is linear rather than logarithmic — making narrow paths disproportionately slow.
Screen edges and corners are the highest-value target locations due to the infinite-width effect.
Input device choice (mouse, touch, keyboard) changes the motor control constraints and should influence interface design.