Fitts's Law

011 min

Fitts's Law at a glance

• What it is: Movement time rises with distance to the target and falls as target size grows.
• Where it comes from: Paul Fitts, 1954; reformulated for computer interfaces by Scott MacKenzie in 1992.
• Apply it by: Make frequent targets bigger, move them closer, or pin them to a screen edge.
• It backfires when: Enlarging every target leaves less room and forces more scanning.

022 min

What Fitts's Law actually says

Fitts's Law states that the time it takes to move a pointer to a target depends on two things: the distance to the target, and the target's size. The version used across interface design today, credited to computer scientist Scott MacKenzie, writes the relationship like this:

MT is the movement time. A is the amplitude, the distance from the pointer's starting position to the center of the target. W is the width of the target, measured along the line the pointer travels. The constants a and b come from fitting the formula to measured data for one specific input device — a mouse and a finger on glass do not share the same a and b, so the formula predicts a relationship between distance, size, and time, not a fixed number of seconds by itself.

The term log2(A / W + 1) is called the index of difficulty, measured in bits. Because it uses a logarithm, doubling the distance to a target does not double the time it takes to reach it: the effect of distance shrinks as distance grows, and the same is true for width in reverse.

A worked comparison shows what this means in practice. A target 40 millimetres away and 15 millimetres wide has an index of difficulty of log2(40/15 + 1), about 1.9 bits. A target 150 millimetres away and 5 millimetres wide has an index of difficulty of log2(150/5 + 1), about 4.9 bits. The second target is less than four times farther away, but its index of difficulty is more than two and a half times higher. It is also three times narrower, and size matters as much as distance.

Reaching for a target happens in two phases. An initial, fast movement carries the hand roughly toward the target, driven mainly by the distance to travel. Then a slower, second phase corrects for aim: the pointer needs to slow down in order to not overshoot and click beyond the target. A narrow target forces a longer, more careful second phase, which is why width and distance combine into one difficulty score instead of acting on their own.

031 min

Where Fitts's Law comes from

Paul Fitts published the finding in 1954, in the Journal of Experimental Psychology, in a paper titled "The information capacity of the human motor system in controlling the amplitude of movement." Fitts asked participants to move a stylus back and forth between two metal plates on a table as fast as they could, an exercise now called the reciprocal tapping task. He varied the plates' width and the distance between them across trials, then measured how those two changes affected the average time per tap.

Fitts modeled the task using an analogy to Claude Shannon's information theory. The plate's width represented a channel's tolerance for error. The distance represented the signal being sent. His original formula for the index of difficulty was ID = log2(2A / W). The factor of 2 is absent from the version used today.

In 1992, Scott MacKenzie published a reformulation aimed specifically at human-computer interaction, in a paper called "Fitts' law as a research and design tool in human-computer interaction." MacKenzie proposed ID = log2(A / W + 1) instead. The form is often called the Shannon formulation, after its resemblance to a theorem in Shannon's own work, and it is the version cited across most interface-design writing today, including the formula given earlier in this article.

042 min

How Fitts's Law shows up in real interfaces

Applied: the menu bar at the screen edge

In macOS, the application menu sits at the very top of the screen rather than inside each window. Because the cursor cannot travel past the top edge of the screen, the menu bar behaves as though it has no upper boundary at all. Interaction designer Bruce Tognazzini, who worked on the original Macintosh, wrote that since the menu bar lies on a screen edge, it has an infinite height. A user can throw the pointer toward the top of the screen without slowing down, because there is nothing past the edge for the pointer to overshoot into.

The same idea explains why operating systems place frequently used controls in screen corners. A corner is pinned on two sides at once, so it stops the pointer regardless of the angle it arrives from. The four corners of a screen are the fastest targets on any display, because of their pinning action, faster even than a straight edge.

Violated: the gap between two monitors

The edge advantage depends on the screen edge actually being an edge. On a desktop with two monitors placed side by side, the boundary between them is no longer a wall: the cursor slides straight through from one screen to the other. A window positioned near that shared boundary loses the stopping effect a single monitor's true edge would have given it, so aiming for a target close to the seam takes the same care as aiming for a target in the middle of a screen. A touchpad with a dedicated strip for scrolling has the same problem in miniature: the strip's inner boundary is not a real edge, so a finger sliding toward it has no way to know it has crossed from the regular tracking surface into the scroll zone until the cursor's behaviour suddenly changes.

051 min

Fitts's Law on touchscreens

Desktop pointing and touchscreen pointing are both covered by Fitts's Law, but they differ in one detail that changes what a designer can rely on: whether there is a cursor that can be physically stopped.

On a desktop, the pointer is a separate object from the hand. It can be thrown toward an edge and left to collide with the screen's physical boundary, which is what makes edge placement free of overshoot risk. On a touchscreen, there is no such intermediary: a finger lands wherever it lands, directly on the glass, and the screen's physical edge does nothing to correct where the finger touches down. Nielsen Norman Group's Raluca Budiu makes the point directly: while the edge placement offers an advantage in mouse- or trackball-driven UIs, it offers no advantage for touchscreens.

The variable that changes between the two cases is the presence of a controllable pointer object separate from the limb doing the pointing. Mouse and trackpad interfaces have one; direct-touch interfaces do not. The lesson for a mobile designer is that pinning a button to the screen edge buys nothing on a phone the way it does on a desktop. The button still has to be genuinely big enough on its own.

061 min

Applying Fitts's Law

A few concrete moves follow directly from the formula.

• Make frequent targets bigger. A primary action such as "submit" or "buy" benefits more from extra size than a rarely used option does.
• Move related targets closer together, and closer to wherever the pointer or finger is likely to already be, such as directly under the text just typed.
• Pin persistent controls to a screen edge or corner on desktop, where the cursor can be stopped by the boundary. This benefit does not carry over to touchscreens, as covered above.
• Add a text label next to an icon rather than relying on the icon alone. The combined icon-plus-label area is larger than the icon by itself, so it is both easier to understand and faster to hit.
• Keep destructive controls, such as "delete account," smaller and farther from the primary action, so a fast, careless movement is less likely to land on them by accident.

None of this means every target should be made as large as possible. A screen where every button competes for maximum size runs out of room fast, and forces more scanning to find the right one — a cognitive load theory problem, not a Fitts's Law one. The moves above work because they are selective: they make the few targets used constantly a little easier to hit, without doing the same for everything else.

071 min

Fitts's Law and accessibility

Fitts's Law gives target size a precise, measurable link to how hard a control is to hit, which is why accessibility standards for pointer input are written directly against target size.

WCAG 2.2's Success Criterion 2.5.8, Target Size (Minimum), says the size of the target for pointer inputs is at least 24 by 24 CSS pixels. A short list of exceptions covers inline targets and a few other cases. A stricter Level AAA criterion in the same standard, 2.5.5 Target Size (Enhanced), asks for at least 44 by 44 CSS pixels. Both numbers exist because the same relationship Fitts described in a research lab in 1954 still holds on a touchscreen or a mouse-driven page: a smaller target takes longer to hit accurately and produces more missed taps.

This matters most concretely for people with a tremor, reduced fine motor control, or low dexterity, for whom a target below the minimum size is not just slower to hit but frequently missed altogether. It also affects anyone using a phone one-handed, or anyone whose hand is less steady at the moment for any reason, including being in a moving vehicle. Checking a design against accessibility guidance for target size is a direct, testable application of Fitts's Law, not a separate concern layered on top of it.

082 min

Fitts's Law vs. nearby concepts

Fitts's Law is most often confused with Hick's Law, because both give a time prediction from a logarithm of some quantity, and both come out of the same mid-twentieth-century information-theory tradition. The difference is what each one is measuring. Fitts's Law predicts how long it takes to move to a target you have already chosen, once its distance and size are fixed. Hick's Law predicts how long it takes to decide which option to choose in the first place, as a function of how many options are on offer. A dropdown menu with too many items is a Hick's Law problem. A button that is too small to hit reliably is a Fitts's Law problem, even if both happen to sit on the same screen.

Fitts's Law is also sometimes confused with the Steering Law, derived for tasks such as navigating a hierarchical pull-down menu, where the user must generate a trajectory with the pointing device that is constrained by the menu geometry. Fitts's Law covers a single, direct move to one target. The Steering Law covers movement that must stay inside a bounded path the whole way, which is a different problem even when both apply to the same menu.

092 min

Where the evidence for Fitts's Law is contested

Fitts's Law itself is not seriously disputed. It has been replicated many times since 1954, across different limbs, devices, and physical environments. What is contested is how reliably it gets measured in practice.

Scott MacKenzie, the same researcher behind the 1992 reformulation, has pointed out that measured throughput for the same device varies far more across published studies than the underlying theory would suggest. TPs range from about 1 bits/s to 60 bit/s across different point-select studies using an ordinary mouse, roughly a sixty-to-one spread for a task that should produce comparable numbers if it were measured the same way each time. MacKenzie attributes the gap to differences in study methodology rather than to any flaw in the underlying theory.

Part of the disagreement is which formula to use for the index of difficulty in the first place. At least four different versions of the formula circulate in the literature: some include the factor of 2 from Fitts's original formula, some use a plus-one or a plus-zero-point-five inside the logarithm, and one drops the logarithm entirely. Studies built on different formulas are not strictly comparable to each other, even when they are nominally testing the same law.

For a practitioner, the unresolved part is measurement convention, not whether bigger and closer targets are genuinely faster to hit. That relationship has held. What has not settled is a single standard way to report how much faster.

102 min

Frequently asked questions about Fitts's Law

?8 questions

Questions people ask

▶3 videos

Watch

• 

  Understanding Fitts's Law

• 

  Fitt's Law Explained

  Explified Labs

• 

  Learn Fitts' Law for strong UX design in Hindi | #xdtutorial #uidesign #fittslaw

  Pelfizz Studio

§9 sources

Sources