Typing speed test

The five-character word is older than the personal computer and is a convention rather than a measurement. MacKenzie’s survey of text entry evaluation dates it precisely: "it has been a convention since about 1905 to standardize the computation of text entry speed in words per minute, where a word is defined as five keystrokes", citing Yamada, and noting that this "includes letters, spaces, punctuation, and so on". Fixing the yardstick is what makes one typist comparable to another, and it also means the same typist scores lower on German than on English, because German words are longer while the five-character word is not.

Scoring here is positional, and that has a consequence worth understanding before you read your result. Your text is compared against the target one character at a time, in place: position one against position one, position two against position two. Miss a single letter and every character after it is offset by one, so the rest of the line is marked wrong even though you typed it correctly. That is not a quirk to work around; it is why the test rewards fixing an error immediately. Backspace restores the alignment, and the two or three keystrokes it costs are cheaper than a whole line scored as errors. A word-level test would forgive the drift, which is one reason results differ between typing sites.

The two speeds answer different questions. Raw counts every character sitting in the box and divides by five; net counts only the ones that landed in the right place. Neither is a keystroke tally: backspace deletes a character and it leaves both counts, which is why correcting an error costs you nothing but the time. A large gap between them means the keyboard is not the limit, your correction rate is, and the counterintuitive fix is to slow down: the net figure usually rises. A small gap at a low speed means the opposite, that accuracy is already fine and speed needs practice rather than care.

The words are drawn from a list of 236 common English words, shuffled fresh each run, with 140 of them dealt per attempt. Common-word tests measure typing rather than reading, which is the point: a passage of unfamiliar prose measures how fast you can parse it as much as how fast you can type it, and that is a different skill. It also means this test flatters anyone used to typing common English and does not reflect how you would fare on code, on names or on anything with punctuation and capitals in it.

Two things the number cannot tell you. It measures a burst, not a working day, and sustained typing over an hour runs well below a thirty-second peak for almost everybody. And it says nothing about how you type: a fast hunt-and-peck typist and a slow touch typist can score identically while having entirely different ceilings, because only one of them can improve without looking at the keyboard. If the goal is to get faster, the accuracy figure and the layout you use are more informative than the headline speed.