How a Typing Speed Test Actually Measures You
The clock in a typing test starts on your first keystroke rather than when the page loads, and that one decision is worth several WPM on a short run. Everything else that separates two tests of the same passage comes from a similar list of small choices, each defensible on its own, none of them visible on the result screen.
What starts and stops the timer
Starting on the first keystroke means the time you spend reading the passage, settling your hands or finding the text box costs you nothing. A test that started at page load would measure your reading speed as well as your typing.
Stopping is decided by the mode. A fixed 15, 30 or 60 second run ends when the time is up, wherever you are in the text. A full text run ends when you reach the last character, so its duration is your own. The final figure uses the elapsed milliseconds actually recorded, not the nominal length, which is why a 60 second run occasionally reports 61 seconds.
What backspace does, and what it does not undo
Backspace is where most of the confusion lives. In our typing test it moves the cursor back one position, clears the verdict on that character so a retype can count as correct, and adds one to a separate corrections counter.
What it does not do is return the time you already spent, and it does not remove the error from your accuracy. A mistake stays in the accuracy figure even after you repair the text. Accuracy is calculated as keystrokes minus total wrong characters, divided by keystrokes, so it describes the whole session rather than the final state of the paragraph.
That gives correction a specific price. Fixing one character costs one backspace, one retype and the pause before both, roughly three keystrokes of time for one character of output. Hammering backspace is therefore expensive twice over, which is the mechanism behind a plateau that looks inexplicable on paper.
Raw against corrected
Two figures come out of the same run. The corrected figure counts only the characters that matched the source, divided by five, divided by the minutes elapsed. The raw figure counts every character you entered, right or wrong, through the same arithmetic.
Raw is always the higher number and it describes hand speed. The corrected figure describes produced text. A wide gap between the two means the hands are ahead of the eyes, which is the most common finding on a first measurement and the one that responds fastest to slowing down slightly.
The relationship between these figures and the units employers publish is set out in what WPM means and in CPM and keystrokes per hour.
Six reasons two tests disagree on the same text
- Character comparison against word comparison. Character by character marking counts one wrong letter as one error. Word level marking condemns the whole word, so the same typing produces a lower score.
- Whether wrong input is allowed. Some tests refuse the keystroke until you type the right character, which prevents a run of errors and quietly raises accuracy.
- Whether the trailing partial word counts. Ending a 60 second run mid word can add or drop a fraction of a word.
- Which characters are counted. Correct characters, all characters or completed words are three different denominators.
- Whether paste and autocorrect are blocked. Ours blocks pasting outright, and treats composed input from an accent sequence as text rather than as a shortcut.
- The passage itself. Vocabulary, punctuation density and capital letters change a result more than most people believe.
That last point deserves a number. Our corpus holds Whitman's There was a child went forth every day at 225 characters of short repeated phrases, and Irving's page on going into the country at 499 characters of semicolons and clauses inside clauses. The same typist will not produce the same figure on both, and neither result is the wrong one.
Before you blame yourself
Repeat a test on identical text and you measure familiarity with that text, not skill, because the second run benefits from knowing what is coming. Comparing across durations does the same thing in reverse.
If one particular character keeps failing rather than one particular finger, the problem may not be in the measurement at all. A key that misses presses or repeats them shows up clearly in the keyboard test, and it is worth ruling out before adjusting your technique.
Then take one deliberate run at something demanding. Michael Faraday's account of the shadow of a candle flame, taken down at his Christmas lectures and printed in 1861, packs 460 characters into a chain of clauses held together by semicolons, so punctuation has to be placed exactly. Washington Irving's page on going into the country, from The Sketch Book of 1820, runs 499 characters on semicolons alone. Beside them, Walt Whitman's There was a child went forth every day, from the first Leaves of Grass in 1855, is 225 characters of repeated plain phrases, which is the other end of the same scale. Typing one after the other shows, more clearly than any single passage, where your rhythm goes when a sentence stops being predictable. All three are waiting in the typing test, each with its author, year and source on its own page.