Forty Words a Minute: What a Typing Test Actually Measures

Ask what a good typing speed is and you will be told about 40 words a minute, which is roughly true and almost entirely misunderstood. The misunderstanding is in the unit. A “word” in typing measurement is not a word. It is five characters, spaces included, and it has been defined that way for about a century.

The convention exists for a good reason. If a word meant a word, a test full of short words would produce a far higher score than the same typing effort spent on long ones, and no two tests would be comparable. Fixing the unit at five characters makes the number mean something across different texts: 200 characters typed in a minute is 40 words per minute whether those characters were “cat dog sun” or “extraordinary”. English averages a shade under five letters per word plus a space, so the figure stays close enough to intuition to be useful.

With that settled, the real numbers are worth knowing. Casual typists on a full keyboard land somewhere around 40 words per minute. The largest study of the question — an online test taken by 168,000 volunteers, analysed by researchers from Aalto University and Cambridge — put the average at about 52, though a self-selecting group of people who choose to take a typing test is likely to skew a little quick. Around 65 to 75 is proficient, the range where typing stops being a bottleneck for most office work. The top one percent of that study exceeded 120. Competitive typists run higher still, and the most-cited record belongs to Barbara Blackburn, clocked at over 200 words per minute in short bursts on a Dvorak layout.

So if you score 45, you are unremarkable in the most literal sense, and it is almost certainly not what is slowing you down at a keyboard.

Accuracy is the part people underrate, because the arithmetic is unforgiving. Gross speed is simply characters divided by five divided by minutes. Net speed subtracts your mistakes, and the subtraction is larger than it looks: an error does not merely cost you the wrong keystroke, it costs the backspace, the retype, and the half-second of attention spent noticing. A typist at 70 gross with sloppy accuracy routinely finishes behind a typist at 55 who rarely has to go back. Backspace is the most expensive key on the keyboard.

The other received wisdom worth examining is touch typing. The claim is that proper technique — eight fingers on home row, eyes on the screen — roughly doubles your speed over hunt-and-peck, and as a general direction that holds. But the large study found the picture messier than the classroom version. Only about half the participants had ever taken a typing course, and plenty of self-taught typists using six fingers or fewer were reaching perfectly respectable speeds. What separated the fast from the slow was less about how many fingers moved and more about rollover — beginning the next keystroke before the last one has finished — and about consistent finger-to-key habits, whatever those habits happened to be.

That is an encouraging finding if you never learned properly. Your unorthodox technique is probably not your ceiling. Looking at the keyboard is.

While we are dismantling things, the QWERTY story deserves a correction. You will often hear the layout was designed to slow typists down. It was not, quite. Early typewriters jammed when neighbouring typebars were struck in quick succession, so common letter pairs were separated to keep the bars from colliding — the goal was to prevent jams, which in practice let people type faster rather than slower. The layout is a compromise shaped by mechanical constraints that vanished a century ago, and we kept it because everybody already knew it.

Dvorak, the usual proposed replacement, is a genuine improvement on paper, with more typing on the home row. The famous studies showing large gains were run by Dvorak himself, and later re-examinations found the advantage much smaller than advertised. The honest summary is that switching layouts costs weeks of frustration to buy a few percent, which is why almost nobody does it and why it does not matter much that they don't.

Phones are where the interesting change has happened. Thumb typing used to be dramatically slower than a keyboard, and it is closing: research from the same group found average mobile speeds in the high thirties, with young, heavy users at the top of that range and not far off what many people manage on a full keyboard. Predictive text and autocorrect do a share of the work, which is either cheating or using the tool as designed, depending on your mood.

There is one more thing every typing test hides, and it is the reason your score flatters you. A test measures transcription — the words are given to you, they are common, they are correctly spelled, and you are not deciding anything. Real typing is composition. You are choosing words, catching yourself, reformatting, tabbing between fields, and typing proper nouns and technical terms that no predictive system anticipates. Almost nobody writes at their test speed, and the gap is not a failure of the test so much as a reminder of what it measures: your hands, not your head.

Which brings us to the version on CerdikCerdik, and this one needs a caveat larger than usual. The test runs for 60 seconds and gives you words one at a time from three lists — three-letter words on Easy, five-letter on Medium, six and seven-letter on Hard. Your score is words genuinely completed per minute, and those are real words rather than five-character units.

That means the number here is not directly comparable to a standard WPM score, and, more importantly, not comparable between difficulties. Forty short words on Easy is about 160 characters, which a conventional test would report as roughly 32. Forty words on Hard is nearer 280 characters, or about 56. Same score on screen, a difference of almost double in the standard unit. It is why the leaderboard keeps a separate table per difficulty, and why comparing your Easy best to your Hard best tells you nothing at all.

The other design decision is that accuracy is not reported, because it cannot go wrong. A word is only accepted when it matches exactly, so a typo simply leaves you stuck on the word until you fix it. Accuracy is a gate rather than a score. That makes the game a cleaner measure than a test that lets you race through errors and docks you afterwards — but it also means the clock is punishing typos the whole time you are hunting for the mistake, which is the real lesson about backspace delivered without a lecture.

If you want to get faster, the advice is dull and it works. Slow down until you stop making errors, because speed built on corrections has a low ceiling. Keep your eyes on the screen even when it feels slower, since looking down is the habit that caps you. And accept that improvement is not linear: most people jump quickly at the start, then sit on a plateau for a long time before anything moves again.

Try a round at cerdikcerdik.com/games/speed-typing — pick one difficulty and stay on it long enough to have a real baseline, rather than collecting three scores that mean three different things. If you would rather measure your hands than your habits, cerdikcerdik.com/blog/average-human-reaction-time covers the part of your speed you genuinely cannot train.