Does Typing Speed Drop Late at Night

Quick answer

> Quick answer: Yes, and the drop is biggest between roughly two and five in the morning. Afternoon and early evening are where most people are quickest. Early morning is worse than people expect.

The pattern follows reaction time rather than tiredness, so you feel fine while typing badly.

I read the research on daily performance rhythms before writing this. The gap between the worst hours and the best is larger than most gear differences we've measured.

Typing is a reaction time task

That's what explains everything below.

Touch typing isn't thinking. It's a fast chain of see a letter, move a finger, repeated a few hundred times a minute.

So it follows the same daily curve as any reaction time measurement rather than the curve for concentration or creativity.

Which means the hours when you feel sharp mentally aren't necessarily the hours your hands are quickest.

The shape of the day

Performance on reaction and attention tasks is lowest in the small hours, climbs through the morning, and is strongest through the afternoon and into the evening.

Research on daily rhythms in attention puts the worst window at roughly two to five in the morning and the best in the afternoon.

That's a long way from the popular idea that mornings are always best.

Mornings are best for some things. Not for this one.

Why two in the morning is so bad

Two things stack at that hour.

Your body clock is at its lowest point, which slows reaction time on its own even in somebody who slept well earlier.

And if you're awake at that hour, you've usually been awake a long time, so tiredness adds on top.

The two effects are separate and they multiply. That's why a late night session feels much worse than one hour of lost sleep should explain.

The early morning problem

Sleep inertia is the grogginess in the first half hour or so after waking, and it hits reaction speed hard.

Typing a test five minutes after getting out of bed produces one of the worst numbers you'll ever record.

It clears on its own, and the usual advice of an hour is about right for most people.

So if you have a typing test at nine, don't get up at ten to nine.

Five minutes of easy typing does most of the clearing. It isn't warming up your hands so much as waking up the part of you that reacts.

The dip after lunch

Most people have a real slump in the early afternoon, and it isn't only the food.

Human daily rhythms have a second smaller cycle within the day, which is why a dip appears in the early afternoon even in people who ate nothing.

A heavy lunch makes it worse rather than causing it.

If you want an afternoon test, take it later rather than immediately after eating.

How big is the difference

For most people it's a few words a minute between their worst normal hour and their best, and considerably more if the worst hour is the middle of the night.

I compared that against the hardware gaps collected in the garage, and the hour of the day wins against most of them.

And unlike a keyboard it costs nothing to fix.

You just take the test at a different hour.

Your chronotype shifts the window

The curve is the same shape for everybody. Where it sits on the clock isn't.

A genuine early riser peaks a couple of hours before a genuine night person. Chronotype is the word for that difference, and it's largely not a choice.

So generic advice about the best hour is nearly useless. The shape is general. The timing is yours.

Measure your own rather than trusting an article, including this one.

Finding your own peak

Take five tests at the same hour, three days running, then repeat at a different hour.

Four time slots covers a day well enough. Mid morning, early afternoon, early evening, late evening.

Compare the middle value from each slot rather than the best.

It takes under an hour of testing spread over a fortnight, and the answer is worth having, because you'll use it every time your speed matters.

Write each slot down as you go. Doing this from memory a fortnight later produces a confident answer that happens to be wrong.

Keep everything else fixed

Same keyboard, same chair, same test length, same room.

If you test in the morning on a laptop in the kitchen and in the evening on a desktop at your desk, you're measuring the desks.

Caffeine has to be controlled too. Testing at nine with coffee and at nine in the evening without it compares the coffee, not the hour.

We went through the caffeine effect separately in our caffeine piece.

Sleep debt flattens the whole curve

A well slept person has a clear peak and a shallow trough. Somebody running on five hours a night has a lower peak and a much deeper trough.

The rhythm survives sleep loss. It just moves the whole line down.

So the best hour of a badly slept week is often worse than the worst hour of a well slept one.

If your speed matters this week, sleep matters more than the clock.

Room light changes it too

Bright light in the morning pulls your body clock earlier, and bright light late at night pushes it later.

Which matters if you type at night in a dark room lit only by a screen. That light is enough to delay the clock over weeks and make your mornings worse.

It's a slow effect, not a same night one.

A lamp on in the evening room and daylight in the morning does more for next month's typing than any drill.

Accuracy falls before speed does

The same pattern as fatigue, cold hands and everything else that degrades typing.

At a bad hour your rate often looks close to normal while your error count doubles. The raw number hides it and the corrected number doesn't.

So judge the time of day question on accuracy, not on the headline figure.

We set out the two scores in gross versus net.

What this means for a job test

If you get to choose the slot, pick mid to late afternoon, and pick a weekday when you slept normally.

If the test is at eight in the morning, be awake by seven and type something for five minutes first.

If it's an online test you take at home, don't take it at midnight because that's when you got round to it.

That one decision is worth more than a fortnight of practice.

Practising at a bad hour is still useful

Nothing here says avoid typing at night.

Drills at a bad hour still build the movement. They just produce worse numbers, and the numbers are the only part affected.

So practise whenever you can. Measure when you're at your best.

Separating those two things is the whole trick, and it's the same rule as testing versus training generally.

Racing at night

Late night races are where most online typing happens, because that's when people are free.

You're slower, and so is everybody else, so the competition is fair even if the numbers are lower.

Don't read a late night race result as your speed dropping. Read the field.

A race at midnight and one at six in the evening are different events with the same rules.

The leaderboard shifts with the hour too, so a rank at midnight and a rank at teatime are not the same achievement.

The short version to act on

Test in the afternoon or early evening.

Don't test within an hour of waking.

Don't test between two and five in the morning and then worry about the result.

Find your own peak with four time slots over a fortnight.

And if you only change one thing, stop taking tests at midnight and reading the number as if the hour didn't matter.