Why "14 Minutes Left in This Chapter" Is Usually Wrong
“14 minutes left in this chapter” is a useful thing to know at 23:40. It is only useful if it is true, and in most apps it is a guess dressed as a measurement.
The usual method is to count the words remaining and divide by a fixed reading speed — commonly 250 words per minute, a figure from studies of adult readers of English prose. Applied to you, reading this book, tonight, it is wrong in at least three directions at once.
Why a fixed number cannot work
People differ enormously. Comfortable adult reading speeds span roughly 150 to 400 words per minute. A single constant is wrong for almost everyone by a wide margin.
The same person differs. You read a thriller faster than a philosophy paper. You read faster at 10:00 than at 23:00. You read slower in a second language, and slower again on a bus.
The unit does not travel. Words per minute is meaningless for Japanese or Chinese, where text is not delimited by spaces. Counting “words” in a language without word spacing produces a number that is not wrong so much as unrelated to anything.
Measuring instead of assuming
Aurora Reader estimates from your own history. It records reading sessions — how long you read and how many pages you got through — and derives your actual pace from them.
Two details in how it does this are worth knowing, because they explain the behaviour you will see.
Sessions under 30 seconds are ignored. Opening a book to check something is not reading, and counting it would push your apparent speed toward infinity.
Before there is data, it assumes 0.5 pages per minute. That is a deliberately conservative placeholder, not a claim. It is why the first estimates you see are wrong, and usually pessimistic.
So the honest description is: the first few sessions produce a bad estimate, and it converges. If you have just installed the app, the number on screen is a default, not a measurement of you.
What throws it off later
Skimming. Twenty minutes of flipping through to find a passage tells the average you read at four times your real pace, and the estimates get optimistic for a while.
Leaving the book open. Reading for ten minutes and leaving the app in the foreground for forty makes you look slow. Sessions measure elapsed time with the book open, not attention.
Switching genres. A pace learned on novels does not describe a technical manual. The average adapts, but it lags by a book or two.
None of these are avoidable by the app. They are the price of measuring rather than assuming, and the measurement is still much closer than a constant.
What the number is good for
It answers one question well: can I finish this chapter before I have to stop? At a train stop, at bedtime, before a meeting, that is genuinely worth having.
It answers a different question badly: how much have I achieved? Time with a book open is not comprehension, and a book you read slowly because it was worth slowing down for is not a worse result than one you finished quickly. Reading statistics measure the easy thing because the interesting thing cannot be measured.
Treat the estimate as a train timetable and not as a report card, and it earns its place on the screen.