Readability formulas estimate how hard a text is to read by measuring two things: how long the sentences are, and how long or syllabically complex the words are. That is the whole mechanism. Understanding that is what stops the score being misused, because it explains both why the numbers are useful and why they can be badly wrong.
The formulas
Flesch Reading Ease returns a score from 0 to 100, where higher is easier. Roughly: 90–100 is very easy (around age 11), 60–70 is plain English, 30–50 is difficult, and below 30 is very difficult. Most general-audience writing aims for 60–70.
Flesch–Kincaid Grade Level converts the same inputs into a US school grade. A score of 8 means an average American 13-to-14-year-old should manage it. General-audience writing usually targets grade 8–10.
Gunning Fog estimates the years of formal education needed, weighting “complex” words of three or more syllables. Business writing typically aims below 12.
SMOG predicts the education level needed for full comprehension. Widely used in healthcare, where it is often the required measure for patient materials.
Coleman–Liau and Automated Readability Index use characters per word instead of syllables, which makes them more reliable for machine processing because character counting is unambiguous while syllable counting is not.
They generally agree with each other. Where they diverge sharply, it is usually because the text contains many long words that are nonetheless familiar, or short words that are not.
What they do not measure
This is the important part.
Meaning. A formula cannot tell whether a sentence makes sense. “The cat sat on the quantum” scores as very easy.
Word familiarity. “Cat” and “gnu” are both one syllable. So are “run” and “verb”. A text full of short but obscure words scores as easy and reads as impossible.
Structure and flow. Whether ideas are ordered sensibly, whether paragraphs connect, whether the reader is given context before the thing that needs it — none of this is visible to the formula, and all of it matters more than sentence length.
Audience knowledge. A cardiology paper reads easily for cardiologists at any Flesch score.
Jargon. Domain terms are frequently short.
The consequence is that you can improve a score without improving the writing — chop sentences in half at arbitrary points and swap a few long words for short ones, and the number rises while the prose gets worse. Treat the score as a prompt to reread, never as a target.
Where people go wrong
Optimising for the number. The most common failure. A score of 60 achieved by butchering sentence rhythm is worse than a score of 45 that flows.
Applying English formulas to other languages. These were built and validated on English. German compounds and Finnish agglutination break the syllable assumptions completely, and the output is a plausible-looking number that means nothing.
Scoring the wrong text. Headings, bullet lists, code blocks and tables distort the result, because the formulas assume continuous prose. Score the body copy.
Scoring very short samples. Below about 100 words the numbers are unstable — one long sentence can swing the grade level by several years. Use at least a few paragraphs.
Assuming lower is always better. Writing for a specialist audience at grade 6 reads as condescending. Match the audience, not the minimum.
Using it well
Run the score, then read the flagged long sentences and ask whether each one is carrying more than one idea. Usually it is, and splitting it at the join improves both the prose and the number. That is the useful loop. The word counter covers length; this covers density.
Your text is processed entirely in your browser and never transmitted — see the privacy policy.