Sentence Length & Passive Voice Checker

How it works

Average length = words ÷ sentences; long-sentence share = sentences over the threshold ÷ sentences; passive estimate = sentences containing "to be" + past participle; adverb density = -ly words ÷ words

Readability formulas summarise a text in one number; editors instead look at the distribution behind it. A passage with a comfortable average can hide a few 50-word sentences that lose every reader, so the share of long sentences and the longest offenders are listed for surgery. Passive voice is detected the way grammar checkers do it — a form of "to be" followed by a past participle — which catches most cases and a few false positives ("was tired"); a passive rate above 10–15% in ordinary prose usually means the writer is hiding who did what. Adverb density and repeated words are the remaining quick tells of padding. None of these is a rule: a long sentence can be lucid and a passive can be exactly right, but the numbers show where to look.

Worked example

The sample has five sentences averaging about 17 words, one of them over 25 words, one containing passive constructions ("was written by", "were presented", "had been informed") — a 20% passive rate — and three -ly adverbs in 83 words, so the last two sentences are the ones to rework.

Frequently asked questions