How the Text Statistics Dashboard Works
AnyTool Text Statistics analyses text entirely in the browser. It tokenises words with a Unicode-aware regular expression, counts sentences by terminal punctuation, paragraphs by blank lines, and estimates syllables with a vowel-group heuristic. From a single pass over the tokens it derives counts, averages, lexical diversity, distributions and every readability score, so the full report updates live as you type.
- Words are Unicode letter/number runs; non-English text is counted correctly
- Sentences split on . ! ? and ellipses; paragraphs split on blank lines
- Syllables estimated with a vowel-group heuristic, as in most readability tools
- Reading time at 225 words per minute, speaking time at 130 words per minute
- All computation is client-side JavaScript — no server round-trip
Multiple Readability Formulas and a Consensus Grade
The tool computes six readability indices: the Flesch Reading Ease score (0 to 100, higher is easier) and five grade-level formulas — Flesch-Kincaid, Gunning Fog, SMOG, Coleman-Liau and the Automated Readability Index. The grade formulas are averaged into a single consensus grade with a plain-language band so writers can target a reading level. SMOG is most reliable from around thirty sentences.
- Flesch Reading Ease with a plain-language band label
- Flesch-Kincaid, Gunning Fog, SMOG, Coleman-Liau and ARI grade levels
- Consensus grade averages the five grade formulas
- Gunning Fog discounts words that are complex only due to -es, -ed or -ing suffixes
- Each score maps to a US grade and reading-age interpretation
Distributions, n-grams and Lexical Diversity
Beyond readability, the dashboard charts the word-length and sentence-length distributions and breaks characters into letters, digits, punctuation, whitespace and case. It ranks the most frequent words and the top bi-grams and tri-grams, and reports lexical diversity through the type-token ratio, a length-stable root TTR, and the count of hapax legomena. Charts are drawn as lightweight in-page bars, and large inputs are debounced so the editor stays responsive.
- Word-length and sentence-length histograms plus a character-class breakdown
- Top words and top bi-grams / tri-grams, counted within sentence boundaries
- Type-token ratio, root TTR and hapax-legomena counts for vocabulary richness
- Longest, shortest and hardest sentences are surfaced automatically
- Report can be copied or downloaded as .txt or .csv — nothing is uploaded
