Word Length Research
Distribution From 2 to 15 Letters — 74,872 Words
74,872 words analyzed. Peak at 8 letters (11,976 words). Average: 8.11 characters. Fingerprint: 916f47949b41. Generated: 2026-07-27.
Length Distribution Table — 2 Through 15 Letters
Each word counted by Python len(). All 74,872 words included. Sorted by length ascending. Percentage of total database shown for each length.
| Length (letters) | Word count | % of database | Distribution |
|---|---|---|---|
| 2 | 301 | 0.4% |
|
| 3 | 1,040 | 1.39% |
|
| 4 | 3,181 | 4.25% |
|
| 5 | 6,055 | 8.09% |
|
| 6 | 9,234 | 12.33% |
|
| 7 | 11,882 | 15.87% |
|
| 8 | 11,976 | 16.0% |
|
| 9 | 10,509 | 14.04% |
|
| 10 | 8,269 | 11.04% |
|
| 11 | 5,626 | 7.51% |
|
| 12 | 3,516 | 4.7% |
|
| 13 | 1,961 | 2.62% |
|
| 14 | 899 | 1.2% |
|
| 15 | 423 | 0.56% |
|
Methodology: Python len() on every word in the 74,872-word database. Fingerprint: 916f47949b41. Generated: 2026-07-27.
Distribution Shape — Right-Skewed Peak at 8 Letters
The length distribution is unimodal with a peak at 8 letters (11,976 words, 16.0% of the database). The distribution is right-skewed: the tail extends toward longer words rather than shorter ones. This shape reflects the morphological structure of English vocabulary — there are relatively few stable root words shorter than 5 letters, while the productive suffixes (-NESS, -TION, -ATION, -MENT) push many words into the 8-12 letter range.
The decline from the peak toward shorter lengths (2–5 letters) is steeper than the decline toward longer lengths (9–15 letters). This asymmetry occurs because the short end of the distribution is bounded (words shorter than 2 letters do not exist as dictionary entries), while the long end has a biological upper limit (human short-term memory and articulatory capacity) rather than a hard linguistic boundary.
For Scrabble players, the distribution shape matters because it determines the probability of finding a valid word from a given rack. With a 8-letter rack (after drawing 7 tiles), the single most likely word length you can form from subsets of your tiles is also in the 7-9 letter range — exactly where the database density peaks. This is one reason the Scrabble rack size of 7 tiles is well-calibrated to English vocabulary structure.
Practical Implications for Word Game Players
Word Length and Vocabulary Type — Short, Medium, and Long Word Zones
The database distributes into three natural zones based on word length and vocabulary type. The short zone (2-4 letters, 4,522 words, 6.0% of database) consists of two-letter Scrabble words (AA, QI, ZA), three-letter words (THE, AND — functional words), and four-letter root words (PLAY, BOOK, FISH). These short words are common in everyday English and essential as Scrabble hooks and parallel plays.
The medium zone (5-9 letters, 44,955 words, 60.0% of database) is the most productive zone for word games. This zone contains single-root words with one productive suffix or prefix: PLAYER (6), PLAYING (7), PLAYERS (7), PLAYFUL (7), PLAYBOOK (8), PLAYFULLY (9). The 7-letter segment of this zone (11,882 words) is the Scrabble bingo target — all 7-letter plays earn the 50-point bonus, and 11,882 candidates represent an enormous play space.
The long zone (10-15 letters, 25,395 words, 33.9% of database) contains words with multiple morphological extensions: PLAYFULNESS (11), UNPLAYFULLY (11), SPORTSMANSHIP (13), INTERCOLLEGIATE (15). These words are common in Collins Scrabble Words (which includes more technical and extended vocabulary than TWL). The long zone is used in Scrabble when playing through multiple existing board tiles to form extensions.
The even distribution across zones (6%, 60%, 34%) reflects the combined database composition. A TWL-only database would show lower long-zone representation because Collins adds disproportionately many long, technical, and archaic words. A general everyday English vocabulary database would show higher short-zone representation because common conversational words (THE, IS, ARE, AT, IN) would dominate. The current balance reflects Scrabble-weighted vocabulary combined with British English supplement.
For crossword construction, word length distribution is a critical design constraint. A typical crossword grid contains many 5-7 letter words (the medium-length zone where most crossing opportunities arise) with some long theme answers (10-15 letters) and short connective words (3-4 letters). The database length distribution maps well onto standard crossword grid structure, making it a useful reference for crossword constructors as well as solvers.
Methodology — How Length Distribution Was Computed
Word length distribution was computed by applying Python len() to every word in the 74,872-word database. Words were grouped by character count using a Counter object. The result is a dictionary mapping each integer length (2 through 15) to the number of words at that length. Percentages are each count divided by 74,872. The average word length is the sum of (length × count) for all lengths, divided by the total word count: sum(l * c for l, c in distribution.items()) / 74,872 = 8.11 characters.
Limitations: character count is not equivalent to syllable count. Eight-character words may be monosyllabic (STRENGTHS, 9 chars, 1 syllable) or polysyllabic (CALENDAR, 8 chars, 3 syllables). The distribution shows character-length, not morphological complexity. Additionally, the database was not filtered by frequency — a word appearing once in a corpus contributes equally to the distribution as a word appearing a million times. This vocabulary-frequency approach gives longer, rarer words equal weight with shorter, common words.
All figures are reproducible from fingerprint 916f47949b41 by running the computation against the same wordlist.js file used by the live tools. The data table on this page is the definitive source for word length statistics from this database version.
Summary — Key Length Statistics
Key figures: peak at 8 letters (11,976 words, 16.0%), average 8.11 characters, range 2–15 letters, total 74,872 words. Short zone (2-4 letters): 4,522 words (6.0%). Medium zone (5-9 letters): 44,955 words (60.1%). Long zone (10-15 letters): 25,395 words (33.9%). All data from fingerprint 916f47949b41, generated 2026-07-27.
Distribution Comparison — Database vs Everyday English
The 8.11-character average in this database is notably higher than everyday spoken English (average approximately 4.5 characters in conversational speech) and written English (approximately 5.1 characters in news text). The gap is explained by the database composition: Collins Scrabble Words includes many long technical and archaic words absent from everyday usage. Common function words (THE, IS, ARE, IN, ON, AT) are present in the database but are outnumbered by the large number of long morphologically-extended forms that make up the majority of Scrabble vocabulary. The 8-letter peak reflects the productivity of English derivational morphology — adding one suffix or prefix to a common 5-7 letter root typically produces an 8-10 letter word, which is exactly where the database is densest.