Anchor, instantly. Gravity, via a falling apple. Nonetheless — there’s nothing to draw.
Put a word on a game tile — or a flashcard, a button, a sign — and you’ll soon want a little picture on it. Some words are pictures — clock, dog, umbrella. Others need a metaphor (an idea is a lightbulb), and a large share of English defeats any artist, amateur or professional. We scored 598,677 words and phrases from 0 to 1 for how well each one becomes a small, unmistakable icon.
Behind each score sits a rubric — could a 7-year-old draw it? could a pro? would the sketch be mistaken for something else? — plus a one-line brief for the artist. Here is an illustration of higher vs. lower iconability scores. Lower scores are dimmer:
Among the 24,089 most familiar words, the median score is 0.048 — half of everyday English sits in the bottom band. Concrete nouns crowd the top of the scale; grammar and abstraction hold the floor.
The scores combine two views: whether a word already lives in the major public icon libraries, and a nine-part judged rubric covering drawability by age, professional feasibility, ambiguity, and how the image survives at tile size.
If your game art pipeline starts from a word list, this column is the production schedule: which tiles deserve commissioned art, which need a metaphor, and which should stay text. And the score answers the broader question — can this concept be represented visually? — wherever it comes up: kids’ and accessibility interfaces (words a 7-year-old can draw are words a pre-reader can recognize), flashcards and language learning, signage and wayfinding, generative-art prompts.
Iconability ships as a per-word signal in the scores table, alongside game-suitability and the accessibility profile. See real rows from every table, or take the whole thing — it’s public domain.