Wordle Pattern Engine
Real Green / Yellow / Grey Constraint Filtering
Enter your green letters (correct position), yellow letters (wrong position), and grey letters (not in word). The engine applies all three constraints simultaneously and returns only the 5-letter candidates from 74,872 words that satisfy every constraint. Not affiliated with NYT.
How the Constraint System Works
Green: a green letter at position N means the word's Nth character must be exactly that letter. Five green letters specify the word completely. Any combination of filled positions reduces the candidate pool proportionally.
Yellow: a yellow letter means the word contains that letter at least once, but not in the position where you guessed it. The engine requires the letter to appear somewhere in the word and excludes any word where the letter only appears at the guessed positions.
Grey: a grey letter means the word does not contain that letter at all. Every word containing a grey letter is eliminated. After 2–3 guesses with informative first words, grey constraints alone typically eliminate 60–80% of the starting candidate pool.
All three constraint types are applied simultaneously. The engine returns only words satisfying every constraint you have entered. Enter constraints from multiple guesses combined — there is no limit to how many constraints you can specify at once.
Scope and Limitations
Understanding the Three Constraint Types
The three constraint types — green, yellow, grey — form a complete information system for narrowing five-letter word candidates. Each type encodes a different category of information from a guess. Together, three or four well-chosen guesses can reduce the starting pool of 6,055 five-letter words to a handful of candidates or fewer. The key to efficient solving is choosing guesses that maximise information yield: words containing common letters (E, A, R, I, O, T, N, S) in positions that are most likely to be informative.
Green constraints are the most restrictive. Each green letter eliminates every word that does not have that exact letter at that position. A single green letter at position 3 eliminates all five-letter words where position 3 is not that letter — typically 95% or more of the database. Two green letters at known positions narrow the field dramatically. Five green letters (a correct guess) leaves exactly one candidate.
Yellow constraints carry two pieces of information simultaneously: the letter is present somewhere in the word (inclusion constraint), and the letter is not at the guessed position (position exclusion constraint). This dual nature makes yellow letters more informative than they first appear. A yellow E at position 2 eliminates all words without E, and additionally eliminates all words where E appears only at position 2. The engine applies both halves of the yellow constraint: it requires the letter to be present and excludes the guessed position.
Grey constraints are the broadest eliminator. A grey letter removes every word containing that letter. After two or three guesses, accumulating grey letters typically eliminates 70–85% of the starting candidate pool. The most efficient first guesses test letters that appear in the largest number of five-letter words: E (present in over 60% of five-letter words), A, R, I, O, T. Making all five letters of a first guess high-frequency and non-overlapping maximises the information from grey responses.
Combining constraints from multiple guesses: this engine applies all constraints simultaneously. After two guesses, enter all greens from both guesses, all yellows from both guesses, and all greys from both guesses in a single operation. The result is the intersection of all constraints — only words satisfying every single constraint across both guesses will appear in results. This is more efficient than applying constraints sequentially and re-searching each time.
The tool does not recommend which word to guess next. It only filters to valid candidates given the constraints you provide. The strategy of choosing the next guess from the candidate pool (or from outside it for maximum information) is a human decision. Entering constraints from all previous guesses and then examining the remaining candidates is the standard workflow for efficient solving.
Database Coverage vs Game Word Lists
The 74,872-word database used by this tool is substantially larger than the curated word list used by any specific word game. Wordle draws from a list of approximately 2,300 common English words — a small fraction of the 6,055 five-letter words in this database. After applying constraints from two or three guesses, the remaining candidates from this tool will include valid English words that the game would never select as its answer. The candidate list narrows the search space but does not guarantee the answer is the most familiar word in the remaining set. Consider word frequency and commonness when selecting between candidates.
This tool covers Collins Scrabble Words alongside the TWL, so valid Scrabble words that are not common everyday words will appear in results. Words like AALII, ABAFT, or ABELE may appear as pattern matches but would not appear in a general-knowledge word game's answer set. This is a known database characteristic rather than a bug.