Choose the page by the decision in front of you. The three tools below have deliberately separate jobs: finding candidates after feedback, comparing starters before feedback, and interpreting duplicate-letter colors.
Combine a six-position green pattern, yellow-letter bans, gray letters, and minimum or maximum counts against the 682-word answer pool.
Compare measured starter wordsReview a reproducible 1,485-guess by 682-answer experiment with entropy, expected candidates, and worst-case buckets.
Decode repeated-letter feedbackSee two interactive, count-aware examples and learn when a gray duplicate establishes a maximum instead of an exclusion.
Load vowels earlyBrowse all 401 accepted guesses with three or more vowels, including the 19 four-vowel entries, and see when vowel-heavy openers pay off.
Three intents, three destinations
The candidate solver starts from colors already on the board. It returns only answer-pool records that obey the entered constraints. It is not a blank-board opener ranking.
The starter-word experiment starts with no feedback. It measures how each accepted guess partitions a fixed answer pool and exposes the tradeoff between entropy, expected remainder, and worst-case size. It is not a today’s-answer predictor.
The repeated-letter guide explains the scoring rule itself. Use it when one copy of a letter is green or yellow while another is gray, then carry the proven minimum and maximum into the solver.
Background analysis from the same data
Four reference articles explain the numbers behind these tools. Why 6-Letter Wordle Is Harder Than the 5-Letter Game measures suffix predictability, the 47.8 percent repeated-letter rate, and realized entropy to locate where the extra difficulty actually comes from. Letter Frequency in 6 Letter Words tallies all 4,092 letter slots in the answer pool into overall, first-position, and final-position rankings that opening guesses can be built from. 6 Letter Ending Patterns maps every ending family across the pool and measures a follow-up probe for the 81-word -ER group. Second Guess Strategy simulates what a strong opener leaves behind and settles, bucket by bucket, when a probe beats filtering.
For a known outside letter, browse the starting-letter lists or ending-letter lists. The methodology page records the licensed source, pool fingerprints, formulas, algorithm version, and limitations shared by these tools.
Frequently asked questions
Which page should I use after I have green and yellow tiles?
Use the candidate solver. It combines exact positions, forbidden yellow positions, gray letters, and duplicate-count bounds against the site's answer pool.
Which page compares first guesses?
Use the starter experiment. It compares every accepted guess against every answer before feedback and reports several metrics rather than claiming one universal best word.
Where can I check repeated-letter feedback?
Use the repeated-letter guide for two interactive examples and rules for converting duplicate green, yellow, and gray tiles into minimum and maximum counts.