Wordle analyzer: This free Wordle analyzer is a post-game review tool. Enter the known five-letter answer and your guesses in order; it reconstructs duplicate-aware green, yellow, and gray tiles, counts which answers would still fit after every row, measures the observed information, estimates the average branch size of your guess, and offers a one-step alternative. It also checks whether later guesses reused revealed letters. Everything runs locally, so no answer or play history is uploaded. The analysis is independent and is not the official WordleBot.
How the Wordle analyzer reviews a finished game
Wordle analyzer: Start only after the answer is known. Type the answer in the first field, then copy each guess from row one onward without gaps. The Wordle analyzer scores every letter against that answer using the same two-pass logic needed for repeated letters: exact-position matches are assigned first, then misplaced matches consume only the remaining copies. That prevents a word with one E from producing two yellow E tiles. The reconstructed row is therefore the evidence used for filtering, rather than a color pattern guessed from memory.
Wordle analyzer: Before each row, the Wordle analyzer keeps a candidate list consistent with all earlier feedback. It starts from this site's current Core pool of 2,000 answers. A valid entered answer outside that pool is added for this review so the real path does not disappear. After a guess, every candidate producing a different feedback code is removed. This makes the displayed before-and-after counts auditable. It does not claim to reproduce a hidden official answer list, editorial schedule, or proprietary rating.
Enter the completed path
Supply one known answer and one to six accepted guesses in the exact order played.
Rebuild every tile
A duplicate-aware scorer creates the five feedback states that constrain the next candidate set.
Compare the branches
Read actual survivors, expected survivors, observed information, and a local one-step alternative.
What each Wordle analyzer metric means
Wordle analyzer: Candidates left is the observed branch: answers in the current model that return exactly the colors you saw. Expected left asks a different question before the answer is known. The Wordle analyzer partitions every candidate by its possible feedback pattern, squares each bucket size, adds those squares, and divides by the number of candidates. A smaller result means that, on average, the guess creates narrower branches. One-step efficiency compares the model's smallest expected branch with your guess's expected branch and caps the display at 100 percent.
Wordle analyzer: Observed information is log base two of candidates before the move divided by candidates after it. Removing half the possibilities is one bit; shrinking 2,000 candidates to one is about 10.97 bits. A move can receive substantial observed information because the actual answer happened to land in a tiny branch even when its average branch was ordinary. That is why the Wordle analyzer shows observed information beside expected survivors instead of turning both into a fictional luck score. Branch efficiency is descriptive, not a universal grade of player skill.
Wordle analyzer: Observed path
Candidate count and information describe the particular feedback your finished game produced.
Wordle analyzer: Expected branching
Expected survivors compare guesses before anyone knows which candidate is the answer.
Wordle analyzer: Model alternative
The suggested word minimizes one-step expected survivors in the current local candidate set.
Repeated letters, clue reuse, and Wordle analyzer limits
Wordle analyzer: The clue-use check is a practical review, not official hard-mode certification. For every later guess, the Wordle analyzer requires green letters to remain in position, yellow letters to move away from the position where they were yellow, and enough copies of every previously positive letter to be present. It intentionally does not infer every maximum count from gray duplicates, and it cannot know whether a valid guess was strategically forced. A Review label means the entered row did not retain one of those visible positive clues; it does not say the original game rejected that word.
Wordle analyzer: The local model evaluates 9,983 accepted guesses against the surviving Core candidates and chooses only one move ahead. It does not use the exact multi-turn dynamic program, official solution schedule, personal settings, or editorial knowledge. Candidate membership can therefore differ from another service. Treat the alternative as a reproducible comparison under a stated model, not as proof that your move was wrong. Strong post-game review asks what a number represents, which word bank produced it, and whether the same rule is applied to every row.
Local review compared with the published exact model
Wordle analyzer: The analyzer borrows the useful idea of partitioning candidates by feedback, but its product goal and computation are deliberately smaller than the research model.
Wordle analyzer differences from Shortle research
| Dimension | Published Shortle model | This local analyzer |
|---|---|---|
| Wordle analyzer: Candidate answers | 2,315 solution candidates in the documented instance | 2,000 Core answers, plus a valid entered answer when needed |
| Wordle analyzer: Allowed guesses | 10,657 valid guesses in the documented instance | 9,983 accepted guesses in the current site build |
| Wordle analyzer: Feedback space | Up to 243 five-tile patterns | The same 243 encoded outcomes, filtered with duplicate-aware scoring |
| Wordle analyzer: Optimization | Exact dynamic programming over future game states | One-step minimum expected survivors for each reviewed row |
| Wordle analyzer: Output | A full policy and a reported starting word, SALET | Per-turn branches, information, clue reuse, and a local alternative |
Evidence and concrete numbers behind the analyzer
Wordle analyzer: Alex Selby's MIT Sloan paper formulates Wordle as a finite decision problem and documents an instance with 2,315 candidate solutions, 10,657 allowed guesses, and at most 243 possible feedback patterns for a five-letter row. The paper uses exact dynamic programming rather than the one-step calculation on this page. Its reported optimal policy starts with SALET, averages 3.421 guesses, and solves every candidate in no more than five guesses under that stated model.
- Wordle analyzer: A five-tile row has at most 243 encoded green, yellow, and gray combinations before impossible patterns are removed.
- Wordle analyzer: The current local review begins with 2,000 Core answer candidates and searches 9,983 accepted guess words.
- Wordle analyzer: The cited exact model reports a 3.421-guess average and a five-guess worst case for its own 2,315-answer instance.
- Wordle analyzer: Expected survivors use the auditable formula Σ(bucket size²) ÷ current candidates; smaller is better for one-step branching.
Wordle analyzer: Read the MIT Sloan working paper and its documented model, algorithm, word counts, and results.
Wordle analyzer: Read Shannon's original mathematical treatment of messages, choices, and information.
A short source definition of the choice being measured
Wordle analyzer: Shannon frames communication as choosing among possible messages. Here, the possible messages are modeled answers, and a feedback row narrows that set.
“The actual message is one selected from a set of possible messages.”
Claude E. Shannon, A Mathematical Theory of Communication, 1948
Short quotation from the original English paper; its application to this independent game tool is our interpretation.
A reliable post-game analysis checklist
Wordle analyzer: Use the following checks before treating any number or suggested alternative as meaningful.
- Wordle analyzer: Enter the known answer and every guess in order, with no missing middle row.
- Wordle analyzer: Confirm that repeated letters receive only as many green or yellow tiles as the answer contains.
- Wordle analyzer: Separate observed candidates from expected survivors; they answer different questions.
- Wordle analyzer: Compare alternatives only within the disclosed answer pool, guess bank, and one-step objective.
- Wordle analyzer: Use clue-reuse labels as a review prompt, not as an official hard-mode ruling or skill score.
Primary sources and model documentation
Wordle analyzer: These sources support the optimization facts and information concept. They do not endorse this independent analyzer or make its pool official.
- Wordle analyzer: MIT Sloan: The Shortle game and an optimal strategy for Wordle
- Wordle analyzer: Operations Research: The Shortle game and an optimal strategy for Wordle
- Wordle analyzer: Claude E. Shannon: A Mathematical Theory of Communication
- Wordle analyzer: Dont Wordle word-list scope and current playable pools
Wordle analyzer FAQ
Wordle analyzer: Is this the official WordleBot?
No. This Wordle analyzer is an independent local tool with a disclosed Core answer pool, accepted guess bank, and one-step expected-survivor model.
Wordle analyzer: Does the analyzer reveal today's answer?
No. It requires you to enter an answer from a finished game. It does not fetch a daily schedule or predict an unpublished solution.
Wordle analyzer: How are repeated letters scored?
Exact-position matches consume answer letters first. Misplaced matches are then awarded only while another copy remains, preventing extra yellow tiles.
Review the branch, then keep the context
Wordle analyzer: A useful Wordle analyzer makes its assumptions visible. Reconstruct the colors, inspect the actual candidate path, compare expected branches, and check how revealed clues carried forward. Then keep the conclusion proportional: one suggested alternative under one disclosed pool is a learning prompt, not a verdict on the player.