One word, two opposite objectives, one disclosed local model

Wordle Starting Word Ranker

Test any accepted five-letter guess, compare two openers, and see how the same word ranks for ordinary Wordle solving and Dont Wordle survival.

Ordinary WordleDont Wordle

Rank any five-letter starting word

Compare one or two accepted words across ordinary Wordle and Dont Wordle objectives.

Enter any accepted five-letter guess.
Optional. Add a second word for a direct comparison.

Your words stay in this browser. The ranking file is calculated during the site build.

Independent local model: 2,000 Core answers and 9,983 accepted guesses, not official Wordle or WordleBot data.

Wordle starting word ranker: This free Wordle starting word ranker measures every accepted guess against the same 2,000-answer Core pool. Enter one word or compare two. For ordinary Wordle, a smaller expected remaining set is better because it narrows the search. For Dont Wordle, a larger expected set is better because it preserves more possible answers and delays an accidental solution. The result includes both ranks, average and worst branch sizes, Shannon entropy, observed feedback patterns, colored-clue probability, distinct-letter count, and immediate target risk. All 9,983 rows are precomputed during the site build; your input stays in the browser. This is an independent model, not official Wordle or WordleBot data.

How the Wordle starting word ranker measures an opener

Wordle starting word ranker: The calculation asks what would happen if each of the 2,000 modeled answers were hidden. For one proposed guess, the Wordle starting word ranker scores every answer with duplicate-aware green, yellow, and gray logic. Answers that produce the same five-tile response enter the same bucket. A strong ordinary opener makes many small buckets; a branch-preserving Dont Wordle opener deliberately leaves fewer, larger buckets. Because both views use the identical answers and feedback function, the contrast comes from the objective rather than from changing the evidence.

Wordle starting word ranker: Expected answers left is computed as the sum of every bucket size squared divided by 2,000. That is the average branch a player enters before knowing which feedback will appear. Worst remaining is the largest single bucket. Entropy summarizes how evenly probability is spread across observed patterns, while pattern count reports how many of the theoretical 243 feedback codes actually occur in this local pool. The Wordle starting word ranker reports all four rather than compressing different ideas into an unsupported universal score.

1

Wordle starting word ranker: Choose one or two words

Wordle starting word ranker: Enter any five-letter word accepted by this site's current guess bank, then add an optional comparison.

2

Wordle starting word ranker: Read both objectives

Wordle starting word ranker: Use the lower expected branch for ordinary solving and the higher expected branch for Dont Wordle survival.

3

Wordle starting word ranker: Check the model boundary

Wordle starting word ranker: Confirm Core membership, target risk, word-bank size, and methodology before treating a rank as advice.

What each starting-word metric says—and does not say

Wordle starting word ranker: Solve rank orders all 9,983 accepted guesses by expected remaining answers from smallest to largest, then uses worst branch and entropy as transparent tie-breakers. Branch-preserving rank first favors words outside the 2,000-answer Core pool, because such a word has zero modeled chance of ending the round immediately. It then orders expected remaining answers from largest to smallest. The Wordle starting word ranker therefore describes two explicit one-move objectives. It does not predict editorial answer schedules, personal vocabulary, keyboard speed, or the complete value of later turns.

Wordle starting word ranker: A colored-clue probability is the share of modeled answers that produce at least one green or yellow tile. That value is not automatically good or bad. Ordinary solvers usually welcome clues, yet the exact size of the resulting branch matters more than color alone. Dont Wordle players may prefer an all-gray result that leaves many candidates, but they must still follow the revealed constraints on later turns. Distinct letters is similarly descriptive: five unique letters often sample more of the alphabet, while repeated letters can intentionally reduce information. The Wordle starting word ranker keeps these facts separate from its two ranks.

Wordle starting word ranker: Expected branch

Wordle starting word ranker: The probability-weighted average number of answers remaining after the first feedback row.

Wordle starting word ranker: Worst branch

Wordle starting word ranker: The largest answer group that can survive one feedback pattern under the local model.

Wordle starting word ranker: Entropy and patterns

Wordle starting word ranker: Complementary descriptions of how many outcomes occur and how evenly answers occupy them.

Ordinary Wordle and Dont Wordle reward opposite branches

Wordle starting word ranker: The identical feedback partition can support two valid rankings because the games ask the player to move in opposite directions.

How to interpret the Wordle starting word ranker

How to interpret the Wordle starting word ranker
MeasureOrdinary Wordle objectiveDont Wordle objective
Wordle starting word ranker: Expected answers leftPrefer a smaller average branch so the next move starts with fewer candidates.Prefer a larger average branch so more possible answers survive.
Wordle starting word ranker: Worst remainingA smaller largest bucket protects against an unhelpful first response.A larger bucket can preserve room, although the actual feedback still controls the path.
Wordle starting word ranker: EntropyHigher information often helps split answers, but it is not a full multi-turn solution policy.Lower information can preserve uncertainty, but it does not remove future clue obligations.
Wordle starting word ranker: Core target membershipA possible immediate answer is usually welcome.A Core answer has a modeled 1 in 2,000 immediate-loss risk; a non-Core guess has zero in this pool.
Wordle starting word ranker: Practical word choiceBalance the rank with a word you know and a dictionary your game accepts.Balance survival with legality, memory, and the ability to continue under reverse clues.

Published evidence, concrete counts, and scope

Wordle starting word ranker: Alex Selby's MIT Sloan paper formulates Wordle as a finite decision problem. Its documented instance uses 2,315 candidate answers, 10,657 allowed guesses, and up to 243 encoded feedback results for five tiles. The paper reports that an exact dynamic program begins with SALET, averages 3.421 guesses, and needs at most five under that specific model. This Wordle starting word ranker uses a different 2,000-answer local pool and a one-move branch calculation, so its ranks should not be described as Selby's optimal policy.

  • Wordle starting word ranker: The current build measures 9,983 accepted guesses against exactly 2,000 local Core answers.
  • Wordle starting word ranker: Five tiles produce at most 243 encoded gray, yellow, and green combinations before impossible cases are removed.
  • Wordle starting word ranker: The readable suggestion panels contain 40 curated starter words; full-bank ranks still cover every accepted guess.
  • Wordle starting word ranker: A Core answer has an immediate modeled target risk of 1 ÷ 2,000 = 0.05%; a non-Core accepted guess has zero risk within this pool.
  • Wordle starting word ranker: The cited exact model reports 2,315 answers, 10,657 guesses, a 3.421-guess average, and a five-guess maximum for its own instance.

Wordle starting word ranker: Read the MIT Sloan working paper for its exact model, algorithm, word counts, and reported optimal policy.

Wordle starting word ranker: Explore the 3Blue1Brown lesson for a visual explanation of feedback distributions and information.

A short result from the cited exact model

Wordle starting word ranker: Selby's statement belongs to the paper's documented dictionary and full dynamic program, not to this site's local first-move rank.

“the best starting word is SALET”

Alex Selby, The Shortle Game and an Optimal Strategy for Wordle, MIT Sloan

Short quotation preserved in English. The surrounding interpretation and all local ranks are independent.

Checklist before choosing a ranked starting word

Wordle starting word ranker: A rank is useful only when the objective, word bank, and practical constraints match the game you plan to play.

  1. Wordle starting word ranker: Decide whether you want to narrow candidates quickly or preserve as many branches as possible.
  2. Wordle starting word ranker: Check whether the word belongs to the Core answer pool and whether immediate target risk matters for your rules.
  3. Wordle starting word ranker: Compare expected and worst branches instead of relying on entropy or colored tiles alone.
  4. Wordle starting word ranker: Confirm that the target game accepts the word; dictionaries and editorial lists can differ.
  5. Wordle starting word ranker: Prefer a familiar benchmark when a rare mathematical extreme would be difficult to reuse under later clues.

Primary sources and local model documentation

Wordle starting word ranker: The external sources support the optimization and information concepts. They do not endorse this independent product or its word bank.

Wordle starting word ranker FAQ

Wordle starting word ranker: Is the number-one word guaranteed to be the best Wordle opener?

No. The solve rank minimizes one-move expected survivors in this site's disclosed Core model. A full multi-turn policy, another dictionary, hard mode, or an editorial answer schedule can produce a different choice.

Wordle starting word ranker: Why is the best Dont Wordle rank sometimes a strange word?

The full-bank survival objective rewards accepted non-answer words that produce broad, low-information branches. The separate familiar-starter list is a practical filter, not a change to the underlying measurement.