How to Solve 5-Letter Wordle Words Consistently — Elimination Tree & Decision System
Solving 5-letter Wordle words consistently is not about luck or broad vocabulary. It is about executing a systematic candidate compression sequence across the 12,478-word verified dataset. Each guess is a decision node — the tile return routes you to the next action. This guide presents that decision system as a branching elimination tree, grounded in the structure of the 12,478-word verified dataset. words sorted by vowel count
This guide is for players who want a repeatable process for narrowing 5-letter words — not general tips. Every step has a defined condition and a defined response.
TL;DR
Play a Tier 1 opener (SLATE, STARE, CRANE, RAISE). Read the tile return and follow the branch. All grey → second sweep word. Green tiles → filter by position. Yellow tiles → record constraint, continue compression. By Guess 3, the surviving pool is typically small enough to resolve within three more guesses. → Wordle Solver applies all constraints simultaneously.
Best for: Wordle players who want a systematic decision process rather than general advice — a repeatable elimination tree that produces consistent solve outcomes regardless of the specific answer.
The Elimination Tree — Full Decision System
Five decision nodes. Each tile return is a branch condition that determines the next action. Follow the tree from Guess 1 through to resolution.
Wordle Elimination Tree — Full Solve Sequence
GUESS 1Tier 1 Opener
All grey
Play a second no-repeat word covering five entirely new high-frequency letters. No overlap with Guess 1. → Guess 2 node (sweep path)
Letters confirmed present, positions unresolved. Record each yellow as: "present, not in position X." Continue compression — do not guess a likely answer yet. → Guess 2 node (compression path)
3+ green
Rare. High-information return. Apply all confirmed positions + eliminations. Surviving pool is likely under 20. → Skip to Guess 3 node (resolution path)
↓
GUESS 2Compression or Adaptation
Sweep path
After all-grey Guess 1: play a second five-letter no-repeat word covering new high-frequency letters. After two sweeps, 10 letters tested. Surviving pool is defined by 10 letter absences.
Narrow path
After greens: guess within constraints. Apply confirmed positions + grey eliminations. No-repeat preference still applies at Guess 2 — avoid burning two positions on a repeated letter.
Compression path
After yellows: play a word that places each yellow letter in a new untested position. Do not guess an answer yet — resolve yellow constraints into green confirmations first.
↓
GUESS 3Pool Narrowing
Pool > 100
More compression needed. Identify the most ambiguous remaining position — the slot where multiple different letters are still possible — and test it directly. Still information-first at this stage.
Pool 20–100
Transition zone. Apply Wordle Solver to list remaining candidates. Choose a guess that splits the surviving pool most evenly — maximise the information return regardless of which answer is correct.
Identify the letter or position most ambiguous in the surviving pool. Test it — even if that word is not in the candidate pool (standard mode only). Do not guess randomly among candidates.
2–4 candidates
One differentiating feature remains. Test the letter that splits candidates (example: if NIGHT/LIGHT/MIGHT remain, test N vs L vs M in a single non-answer guess, or accept one-in-three odds).
1 candidate
Guess it. You have solved the puzzle systematically.
↓
GUESS 6Endgame
2+ remain
Apply all positional frequency data. In standard mode, choose the candidate most likely given letter frequency at the unresolved position. In Hard Mode, all confirmed letters must appear — choose the candidate that satisfies all constraints and has the highest positional frequency at the ambiguous slot.
1 remains
Guess it. System complete.
How the Pool Shrinks — Expected Candidate Counts
Following the elimination tree under typical conditions, the candidate pool shrinks at a predictable rate. The exact numbers vary by opener and answer, but the directional pattern is consistent across most solve paths.
12,478Before Guess 1
1,000–3,000After Guess 1
100–500After Guess 2
10–80After Guess 3
1–15After Guess 4
The widest range is after Guess 1 — the return can vary from an all-grey (small compression) to multiple greens (large compression). By Guess 3, the tree has typically narrowed the pool enough that the remaining candidates fit on one screen in the Wordle Solver.
Single-tile compression example: A green S confirmed in position 1 immediately cuts the full 12,478-word pool to 1,521 — an 87.8% reduction from one tile. Within the verified dataset, a grey E — E confirmed absent from the answer — removes 1,477 candidates: all words ending in E plus those containing E in other positions. These are structural properties of the dataset, not estimates. Full positional data at the positional frequency analysis.
Tier 1 Openers — The Starting Point
The elimination tree begins with a Tier 1 opener. Opener selection determines the information quality of Guess 1 — the branch conditions at the first node depend entirely on which letters are tested and where. five-letter words without E
Opener
Letters Covered
Vowel Positions
Structure
Best Use
SLATE
S, L, A, T, E
A(3), E(5)
No-repeat
Highest combined letter frequency
STARE
S, T, A, R, E
A(3), E(5)
No-repeat
Strong positional entropy
CRANE
C, R, A, N, E
A(3), E(5)
No-repeat
High elimination rate, mid-frequency consonants
RAISE
R, A, I, S, E
A(2), I(3), E(5)
No-repeat
Three vowels — high vowel coverage in one guess
All four are drawn from the 8,013-word no-repeat pool — the correct opener domain because no-repeat words return five independent data points. Full tier analysis at the opener efficiency guide.
Hard Mode — Adapting the Elimination Tree
Hard Mode adds one constraint to every node after Guess 1: all confirmed letters must appear in every subsequent guess. This does not change the pool or the answer — it changes which guesses are valid.
The elimination tree adapts in two ways under Hard Mode:
Green tiles: No adaptation needed. Green tiles confirm a position — satisfied automatically in every subsequent guess that contains the word's correct structure. Zero coverage debt.
Yellow tiles: Each yellow tile generates positioning debt — the letter must appear in a new position each guess until confirmed green. Under standard mode, you can choose to ignore yellows and continue sweeping for information. Under Hard Mode, you cannot.
Hard Mode Adaptation
Resolve yellow tiles one at a time rather than accumulating multiple yellows before resolution. After each yellow return, choose the next guess to place that letter in a new position while satisfying all other confirmed constraints. Accumulating three or more unresolved yellows by Guess 3 compresses the valid guess space significantly and risks a rhyme-family trap.
The elimination tree resolves most Wordle games within four to five guesses under standard conditions. A small number of answer types create genuine difficulty — not because the system fails, but because the answer pool structure produces late-game ambiguity.
Rhyme-Family Trap
When the answer belongs to a rhyme family — LIGHT, MIGHT, NIGHT, SIGHT, TIGHT — the final letter cluster is identical across multiple candidates. Confirming -IGHT by Guess 4 still leaves four plausible answers. In standard mode, use Guess 5 to test N, M, T in a single word rather than guessing among rhyme-family members. In Hard Mode, accept that this situation may require Guess 6.
Low-Frequency Answer
Wordle's answer set historically favours common English words, but the verified dataset includes valid low-frequency words. If the pool has compressed to under 10 candidates and all are low-frequency, prioritise positional frequency at the unresolved positions rather than word familiarity. The most common letter at the ambiguous position across the surviving pool is the correct target.
Double-Letter Answer
35.8% of the verified dataset (4,465 words) contains at least one repeated letter. The elimination tree does not filter for this — no-repeat preference applies only to your guesses, not to your expectation of the answer. When the surviving pool contains mostly repeated-letter words, work within the pool rather than applying structural filters. The double-letter analysis covers the full repeated-letter candidate class.
Elimination Tree — Decision Reference
① Guess 1 all-grey → second sweep word (5 new high-frequency letters, no overlap)
② Guess 1 greens → filter by confirmed positions immediately, continue compression
③ Guess 1 yellows → record constraint, continue compression, do not guess answer yet
④ Pool > 100 at Guess 3 → more compression needed, test most ambiguous position
⑤ Pool < 20 at Guess 3 → candidate resolution phase, use Wordle Solver to list survivors
⑥ Rhyme-family trap → test differentiating first letters in one guess rather than cycling through candidates
Frequently Asked Questions
Can you solve Wordle every time?
Under standard mode, a systematic elimination approach produces a solve within six guesses across the vast majority of games. The rare exceptions involve high-ambiguity endgames — rhyme families where five or more candidates share a pattern — where Guess 6 becomes a probability judgment rather than a certainty. Hard Mode increases this risk because guess construction is constrained by confirmed letters.
What is the best first word for Wordle?
SLATE, STARE, CRANE, and RAISE are verified Tier 1 openers — each tests high-frequency letters in high-entropy positions. All are no-repeat words. Full tier comparison at the opener efficiency guide.
What do you do if all tiles are grey?
Play a second no-repeat word covering five entirely new high-frequency letters with no overlap to Guess 1. The all-grey return is high-value information — it eliminates every candidate containing any of the five tested letters. After two all-grey guesses, ten letter absences are confirmed and the surviving pool is structurally defined. Examples after CRANE: STOMP tests S/T/O/M/P; FLUID tests F/L/U/I/D.
How do you handle Hard Mode?
Prioritise green confirmations. Each yellow tile generates positioning debt — the letter must appear in a new position every subsequent guess. Multiple unresolved yellows compound the debt and can trap you in rhyme families. Resolve yellows one at a time. Full coverage debt framework at the Hard Mode guide.
What if multiple words fit the pattern?
Identify the letter that differentiates the most candidates and test it in a single guess — even if that word is not in the surviving pool (standard mode). For example, if NIGHT/LIGHT/MIGHT/SIGHT/TIGHT remain, testing N, L, M, T in one word eliminates multiple candidates simultaneously rather than guessing one at a time.
What is the elimination tree approach?
The elimination tree treats each guess as a decision node — the tile return (green/yellow/grey combination) routes to a specific next action based on pool size and constraint type. Following the tree produces consistent outcomes because it prioritises candidate compression over early answer commitment, keeping the pool manageable regardless of the specific answer.