Wordle Guide

Wordle Strategy Guide — Candidate Compression, Positional Entropy & Elimination Logic

Wordle is a constrained search problem. Six guesses to identify one word from a pool that begins at 12,478 and must reach one. Every tile return — green, yellow, or grey — is information that compresses the remaining candidate space. Strategy is the discipline of maximising that compression per guess.

This guide establishes the analytical framework used across the site's Wordle strategy articles: the concepts, terminology, and decision logic that inform opener selection, mid-game routing, and Hard Mode adaptations.

The Core Concepts

Four concepts recur throughout Wordle strategy analysis on this site. Understanding them as a system is more useful than understanding any one in isolation.

Candidate Compression

Each guess reduces the remaining candidate pool. Compression rate measures how aggressively a guess narrows the field — regardless of whether tiles come back green, yellow, or grey. High-compression guesses are valuable even when they return grey tiles, because grey tiles eliminate large candidate groups.

Positional Entropy

Different letter-position combinations carry different information value. E in position 5 eliminates more candidates than E in position 3 because E disproportionately concentrates at word endings. Efficient solving targets high-entropy positions — those where a single tile confirmation or elimination compresses the pool most aggressively.

Coverage Debt

In Hard Mode, confirmed letters must appear in every subsequent guess. Coverage debt is the positioning obligation created by yellow tiles — letters confirmed present but unplaced. Green tiles generate no debt (position satisfied). Yellow tiles generate flexible debt that must be resolved across remaining guesses.

Elimination Asymmetry

Grey tiles on high-frequency letters eliminate more candidates than grey tiles on low-frequency letters. Grey E eliminates roughly 44% of the full dataset. Grey Q eliminates fewer than 1%. Testing high-frequency letters first maximises what grey tiles remove — making the guess efficient regardless of whether the letter appears.

The key principle: A good Wordle guess is not necessarily one that confirms letters. It is one that maximises information return across all possible tile outcomes — green, yellow, and grey. Openers that test high-frequency letters in high-entropy positions create the largest candidate compression per guess, regardless of the specific answer.

The Solve Sequence — Five Decision Stages

Standard Solve Framework
Guess 1
Opener selection. Choose a no-repeat word that tests high-frequency letters in high-entropy positions. The 8,013-word no-repeat pool is the correct opener domain. SLATE, STARE, CRANE, RAISE are established Tier 1 openers — full analysis in the opener efficiency guide.
Guess 2
Compression pass. Apply all Guess 1 tile information. If Guess 1 returned mostly grey, Guess 2 tests a new set of high-frequency letters to cover more of the frequency space. If Guess 1 returned greens or yellows, Guess 2 works within those constraints to compress the surviving pool.
Guess 3
Pool narrowing. By Guess 3 the surviving candidate pool is typically under 200 words. Apply positional constraints from both previous guesses. Use the Wordle Solver to filter remaining candidates if multiple constraints apply simultaneously.
Guess 4–5
Candidate resolution. Pool is typically under 20 words. Each guess should target maximum remaining ambiguity — the letter or position most likely to split the surviving candidates into distinct groups.
Guess 6
Endgame. If candidates remain, use all available positional information to identify the most probable answer. Hard Mode adds the constraint that all confirmed letters must appear — see the coverage debt analysis for endgame Hard Mode logic.

What the Dataset Reveals About Wordle

The verified 12,478-word set gives quantified grounding to strategic decisions that are often described only qualitatively. A few key numbers inform every stage of the solve:

The pool starts at 12,478. A single green tile in position 1 on S cuts it to 1,521 immediately — an 87.8% reduction. A grey E in position 5 eliminates 1,477 candidates. These are not estimates. They are structural properties of the lexical distribution.

64.2% of the pool carries no repeated letters. The 8,013-word no-repeat subset is the correct prior for opener selection — most answers follow the dominant two-vowel, no-repeat structure. Departing from this prior requires tile evidence.

Position 3 is the highest vowel-density slot. Openers placing vowels in positions 2–4 test the highest-frequency vowel slots. This is why RAISE (R-A-I-S-E) and CRANE (C-R-A-N-E) apply efficient positional pressure — their vowels land where vowel concentration is highest.

Positional entropy in practice: A green A in position 3 cuts the search space from 12,478 to 1,206 — a 90.3% reduction from one confirmed tile. A yellow A (A present, position unknown) generates a smaller immediate compression but contributes to coverage debt resolution. Full positional frequency data in the positional analysis guide.

Hard Mode — Coverage Debt Framework

Hard Mode requires every confirmed letter to appear in every subsequent guess. This changes the strategic calculus significantly — not because solving becomes harder per se, but because yellow tiles generate positioning obligations that constrain future guesses.

Green tiles cost nothing in Hard Mode: the position is already correct and the letter is already placed. Every subsequent guess satisfies the constraint automatically.

Yellow tiles generate flexible debt. A yellow S means S must appear in a new position every guess until it turns green. If S appears in positions 1, 2, 3, 4, and 5 with roughly equal frequency, the debt resolution is genuinely flexible. Some letters have more natural resolution paths than others.

The safest Hard Mode approach: prioritise green confirmations over yellow accumulation. Multiple unresolved yellow tiles create compounding debt that limits the candidate pool to words satisfying all simultaneous placement constraints — which can trap you in rhyme families (LIGHT, MIGHT, NIGHT, SIGHT) where no single guess differentiates the remaining candidates. Full coverage debt analysis in the Hard Mode strategy guide.

Strategy Guides — Deep Analysis

Opener Strategy
Best Wordle Starting Words — Tier 1/2/3 Openers & Two-Word Coverage
Pool Analysis
No-Repeat Words — 8,013 Words, Opener Pool & Compression Logic
Hard Mode
Wordle Hard Mode — Coverage Debt, Rhyme Traps & Solve Sequences
Positional Analysis
Wordle Answer Patterns — Positional Frequency & Candidate Compression

Filter Tools — Apply Strategy to Your Current Game

Filter by Tile State
Solve Tools

We use cookies to improve your experience. Cookie policy.