How to Solve Wordle Optimally
Wordle Solver+ is a free tool that helps you find the best guesses for Wordle using three different optimization strategies. Enter the letters you know and their included/excluded positions, and the app will suggest the best next guesses.
Optimal suggestions are based on three different algorithms that evaluate the usefulness of words in different ways. You can choose whether to prioritize average information gain, worst-case certainty, or simple average of remaining words.
Fast elimination when there are many words (>100).
Balanced choice (100–20 words).
Safe progress when few options remain (<20).
Best opening guess: Entropy-based research shows that the following words are particularly effective as first guesses: SLATE • CRATE • TRACE • STARE • RAISE. These words combine high information content, common letters (S, T, R, A, E), and efficiently divide possible answer patterns.
Statistics
Words remaining: How many words match the current filters compared to the entire word list.
Common letters: The five most common letters in remaining words (excludes already tested letters). Numbers show how many words contain that letter.
Most common letter per position: The most common letter for each of the five positions in remaining words. Helps identify position-related patterns.
Distinctiveness (bits): Information-theoretic measure of how many bits are needed to distinguish the remaining words from each other. Higher value = more uncertainty. One bit halves the number of possible words.
How the Algorithm Solves Wordle – Entropy and Strategies Explained
Entropy
Entropy measures how much a guess reduces uncertainty on average. For each word, all possible feedback patterns (green–yellow–gray) are calculated relative to the remaining words. The more evenly words are distributed across different patterns, the more information the guess reveals.
Algorithm:
where pᵢ is the probability of a feedback pattern. Results are reported in bits, which represent the amount of information – one bit means uncertainty is halved. For example, 6 bits ≈ 64 equally likely feedback patterns (2⁶).
Expected Remaining Words
This method estimates how many words will on average remain after the next guess. It doesn't measure information in bits like entropy, but directly calculates the expected number of remaining words. The goal is to choose a guess that leaves as few options as possible on average.
Algorithm:
where pᵢ is the probability of a feedback pattern and nᵢ is the number of words with that pattern. Results are reported as an avg value (e.g., avg 3.4 remaining). Lower value means better guess on average.
Minimax
Minimax doesn't look at averages but the worst possible case. It chooses a word that minimizes the maximum number of remaining options regardless of feedback. This ensures safer progress in the endgame, even if the average result is slightly worse.
Algorithm:
where ng,i is the number of remaining words if guess g receives feedback pattern i. Results are reported as a max value, which indicates the largest possible group size. Lower max means the guess is safer in all situations.
← Back