How Wordle's Algorithm Works: Inside the Answer List

There Are Two Word Lists, Not One

Most people think Wordle has one list of words. It actually has two, and understanding the difference between them changes how you think about the game entirely. The first list contains the 2,309 words that can be the daily answer. The second contains roughly 10,657 additional words that the game will accept as valid guesses but will never be the answer. When you type "FUNGUS" and it's all gray, Wordle recognizes it as a real word and tells you those letters aren't in the answer. But "FUNGUS" was never going to be the answer — it's on the guess list, not the answer list.

This distinction is the single most important thing to understand about how Wordle works under the hood. The answer list was curated for fairness and familiarity. The guess list was built for breadth and strategy. Together, they create a system where you can type almost any real English word and get useful feedback, but the daily puzzle always resolves to something reasonable. It's an elegant solution to a real design problem: how do you make a game that accepts creative guesses without ever making players feel like the answer was unfair?

Josh Wardle Hand-Picked the Answer List

Josh Wardle didn't generate the answer list algorithmically. He and his partner, Palak Shah, went through thousands of five-letter words and manually selected the ones that felt common enough to be fair. The original 2,309 words were chosen because they were words most English speakers would actually know — words you'd encounter in daily life, in books, in conversation. Not words you'd only find in a Scrabble dictionary or a medical textbook.

This curation mattered enormously. A Wordle answer needs to be solvable by a reasonable player in six guesses. If the answer were something like "XYLYL" (a real chemistry term), the game would feel broken. Wardle understood that the fun of Wordle depends on the answer being a word you've heard before, even if it takes you a few guesses to find it. I've played every daily Wordle since January 2022, and the answer-list curation is the main reason the game still feels fair after more than a thousand puzzles. The answers are occasionally tricky, but they're never cruel.

The Guess List Is Much Bigger and Much Weirder

The valid-guess list includes about 10,657 words beyond the answer list. These are words the game recognizes as legitimate English but considers too obscure, archaic, or unfamiliar to serve as daily answers. Think words like "SOARE" (a young hawk) and obscure Scrabble words that only competitive word-game players would know. This dual-list approach is clever because it lets you use uncommon words as strategic guesses to eliminate letters without worrying about whether the game will reject your input. If you want to guess "QAJAQ" to test Q and J, you can. It won't be the answer, but Wordle will still tell you whether those letters appear.

The guess list exists for a reason: it makes the game more strategically interesting. Without it, you'd be limited to only common words for your guesses, which would severely constrain your ability to test specific letter combinations. The ability to guess obscure words like "XYLYL" or "PZAZZ" gives strategic players an extra tool, even if most people never use it. The dual-list system means the game can be both accessible (answers are always common words) and strategically deep (guesses can be anything).

The Daily Word Is Sequential, Not Random

Here's something that surprised a lot of people when the code was first analyzed: Wordle doesn't pick a random word each day. The answer list is in a fixed order, and each day simply advances to the next word in the sequence. October 1, 2021 was the first word, October 2 was the second, and so on. This means that, technically, you could look up what tomorrow's Wordle answer will be. The entire sequence is deterministic.

Many people have done exactly this, and it's why you'll occasionally see spoilers posted before the game resets for some time zones. The sequential approach also explains why Wordle never repeats answers — it just keeps moving down the list. At the rate of one word per day, the original 2,309-word list would have lasted about six and a half years before cycling back. This is a feature, not a bug — knowing you'll never see the same answer twice adds to the daily novelty.

The List Isn't Alphabetical — It's Deliberately Scrambled

Wardle didn't arrange the answers alphabetically or by difficulty. The order appears to be intentionally randomized, which prevents players from predicting patterns. If the list went from A to Z, you'd know that early-year answers tend toward A words and late-year answers toward Z words. The scramble eliminates that possibility. There have been attempts to find patterns in the ordering, and some people have spotted occasional clusters of related words appearing close together. Whether this is intentional or coincidence is unclear. What is clear is that the ordering wasn't designed to build difficulty gradually or to cluster thematically.

What the New York Times Changed

When the NYT acquired Wordle in early 2022, they made several modifications to the word list. Some words were removed from the answer list and replaced with alternatives. The changes fell into a few categories: potentially offensive terms (words that could be read as slurs or had offensive secondary meanings), overly obscure words (some answers that Wardle had included were deemed too uncommon for the NYT's broader audience), and spelling variants (British and American spelling inconsistencies were addressed). The NYT also added words to the valid-guess list over time, expanding the pool of acceptable guesses beyond what Wardle originally included.

The FIBER and FAVOUR Removals

Two specific removals generated significant discussion. "FIBER" was removed because it's an American spelling of "FIBRE," and the NYT was standardizing spellings. "FAVOUR" was removed for similar reasons, being the British spelling that didn't fit with the NYT's standardization toward American English. These changes seem minor, but they highlight a real tension in word games: whose English counts? British, American, Canadian, and Australian players all have different intuitions about which spellings are "natural." The NYT's decision to standardize toward American spellings annoyed some international players, and I understand why. When you've been typing "COLOUR" your whole life, having it rejected or never appear as an answer feels like the game isn't quite for you.

Letter Frequency in the Answer List

Understanding which letters appear most often in the answer list can inform your strategy. The answer list has a different letter frequency distribution than English overall, because it's been curated to remove obscure words (which tend to use unusual letters) and include only common ones. Here's how the most frequent letters stack up:

The Controversy Around List Changes

Any change to the Wordle answer list is going to upset someone. When the NYT removed words, two camps emerged. One argued that the game should stay exactly as Wardle made it, preserving his original curation. The other argued that a game with a mainstream audience has a responsibility to remove terms that could cause harm, even if the harm is from a secondary meaning most players wouldn't know. I land somewhere in the middle. The original list was good, and unnecessary changes risk breaking something that works. But a word that has a harmful secondary meaning isn't "just a word," and removing it seems reasonable when there are thousands of alternatives. The obscurity removals bother me more — part of Wordle's charm is occasionally encountering a word you know but don't use often.

Knowing the Algorithm Doesn't Help You Play Better

People sometimes ask whether understanding the answer list gives you a strategic advantage. The honest answer is: not really, or at least not in a way that matters day to day. Yes, knowing that the answer will always be a common word eliminates the need to guess obscure terms. But most players already operate on this assumption. The bigger strategic insight — that you should focus on eliminating common letters early — is the same whether you know about the dual-list system or not.

The one place where list knowledge genuinely helps is in understanding why certain words feel impossible. If you're stuck and every remaining possibility seems like an obscure word, it's probably because you've narrowed things down to a set of valid guesses that would never actually be the answer. Knowing that the answer is always from the curated list can help you refocus on more common alternatives and break out of the "impossible puzzle" trap.

Possible Answers vs. Valid Guesses: Why It Matters for Strategy

The practical implication of the two-list system is this: you should never guess a word you've never heard of as your potential answer guess. If the only remaining possibilities are words you don't recognize, you've likely eliminated too much and need to reconsider. The answer is always a word you know. Conversely, guessing obscure words to eliminate letters is perfectly valid and sometimes optimal. If you know the answer contains R, A, and E in specific positions and you want to test for S, T, and L simultaneously, guessing a word like "LARES" (a valid guess, never an answer) can be efficient even if you'd never use it in conversation.

SLATE — one of the most efficient opening words, covering 4 of the top 8 most common answer-list letters

What the Algorithm Means for Your Daily Game

Here's my honest take: the algorithm is interesting to think about and largely irrelevant to how you should play. The best Wordle strategy hasn't changed since day one. Open with a strong vowel-heavy word, use your second guess to cover more common consonants, and narrow down from there. The algorithm matters more for understanding why Wordle works as a game than for improving your score. The curated answer list is why the game feels fair. The sequential selection is why spoilers are possible. The dual-list system is why you can type weird words without penalty. These are structural decisions that shape the experience, not tactical tips.

If you want to improve at Wordle, spend your time on strategy, not on studying the word list. The list was designed to reward good guessing, not good memorization. Play the game the way it was intended: one guess at a time, with curiosity about what the answer might be.