Long read · The Daily Five Letters

The five-letter habit that stuck.

Four years on, Wordle is still the polite morning ritual that gets the brain warm. Here’s how to play it well, what the openers really mean, and why the streak isn’t the point.

S L A T E

Wordle landed in late 2021 as a curiosity and stayed as a habit. Four years and one New York Times acquisition later, it’s still the single most-played puzzle game in English — not because the gameplay is profound, but because the format respects your morning. Six guesses, one answer a day, a share button that gives away your performance but never the word. That’s the entire design. It hasn’t needed an update.

This review is the long version of what I tell friends when they ask whether they should bother. The short version is yes, with caveats. Wordle teaches three useful skills — pattern recognition, letter economy, and disciplined elimination — and it teaches them inside a four-minute window. Almost no other puzzle does that, and most of the games that try are worse at it.

i What Wordle is, and isn’t

Wordle is a five-letter word puzzle with six guesses. Each guess returns colour feedback: green if the letter is right and in the right place, yellow if the letter is in the answer but elsewhere, grey if the letter isn’t in the answer at all. There’s one new puzzle per day, the same puzzle for everyone, and no timer. That’s the whole specification.

It is not, despite what you’ll read on LinkedIn, a vocabulary game. The NYT solution list is around 2,300 common five-letter English words, all of which any literate adult already knows. The real challenge is not knowing the words but choosing which one to try when you only have six attempts and you already know three letters. Wordle is a logic game wearing a vocabulary game’s clothes, and once you see it that way, the strategy stops being about being clever and starts being about being efficient.

ii The opening word, with numbers

The opening word is the most-debated and least-important strategic choice in Wordle. Most reasonable openers cluster within a tenth of a guess of each other; the difference between SLATE and CRANE over a year is roughly two extra solves at three guesses. Use whichever one you can remember at seven in the morning.

What does matter is using an opener that covers high-frequency letters in their high-frequency positions. Wordle’s answer list is biased: S, E, A, R, O, T, L, I, N show up disproportionately often, and they show up in specific slots. SLATE tests S in slot one (high frequency), L in two, A in three (very high), T in four, E in five (extremely high). That’s why it benchmarks well, and why a word like HELLO — which technically uses high-frequency letters — benchmarks badly. The two Ls duplicate information, and H in the first slot is uncommon.

What does not matter is testing five distinct vowels. ADIEU is famously bad — it averages 4.08 guesses in our simulation because by the time you know four vowels, you still don’t know any consonants, and consonants are what disambiguate English words.

The opener gets the headlines. The second guess wins the game.

iii The second guess matters more

The opener gets the headlines; the second guess wins the game. Once you have your first board state, the question is whether to commit to your greens and yellows or to use the second guess as another probe. The right answer depends on how many letters you confirmed on the first turn, and the rule is more conservative than most players think.

  • Zero to one greens or yellows: probe again. Use a word that tests five entirely new letters. SLATE followed by POINT covers ten of the twelve most common letters in two guesses, which is information no amount of clever third-guessing can recover from a bad start.
  • Two to three greens or yellows: mixed. If your yellows can plausibly rearrange into a likely answer, commit. If not, probe one more time.
  • Four or more confirmed: commit. Brute-forcing the last unknown slot from a list of three or four candidates is faster than burning a guess on more information.

iv Hard Mode, finally explained

Hard Mode forces every subsequent guess to use the greens and yellows you’ve already revealed. It sounds harder; in practice it’s a coaching mode. By removing the option to use a guess as a pure probe, it pushes you toward better openers and tighter mid-game logic. The downside is that you’ll occasionally fail a puzzle you could have solved by burning a guess on information — because the answer is one of those four-candidate clusters (BATCH, CATCH, HATCH, LATCH, MATCH, PATCH, WATCH) that Hard Mode can’t disambiguate within six guesses.

Recommended path: play standard until your three-month average is under 4.0 guesses, then switch on Hard Mode and accept a temporary regression. After a month, you’ll be better. After three months, you’ll wonder how you ever played without it.

v Common mistakes

The losing patterns are predictable. After a year of editorial coaching, I keep seeing the same four.

  • Using two openers. If you switch between SLATE and CRANE based on mood, you’re not learning either. Pick one and pair it with three or four memorised second-guess templates.
  • Forgetting greys are information. A grey letter rules out five slots, not just one. After three guesses, look at your greys before your greens.
  • Skipping double letters. Roughly 9% of NYT solutions have a repeated letter. If five distinct letters aren’t fitting, try a double of your most-confirmed consonant.
  • Tunnel vision on yellows. A yellow letter is in the answer but not in that slot; people forget the second clause. If E is yellow in slot two, it is not in slot two in any subsequent guess. Mark it on paper if you have to.

vi If you like Wordle, also try…

Wordle’s closest sibling is the daily Lexigo, which trades Wordle’s grid for a connected-letters path puzzle but keeps the daily, one-attempt structure. I write the column for both; when I’m stuck on a Lexigo, I check lexigoanswers.com, which publishes the day’s solution. For the more competitive end, Quordle (four boards) and Octordle (eight) take the same mechanic and stress-test your tracking; they’re reviewed separately in the field guide.

For the broader question of what to play when, and how to fit a puzzle routine into a working day, our long-form field guide is the place to start. It’s edited by Margaux Lindgren and updated every quarter.

P

Priya Venkat

Priya is GameTipster’s mobile puzzle columnist. She came to puzzles from data science and built the Monte Carlo simulator that produces the lab’s opening-word benchmarks. Her Wordle 12-month average is 3.42 guesses, which she will mention only if you ask.