Recommended Free Tools
These clues and answers are for the New York Times Connections puzzle dated March 18, 2024, #281—not a puzzle from another year with the same month and day. Start with the graduated hints below; the complete answer key follows a spoiler warning.
Hints for Connections #281
Connections asks players to sort 16 words into four groups of four based on a shared relationship. Try each hint in order, and check that a proposed group works with all four words rather than just a tempting pair.
- Look for words that can describe a part or period in a process.
- Four entries are names of constellations.
- Several entries suggest spiral shapes found in nature.
- The remaining words connect to the number one in different ways.
Answers for March 18, 2024
Spoilers below: The four category names and their words are revealed here.
| Category | Words | Connection |
|---|---|---|
| Segment of a process | CYCLE, PHASE, ROUND, STAGE | Each can name a part or period in a process. |
| Constellations | CYGNUS, GEMINI, ORION, PEGASUS | Each is a constellation name. |
| Spirals in nature | CYCLONE, GALAXY, SNAIL, SUNFLOWER | Each is associated with a natural spiral form. |
| Associated with “one” | CYCLOPS, MONOLOGUE, SOLITAIRE, UNICYCLE | Each connects to one in a different way. |
Why the “one” group is less obvious
The links are not all the same kind: a cyclops is represented as having one eye, a monologue is speech by one speaker, solitaire is played alone, and a unicycle has one wheel. That variety makes this group less uniform than the other three.
#1 Best Overall
There is a label discrepancy between the two secondary answer archives: Fortnite Insider labels these four words “associated with ‘deep,’” while ConnectionsAnswers.com labels them “associated with ‘one.’” The “one” label fits the word-level explanations above. Neither archive is an official NYT answer page, so this is a discrepancy between secondary sources, not an established NYT correction.
For more on why a plausible connection can mislead, a 2024 paper by Angel Yahir Loredo Lopez, Tyler McDonald, and Ali Emami describes Connections as a task that can tempt quick associations that fail when the full set is considered. The paper reports results on its own benchmark of 358 puzzles: in its one-try setup, humans scored 39.33% and the best-performing model, Claude 3.5, scored 11%; with full hints, the reported scores were 60.67% for humans and 40.25% for Claude 3.5. These are benchmark results under the paper’s evaluation conditions, not general game win rates or results for puzzle #281. Read the paper on arXiv.
Quick Recap
Best Value
Rank #3
Rank #2
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.
