What's the most popular number in Hacker News titles?

Internet math nerds spiraled over one tiny digit, and the comments loved every second

TLDR: A fun data dive into Hacker News titles first said 2 was the most common number, then a better count revealed **1** was the real winner. The comments turned it into a mini spectacle, with jokes about **42**, nerdy law-dropping, and triumphant chants of “We’re number 1!”

A delightfully unserious mystery turned into full comment-section theater after one curious Hacker News user asked a deeply online question: what number shows up most in story titles on the site? The first answer seemed simple — 2 was the winner — and for one shining moment, the case looked closed. But then the plot twisted. The counting method was accidentally slicing up titles like “2.0” and “1,000,” which meant the leaderboard was being haunted by what the author basically called ghost numbers. After a cleaner count, 1 actually took the crown, and suddenly this silly stats game became a miniature scandal about bad counting, software version numbers, and why small numbers rule the internet.

The comments, naturally, stole the show. One user instantly invoked Benford’s Law, the famous “spooky numbers” idea, while another jumped straight to peak nerd folklore with “Probably 42” — because no internet number conversation survives long without a Hitchhiker’s Guide joke. The biggest crowd-pleaser, though, was pure tabloid energy: “You missed a great opportunity to title this ‘The 10 most popular numbers in Hacker News titles’!” That got right to the heart of the community mood: half analysis, half dad-joke Olympics. Meanwhile, another commenter proudly yelled “We’re number 1!”, turning a database correction into a sports victory parade. In the end, the real drama wasn’t the math — it was watching a crowd of smart people get way too emotionally invested in whether 1 or 2 deserved the crown.

Key Points

  • The article uses the Hacker News dataset in ClickHouse to count numbers appearing in story titles.
  • A naive regex (`\d+`) made 2 the most common number but produced misleading results by splitting decimals, comma-formatted numbers, and embedded digit strings.
  • The initial method also inflated the count for 0 because titles containing values like 1.0 and 2.0 were split into separate digits.
  • A revised regex in ClickHouse treated decimals and comma-grouped numbers more accurately and filtered out years from 1900 to 2030.
  • With the improved method, 1 became the most common number, and the article attributes the dominance of small numbers mainly to listicles and software version numbers.

Hottest takes

"Probably 42" — fghorow
"The 10 most popular numbers in Hacker News titles" — mwigdahl
"We’re number 1!" — calvinmorrison
Made with <3 by @siedrix and @shesho from CDMX. Powered by Forge&Hive.