The typical launch at 181 points, drawn from the 2.6% that cleared 100

Be the first to star this exhibit

A launch newsletter we invented charts “what a Show HN launch looks like” and finds the median 2025 launch got 181 points. Its data is every Show HN post that cleared 100 points: 738 of the 28,276 posted that year. Counted whole, the median post got 2 points, and 35.5% never got a vote beyond the poster’s own.

Hacker News Search API by Algolia: every story tagged show_hn created 1 January–31 December 2025 (UTC), 28,276 posts, with points and comment counts as indexed on 19 October 2026; counts are posts, not distinct projects. The newsletter, its advice and both drawings are ours; the counts, shares and medians are our tallies of the API’s records., 2026Added October 19, 2026

01The claim

A Show HN launch typically lands about 181 points and a lively comment thread, so the thing to plan for on launch day is the traffic.

02The trick

Every bar in the newsletter is a true count of real posts, and the median of those posts really is 181 points. The catch is what a post had to do to get into the chart: clear 100 points. It is the kind of cut a “best launches of the year” list makes, and it keeps 738 of the 28,276 Show HN posts submitted in 2025, 2.6%. Everything the newsletter says about “a launch” is a fact about those survivors, and the chart can’t reveal that, because the 27,538 posts that stayed at 100 or below aren’t on the axis to be compared against. The checklist makes the same move in miniature. “42% of launches linked to GitHub” is 312 of the 738 survivors; the share of winners with a trait says nothing about the odds for everyone who had it, since that depends on how many GitHub posts failed, and those are exactly the rows that were dropped. The fine print does state the filter. And the third checklist item, “every single launch in our data got past 100 points”, is true by construction. (Ship & Tell Weekly and its advice are our invention; every count is from the HN Search API.)

03The fix

The redraw puts all 28,276 posts on one zero-based count axis, in two single-point bars and then bands that double in width, so the long tail fits without hiding the head. Hacker News starts every post at 1 point, the poster’s own vote, and the tallest bar is the 10,038 posts, 35.5%, that stayed there. The median post got 2 points and no comments. The newsletter’s whole chart becomes the thin orange sliver on the right. The GitHub advice survives the full count in a smaller form: 312 of 7,969 GitHub-linked posts cleared 100 points, 3.9%, against 426 of 20,307 for everything else, 2.1%. So GitHub-linked posts cleared 100 nearly twice as often (which may say more about the kind of project that links its repo than about the link itself), and 96% of them still didn’t clear it. Before trusting a picture of “what success looks like”, ask what each row had to survive to get in, and say so in the title, not the fine print. Then count the ones that didn’t make it, even when they take up most of the chart.