SIGNALTUBE · ORIGINAL DATA ANALYSIS

What 6,386 Hacker News video records reveal

This page reports calculations SignalTube makes from its catalog snapshot: which HN comments were recommendations, how often recommendations returned in different months, and how videos are classified by format, level, and shelf life. The figures below describe the snapshot.

Catalog videos6,386Indexed video records
Labeled HN comments39,397Comment records linked to catalog videos
Repeat recommendations54.9%3,505 videos recommended in 2+ distinct months
Long-running picks9.0%575 videos recommended in 6+ distinct months
Mean recommenders per video2.314,887 video-recommender links, summed per video
Mean recommendation months2.3Distinct months with recommendations per catalog video

How HN comments use these videos

Across 39,397 stance labels, 3,996 are strong recommendations and 12,020 are recommendations (40.7% combined). The remaining labels mark contextual references, critical mentions, or jokes. A comment that links a video is not automatically an endorsement.

Strong recommendations
3,996 · 10.1%
Recommendations
12,020 · 30.5%
Contextual references
18,927 · 48.0%
Critical mentions
1,761 · 4.5%
Jokes
2,693 · 6.8%

Which topics keep getting recommended?

Among topics with at least 20 catalog videos, Software architecture & design has the highest share of videos recommended in six or more distinct months: 46 of 171 (26.9%). The cutoff avoids ranking tiny topics on a handful of records.

  1. Software architecture & design46 / 171 videos · 26.9% persisted 6+ monthsMean recommendation span: 4.2 months
  2. Mental health & neurodiversity12 / 45 videos · 26.7% persisted 6+ monthsMean recommendation span: 3.9 months
  3. University lectures & courses31 / 125 videos · 24.8% persisted 6+ monthsMean recommendation span: 4.3 months
  4. Productivity & self-improvement22 / 93 videos · 23.7% persisted 6+ monthsMean recommendation span: 4.0 months
  5. Computability & complexity theory8 / 34 videos · 23.5% persisted 6+ monthsMean recommendation span: 3.8 months
  6. Python, Ruby & scripting languages26 / 115 videos · 22.6% persisted 6+ monthsMean recommendation span: 4.1 months
  7. Politics & history (other)7 / 31 videos · 22.6% persisted 6+ monthsMean recommendation span: 3.5 months
  8. Testing, debugging & developer tools33 / 150 videos · 22.0% persisted 6+ monthsMean recommendation span: 3.5 months

Formats in the archive

  • Talk1,423 · 22.3%
  • Explainer1,090 · 17.1%
  • Demo811 · 12.7%
  • Comedy565 · 8.8%
  • Commentary455 · 7.1%
  • Documentary401 · 6.3%
  • Interview338 · 5.3%
  • Lecture312 · 4.9%
  • Other279 · 4.4%
  • News236 · 3.7%
  • Music186 · 2.9%
  • Tutorial173 · 2.7%
  • Podcast93 · 1.5%
  • Course24 · 0.4%

Catalog classifications

Learning level

  • General4,552 · 71.3%
  • Intermediate1,146 · 17.9%
  • Intro391 · 6.1%
  • Advanced297 · 4.7%

Shelf life

  • Evergreen4,511 · 70.6%
  • Dated1,476 · 23.1%
  • News399 · 6.2%

How to read these numbers

“Distinct recommenders” counts people separately for each video. The archive total is a sum of video-level counts, not a count of unique people across all videos. “Recommendation months” counts distinct months in which a video received a recommendation; it does not measure views or watch time.

HackerTube’s export includes 39,834 source comments and caps the labeled comment list at 20 per video. This site indexes 6,386 of the export’s 6,454 video records. Verified YouTube runtime and upload-date coverage is currently 0 and 0 videos; unknown values are left blank rather than inferred from HN years.

Browse the topic archive, compare top HN picks, or read the dataset field guide. Each video page adds a same-topic, same-format peer comparison. Explore the archive with six-month recommendation persistence, evergreen shelf life, and introductory videos.