https://gistil.com/blog What we found measuring YouTube videos: where the padding is and when the useful part starts. en Tue, 01 Sep 2026 00:00:00 +0200 https://gistil.com/blog/why-ai-summaries-are-wrong https://gistil.com/blog/why-ai-summaries-are-wrong Tue, 01 Sep 2026 00:00:00 +0200 Article A summary is a compression, and the things it loses are the things that made the video worth watching. Four failure modes, and the one job summaries are genuinely good at. https://gistil.com/blog/what-gistil-shows https://gistil.com/blog/what-gistil-shows Tue, 01 Sep 2026 00:00:00 +0200 Article A verdict, three numbers, and the receipts behind them — shown next to the video before you press play. Here is every part of the panel and what it is claiming. https://gistil.com/blog/tutorials-that-stopped-working https://gistil.com/blog/tutorials-that-stopped-working Tue, 01 Sep 2026 00:00:00 +0200 Article Six minutes into a tutorial the button it names is gone. Here is how to tell which old tutorials are safe, which are not, and why the date is the wrong thing to check first. https://gistil.com/reports/the-eight-minute-rule https://gistil.com/reports/the-eight-minute-rule Tue, 01 Sep 2026 00:00:00 +0200 Data report Runtimes pile up the moment a video passes eight minutes. The pile is real and easy to see. The penalty everyone assumes comes with it is not. https://gistil.com/reports/sponsor-segments-are-not-the-problem https://gistil.com/reports/sponsor-segments-are-not-the-problem Tue, 01 Sep 2026 00:00:00 +0200 Data report One in five measured videos carries a sponsor read, and it runs about sixty-seven seconds. The thing that actually spends your time is somewhere else entirely. https://gistil.com/blog/limits-of-our-numbers https://gistil.com/blog/limits-of-our-numbers Tue, 01 Sep 2026 00:00:00 +0200 Article Nine places where the measurements on this site stop being trustworthy, written down in one page so they do not have to be rediscovered one argument at a time. https://gistil.com/reports/how-long-does-a-tutorial-stay-true https://gistil.com/reports/how-long-does-a-tutorial-stay-true Tue, 01 Sep 2026 00:00:00 +0200 Data report Measured shelf life runs from 84.8 in science to 37.9 in tech news — a 47-point gap. Runtime does not predict it and neither does difficulty. Subject does. https://gistil.com/reports/does-longer-mean-more https://gistil.com/reports/does-longer-mean-more Tue, 01 Sep 2026 00:00:00 +0200 Data report Measured depth rises by 31 points from the shortest videos to the longest, holds inside every subject we can test, and buys less usefulness than you would expect. https://gistil.com/blog/do-you-need-an-account https://gistil.com/blog/do-you-need-an-account Tue, 01 Sep 2026 00:00:00 +0200 Article Reading a measurement we already have takes no account. Asking for a video nobody has measured yet does. Here is exactly where the line sits, and why it is there. https://gistil.com/reports/chapters-are-not-value https://gistil.com/reports/chapters-are-not-value Tue, 01 Sep 2026 00:00:00 +0200 Data report Section by section across 3,243 videos. The parts a chapter list would call content vary by 34 points inside a single video, and length barely predicts which is which. https://gistil.com/blog/watch-later-list-you-actually-finish https://gistil.com/blog/watch-later-list-you-actually-finish Wed, 26 Aug 2026 00:00:00 +0200 Article The list keeps growing because saving is free and watching is not. One rule, applied at the moment you save, fixes most of it. https://gistil.com/reports/the-shape-of-a-youtube-video https://gistil.com/reports/the-shape-of-a-youtube-video Wed, 26 Aug 2026 00:00:00 +0200 Data report Average measured value climbs fast, holds through the middle two-thirds of a video, then falls sharply near the end. Here is the curve, from 3,179 videos. https://gistil.com/blog/stop-wasting-time-on-youtube-tutorials https://gistil.com/blog/stop-wasting-time-on-youtube-tutorials Wed, 26 Aug 2026 00:00:00 +0200 Article A four-question filter to run before you press play, built from measuring how much of a typical tutorial actually delivers versus how much just restates itself. https://gistil.com/blog/spot-a-bad-tutorial-in-thirty-seconds https://gistil.com/blog/spot-a-bad-tutorial-in-thirty-seconds Wed, 26 Aug 2026 00:00:00 +0200 Article Six signals readable from the page before you commit any time, each tied to something we can actually measure across the library rather than a hunch. https://gistil.com/blog/skip-to-the-good-part-of-a-youtube-video https://gistil.com/blog/skip-to-the-good-part-of-a-youtube-video Wed, 26 Aug 2026 00:00:00 +0200 Article Five ways to find the useful part of a video without watching the setup, ranked from the one everybody already tries to the one that works before you press play. https://gistil.com/blog/search-inside-a-long-youtube-video https://gistil.com/blog/search-inside-a-long-youtube-video Wed, 26 Aug 2026 00:00:00 +0200 Article Browser search only reaches the text on the page, not what's said. Chapters, comments, and the written record of the video each cover a different gap. https://gistil.com/reports/longer-youtube-videos-are-less-padded https://gistil.com/reports/longer-youtube-videos-are-less-padded Wed, 26 Aug 2026 00:00:00 +0200 Data report Padding falls from 45% under ten minutes to 28% over an hour, and measured value rises the same way. Picking the shorter video is, on average, backwards. https://gistil.com/blog/is-2x-speed-helping-you-learn https://gistil.com/blog/is-2x-speed-helping-you-learn Wed, 26 Aug 2026 00:00:00 +0200 Article Speed compresses a video uniformly, filler and substance alike. Where doubling the speed helps, where it quietly breaks comprehension, and what it does not fix. https://gistil.com/blog/find-advanced-youtube-tutorials https://gistil.com/blog/find-advanced-youtube-tutorials Wed, 26 Aug 2026 00:00:00 +0200 Article Search tricks and channel patterns that reliably surface advanced material, and why the ordinary search results skew introductory in the first place. https://gistil.com/blog/evaluate-a-youtube-channel-before-you-subscribe https://gistil.com/blog/evaluate-a-youtube-channel-before-you-subscribe Wed, 26 Aug 2026 00:00:00 +0200 Article Three channel-level signals that predict whether a creator is worth following, separate from whether any single video of theirs was good.