Live data from Hacker News

Show HN: Reversing YouTube’s “Most Replayed” Graph

21–23 of 23 posts

Re: Show HN: Reversing YouTube’s “Most Replayed” Graph

#21

Post by prvt » Fri, Jan 16, 2026, 10:22 AM UTC

levysoft wrote: post #20

Congratulations: from a simple intuition about something that many wouldn’t have noticed, you wrote a beautiful article.

Means a lot, thanks! I am a huge fan of the phrase "the devil is in the details."

Re: Show HN: Reversing YouTube’s “Most Replayed” Graph

#22

Post by Animats » Fri, Jan 16, 2026, 6:36 PM UTC

prvt wrote: post #17

Animats wrote: post #16

So it's a histogram of 100 bars, forced into a spline in SVG to make it look like it has more detail than it does.

That's the big picture idea.

I knew that polynomial curve fitting results in overshoots like that, but didn't think that splines had that problem.

Re: Show HN: Reversing YouTube’s “Most Replayed” Graph

#23

Post by prvt » Sat, Jan 17, 2026, 10:15 AM UTC

Animats wrote: post #22

prvt wrote: post #17

Earlier quoted context omitted.

That's the big picture idea.

I knew that polynomial curve fitting results in overshoots like that, but didn't think that splines had that problem.

Splines are polynomials after all. Also, not every spline has this problem. The one YouTube is using is Cardinal spline.

Post reply on HN