YouTube: MX vs ATV Legends Clips

Status: dormant. Data never pulled.

What this was

A stretch of posting gameplay clips — jumps, tricks, the stuff that looked good. Some of them got real work: editing, timing, cut points, an actual arc. Some of them were close to raw capture.

The thing I came away with is that the effort didn't show up in the numbers. Not weakly — as far as I could tell, not at all. The ones I was proud of and the ones I threw up did roughly whatever they were going to do, and I couldn't predict which was which. I have the same feeling about my Google Maps photos, which is why that one is also written up here.

That's an impression, not a finding. I've never actually looked.

The question

Does production effort correlate with performance at all, and if not, what does?

The honest version is broader: if effort doesn't predict outcome on YouTube, and it doesn't seem to on Maps either, then "why did this one do well" may be the wrong question across the board. What would be left is whether a class of content raises the median, which is a much less satisfying question and probably the real one.

Getting the data

Public per-video stats don't need auth or an API key:

yt-dlp --flat-playlist --dump-json "<channel or playlist URL>" \
  | jq -r '[.id, .title, .view_count, .upload_date] | @tsv'

That gets views, titles, and upload dates. It does not get watch time, click-through rate, impressions, or traffic sources — those are Studio-only and need a CSV export from YouTube Analytics.

The part no tool can supply is the effort rating. That has to be me going down the list and scoring each clip on how much work it took, ideally before looking at its view count.

To fill in

• Channel or playlist URL
• Rough date range of the posting run
• Why it stopped
• Effort score per clip, blind to the numbers if possible