Hacker News Reader

Comment by formerly_proven | Hacker News Reader

On average production quality seems to correlate inversely with accuracy and actual knowledge on Youtube. It kinda makes sense, most Youtube channels seem to be one-man-shows, if the production quality is good, the dude doing it is basically an A/V person at that point and you're looking at P(A/V person AND knowledgeable in X) vs. P(knowledgeable in X).

View on HN

Not only YT, this applies to almost everything. If a website is basic black-on-white HTML with little or no styling, it usually has high-quality content.

So you’re saying us Backend guys are geniuses huh :D

I know it's a joke, but it's not a front-end/back-end thing, but a form/content thing.

Considering the channels I follow, I can't see the correlation you mention, or we have a different definition for "production quality".

Channels that get big for the sake of having amazing quality tend to pick up production quality at some point, probably by hiring people. So if you're watching older/larger channels, the correlation won't be as obvious. On the other hand, we can time warp! This is the oldest video on YouTube from Veritasium: https://www.youtube.com/watch?v=cU2dZz18P0c

I'm sure there were plenty of even gankier ones prior since by that time he already had his nice logo sorted out, a pretty good broadcasting setup, and more. In any case, here's his latest video: https://www.youtube.com/watch?v=Q1bSDnuIPbo

Personally I see three bands of content:

1. Produced-for-broadcast-TV content has the highest production costs, and the lowest technical detail. Example: Scrapheap Challenge / Junkyard Wars.

2. Midrange youtube has medium-low production costs, and medium technical detail. Example: Project Binky.

3. University lecture youtube has low production costs, and high technical detail. Example: Look up 'laplace smoothing'

Of course, it's far from an absolute rule - clearly, you can find low-technical-detail content in any production cost band :)

> Considering the channels I follow

No way for that sample to be biased.

It can be completely biased. This is why my definition is probably different than yours in the first place.

On the other hand, your sample may be also biased, so it goes both ways.

I wonder if it's more of a U-shaped curve - one-person/smaller shows follow the trend you suggest, but larger channels can do well in both areas, as they can afford a team of people rather than requiring someone to be proficient in multiple areas.

That would tend to fit with what I've seen, but that's a pretty small/biased sample.

Reply on news.ycombinator.com