The most comprehensive Product Hunt dataset ever released | Hacker News Reader

The most comprehensive Product Hunt dataset ever released

data.world

59 pointsdimarco9 comments

Besides just a list of submitted projects and tags, is this data worth anything since the entire system the data is gathered from is skewed/biased/tainted by collusion among favored participants?

If anything it might tell you who the gate keepers are, and allow you to participate/navigate in a corrupt system, but this would just feed/grow a system that puts more power in the hands of the few.

You might then ask yourself if in the short term is the distribution afforded to you by your gaming a corrupted system worth it, and from what I've learned about building trust and strength with your audience, the answer is no.

In the long run you should pay into fair systems that act to reflect your philosophies and thus contribute to building healthy, long lasting communities that serve the good of all participants simultaneously.

Some pretty interesting ideas put out in the discussion. These stood our for me.

- Do products in collections get more upvotes on average? - What are 2016's tagline trends vs. 2015 and 2014 (e.g., are we seeing less "uber for" and more "AI for")

THis is where I saw this stuff. https://data.world/producthunt/product-hunt-research/discuss...

I'm excited to see @minimaxir (http://minimaxir.com/) comes up with based on this data.

Sorry I'm responding to this late, but since Product Hunt is rigged, it's impossible to trust any data related to it (particularly in terms of vote/comment counts), unless it's to prove said riggedness.

Yeah, I totally see where you're coming from, though consider that finding the riggedness is interesting enough:

How likely is that they would have the time / patience / skill to manipulate the data to actually be statistically legitimate? Depending on the dataset, a quick check for Benford's law would find said riggedness.

It's not that the data is illegitimate per se, but it makes any conclusions from the data unreliable. (e.g. "When is the best time to post on Product Hunt?" is difficult since the raw data alone does not account for the impact of having an influencer post the submission, and asking for upvotes through networks)

This is comprehensive. it may even be big but is there actually anything insightful that can be inferred from all this data?

How to get your post to the front page of PH?

cool beans!

Reply on news.ycombinator.com