3 ms·
Yeah, that measure was super hacky and I have no justification for it (but it was the simplest/only thing I could think of at the time). Do you have any suggest
by topomorph 15y ago
Yeah, that measure was super hacky and I have no justification for it (but it was the simplest/only thing I could think of at the time). Do you have any suggestions?
One thing I'm thinking of now is seeing how (# points at the xth percentile) / (# points of 10th highest rated post) changes over time. (Or maybe I should just take a closer look at how the overall distribution of points changes to get a better feel first.)
- eru 15y agoI don't know how to come up with a statistic that reflects quality of comments at the top of my head. But the statistic you chose is an interesting one. As is the distribution of points.
- ugh 15y agoThere is no way to get a handle on comment and submission quality without content analysis. You need a list of properties that define good comments and good submissions, you need some way to quantify those properties and then you need to invest obscene amounts of work coding those properties. All of this is hard work and full of uncertainty. Different people want HN to be different things, there can’t be one list of properties that does them all justice. Who gets to define what a good comment or a good submission is? Is it even possible to exhaustively translate someone’s understanding of a good comment or a good submission into a list of properties? Can those properties be quantified, in the best case with as little work as possible? Does the coding process scale to the amount of content that has to be analyzed? This is a Ph.D. thesis worth of work. I wouldn’t try to do it as a hobby.