Could you please publish how many items got which score? I have some hypothesis I would like to test.
Thank you!
I want to see if giving an item upvotes makes it more likely that it gets more upvotes still.
I hope to be able to distinguish statistically between the influence of inherent quality and current upvotes.
My arm-chair hypothesis is that prevalence of inherent quality leads to a geometric distribution. And prevalence of attention grabbing effects should make the distribution look more like a power-law.
If that simple histogram turns out to be inconclusive extending the analysis from the final scores of items to their value over time might help.
P.S. A simple observation started me on this line of thought: I had the feeling that upvoting comments close to mine makes it more likely that mine will be upvoted. I wonder whether such strategic voting really makes a difference.
For each n, please state how many comments got exactly n points.
Is this precisely enough?
Your questions are valuable hints.
Why are you so concerned about the new page? Perhaps I should have made clearer that my focus was on the comments not so much on the articles.
Perhaps I'll just have to write my own toy-parser then.