News for hackers. Stuff that matters.
Re:Senator Al Franken's Letter to Uber [pdf] by bambax on Thursday November 20, 2014 @06:25AM (#8635225)
This story was on top of HN a few moments ago and then seemed to drop like a stone while gaining points: currently at 22nd position with 325 points and 10 hours old, whereas these stories are above it (all as old or older and with less points):

pos id points hours

6 8632209 292 15

9 8633707 129 10

12 8633683 204 14

13 8632405 139 14

17 8632363 140 15

18 8633286 58 12

20 8632043 161 16

21 8632018 141 16

What's happening?

Re:Senator Al Franken's Letter to Uber [pdf] by dang on Thursday November 20, 2014 @06:22PM (#8638849)
You can't derive story rank from displayed points and times alone. That's on purpose, to make it harder to game HN.

The phenomenon you're describing happens routinely, for multiple reasons. One is that we've been experimenting with flushing the front page of older stories periodically when there is a crop of newer ones that, if not for the older ones' inertia, would make the front page. This is related to the work we've been doing to improve the odds for good stories that would otherwise fall through the cracks. We might eventually change the main ranking algorithm to accommodate this, but our approach tends to be to do manual experiments first, software experiments second, and modify the core HN system last.

Re:Senator Al Franken's Letter to Uber [pdf] by bambax on Friday November 21, 2014 @02:44AM (#8640250)
Many thanks for answering. However, this doesn't seem to explain it

> when there is a crop of newer ones that, if not for the older ones' inertia

since the stories that were above this one were older stories, some much older (16 hours vs 10): it seems the story wasn't pushed down to make room for newer ones, but to make room for older ones.

Re:Senator Al Franken's Letter to Uber [pdf] by dang on Friday November 21, 2014 @05:14PM (#8643697)
It isn't that straightforward; I'm afraid that without going into details the explanation is bound to be less than fully satisfying.

Also, there are other reasons why stories fall suddenly. My main point is that it's routine. Sample bias comes into play here in that one mostly notices it with stories one was paying attention to.