Since it's now fixed, when anyone votes on an old article it will be re-ranked to its correct position. Since they've not been re-ranked for hours, most of the moves will be pretty drastic.
Update: We're now discovering that a side effect of this was the automatic killing of a number of submissions that should not have been killed. We're going through logs and addressing those.
Update: Improperly killed submissions have been restored.
Final Update: Turns out it was actually an issue with voting. We were correctly recording all votes (so you saw the arrows disappear), but not correctly counting them. If you're wondering how that's possible, you can thank denormalization.
We've now identified the bug and deployed the updated and fixed code.
..... and this was just killed. I'm going to email dang about it.
Thank you very much for doing so. We often read the emails much sooner than we have time to reply to them, especially if there's an urgent action to be taken, so please don't assume that your message has had no effect if you don't hear back right away.
The title was "The Tech Industry vs. San Francisco " : http://www.newyorker.com/reporting/2014/07/07/140707fa_fact_...
The /new list seems unusually inactive..even at 4AM, even links less than 15 minutes old can accrue a few upvotes. I just checked the /new list and nothing has more than the initial upvote (except for the Source Sans link, which I just upvoted to test)...
It could be algorithmic. It could be a group of users whose first action on the site is to flag weak submissions. It could be an individual with moderation powers. It could be each of them at different times.
Over the years it's been particularly noticeable on Ask, but not uncommon on the front page. But more recently it has been more noticeable on the front page. Personally, I'm all for it. Algorithms can only go so far before they are subject to gaming.
If it is algorithmic, perhaps it is triggered based on trends with new submissions - e.g. such as when few are gaining meaningful adjusting the gravity curve to favor older more popular submissions. This would maintain the quality of the front page and perhaps encourage more effort toward finding quality submissions or bring the focus to discussions.