Right, at https://news.ycombinator.com/item?id=19749130. That's the same study, so I think this submission counts as a dupe. It's a pity that we didn't correct the title on that one at the time.