I love this hypothesis because not only does it fit in perfectly with my world view of Google, but because it’s entirely plausible to consider how their algorithm could inadvertently optimize for such a result simply by looking at click through and bounce rates.

People are likely to click on the most sensational result on the page, and in cases like this they are pretty damn sure going to pick a different search query afterward, which to the algorithm will look like success.

It seems like something easy to study and prove to a high degree of confidence if it is occurring. Particularly the theory that other search engines don’t surface the same negative content, or don’t rank it as highly.