Every time the algorithm changes, the ranking of various sites is going to change. There’s ALWAYS going to be some conservative sites (and some liberal sites, and some My Little Pony fansites) that do worse after the change. That doesn’t tell you anything in itself.
https://twitter.com/johnstossel/status/1308548315315929088
This is the same mistake PragerU made in their failed YouTube lawsuit. They disagreed with some of their videos being flagged as restricted for minors with parental controls turned on, and cried “aha, bias!” without checking whether they got it any worse than other channels.
That said... there is no Ideal Platonic Objective Ranking. It’s not like Breitbart is entitled to the One True and Correct search visibility they had at some prior date, and all subsequent deviations from that baseline show bias.
Suppose (and I doubt this is the case, but suppose) Google really did tweak the algorithm in part because they thought it was giving too much prominence to sites like Breitbart and the Caller relative to other news sources, with the aim of bumping them down.
Does that show the old ranking was fair and the new one is biased? Or does it show the old ranking was biased in favor of those sites and the new one is fair? There’s just no way to answer that independently of SOME value judgment about the sites & what “relevance” should mean.
Should the most popular sites automatically rank highest? But the algorithm itself significantly affects what’s popular. If that’s the rule, you’re committed to a feedback loop that locks in the results of your 1.0 algorithm (and whoever got good at gaming that algorithm).
(The most dramatic shifts in ranking often occur precisely because some sites had mastered a particular set of Search Engine Optimization tricks that suddenly stop working. But it’s not inherently unfair for your SEO tricks to stop working.)
• • •
Missing some Tweet in this thread? You can try to
force a refresh
