alt Hacker News> It seems like they are saying "posts that were posted recently have less views than posts posted before the Yandex strike" which is explained just by time.
It doesn't seem you even bothered to read the article. It shows quite clearly that following the strike post from populist far-right figures had literally a residual level of engagement, which includes both views, comments, reposts, likes, etc.
The screenshots clearly show posts preceding the strike showing literally an order of magnitude more comments, reposts, etc.
I'm not personally tracking there metrics, but I'd expect engagement to normalize a few days after the strike as Russia's propaganda arm is expected to scramble to get it's operation back online. Therefore, "just by time" can accommodate for Russia getting their botfarms back online. But the snapshot covered by this article alone are clear and obvious.
I read the article and that's why i wrote my comment.
For Tice they show a post with 10k views from october 12 and post with 20k views from october 8. There is also post from october 9 which obviously went viral and has hundreds of thousands views, but that's not itself proof because there's a always variability from the algorithm.
Reshare counts are more interesting, 700 for each of oct 8-9 posts but 50 for each of oct 12 posts, that is less likely if we are talking about just the network effects.
anyway we all know russian bots exist. They basically got trump elected after dorsey sold to musk and musk let them run amok. But this is not hard statistics. if they fixed a time period (like, in 3 days posts used to get N engagement avg, now in 3 days posts get M engagement avg) then it could be actual statistics. and they really need to stress reshare counts.