Russia: Assessing Predictions About the Coordinating Council Election · Global Voices
Kevin Rothrock

As RuNet Echo readers know well, the Coordinating Council's elections took place last week, and that body has already convened virtually through Facebook and once again in person. Weeks in advance of the vote, Global Voices offered projections based on Yandex's blogger rating index. After we published that forecast, the opposition's Elections Commission approved an additional 60 last-minute candidates. Those individuals have been added to our blogger rankings, and are now reflected in the updated data below.
In short, the blogger index successfully predicted 56.7% of the winners [ru] in the election's most important general civil category (17 out of 30 seats).
We wondered how this prediction method compares to other forecast approaches, so we also ranked the top-30-rated [ru] debate performances by candidates, aired on TV Rain before the election. That group of 30 individuals included 66.7% of the vote's ultimate winners. In other words, debate performances predicted ten percent more of the results of the election than did candidates’ blogging popularity.
Given that our Yandex blogger index was meant to test the influence of netizen activity, it's important to note that many of the winners missed by the blogger index prediction are still prominent online figures. Dmitri Bykov, Garry Kasparov, and Filipp Dziadko — just to name a few — maintain entire LiveJournal communities, manage their own web portals, and edit magazines with well trafficked websites. Because they don't author individual blogs, however, the Yandex ratings don't capture their online stature. That blogger ratings were almost as accurate a predictor as debate performance is particularly noteworthy insofar as the debates were scored exclusively by registered voters from the Coordinating Council's election. (Meaning that the same people who voted in the actual election were also responsible for ranking candidates in the debates.)
Without further ado, here is a comparison of the final election results, the predictions of the blogger-rating model, and the debate-performance model.