YouTube: Recorded view, like and comment counts for each song. Each of these figures was turned into a percentage based on how much of the total relevant figure it made up – so for example, if Song A had 200 views and all songs combined had 1000 views, then Song A would score 20%, or 20 points, for that metric. I averaged the view, like and comment scores for each song, then boosted the points from this section by 50%, as it combines multiple data sets.
This metric is intended to indicate
overall interest – how many people are engaging with each song?
Predictions: Another combination of things: the winning
odds from an Armenian bookmaker, and the expectations of average Armenians on the street, transcribed
here. For the odds, I've turned the winning odds for each song into an implied win chance (e.g. 4/1 odds = a 25% win chance), then corrected for the bookmaker's profit line by scaling everything down to total 100%. All 13 songs under "any other", which has odds of 17/1, were assumed by necessity to have an equal chance to each other - making their derived odds (13 x 17 = ) 221/1. For the street interviews, I've just taken a percentage of total responses.
Expectations from both the odds and street interviews were then averaged out. This was because the odds are only from a minor bookmaker and haven't been updated in a while, so may not be too reliable, while the street interviews had a low sample size. Both are still interesting to note, though, and so I included them.
This metric is intended to indicate
local predictions – who do Armenians, both expert and otherwise, think has the best shot at winning their selection?
Wiwibloggs poll: Quite simply, this is a pure percentage based on results from the poll.
This metric is intended to indicate
Western Eurofan favourites – which songs are the general favourites of fans outside Armenia?
Facebook reactions: This looks at reactions on Facebook (not the comments, but the little icon thingies), counting everything apart from the angry react as a sign of support or engagement with a song. I've weighted this data with an extra 50%, since by being more Armenia-specific I imagine that it will have more overlap with the set of people who will vote in the NF.
This metric is intended to indicate
local Armenian favourites – what is attracting the most attention from people who are presumably hardcore Armenian fans?
), but with Qami it's a very different story.
