dletter [TotalFark] Quote 2010-11-04 01:55:42 PM
The Facebook political team's initial snapshot of 98 House races shows that 74% of candidates with the most Facebook fans won their contests. In the Senate, our initial snapshot of 19 races shows that 81% of candidates with the most Facebook fans won their contests.
That is still a somewhat large chunk (1 in 5) where it wasn't right. So, I guess it is a good "guess" as to who will win, but, I wouldn't say it is a slam dunk.
More interesting would be an analysis of how closely the percentages of facebook fans vs. eachother was to the actual vote.
CheekyMunky [TotalFark] Quote 2010-11-04 02:07:01 PM
dletter: That is still a somewhat large chunk (1 in 5) where it wasn't right. So, I guess it is a good "guess" as to who will win, but, I wouldn't say it is a slam dunk.
More interesting would be an analysis of how closely the percentages of facebook fans vs. eachother was to the actual vote.
This, pretty much. There were a lot of landslide victories on Tuesday, so if you just look at win/loss it shouldn't be surprising that a lot of them lined up.
If you go a little deeper and look at the correlation with winning margins, or the win/loss predictive accuracy of only the close races... I doubt it would be all that impressive.
King Something [TotalFark] Quote 2010-11-04 02:29:43 PM
Meh.
downstairs [TotalFark] Quote 2010-11-04 04:14:48 PM
dletter: The Facebook political team's initial snapshot of 98 House races shows that 74% of candidates with the most Facebook fans won their contests. In the Senate, our initial snapshot of 19 races shows that 81% of candidates with the most Facebook fans won their contests.
That is still a somewhat large chunk (1 in 5) where it wasn't right. So, I guess it is a good "guess" as to who will win, but, I wouldn't say it is a slam dunk.
More interesting would be an analysis of how closely the percentages of facebook fans vs. eachother was to the actual vote.
I think anyone with above average political knowledge could have guessed at a 74% and 81% clip respectively. Many races were a slam dunk themselves. Only a handful of races were real close, and only a handful were "big national news".
Plus, after guessing right on the slam dunks... flipping a coin gets you near 50% on the close ones.
So Facebook was 20-30 percent above flipping a coin? No big deal.
ultraholland [TotalFark] Quote 2010-11-04 04:50:28 PM
so when the fark does it find Sarah Connor?
Grandmas Candy Dish Quote 2010-11-04 04:50:42 PM
That's impossible. I clearly remember untagging myself from those gangbang photos.
Barakku [TotalFark] Quote 2010-11-04 04:52:14 PM
downstairs: Plus, after guessing right on the slam dunks... flipping a coin gets you near 50% on the close ones.
So Facebook was 20-30 percent above flipping a coin? No big deal.
A simple analysis of easily collected public data was very significantly more accurate than chance. That is big. It wasn't using election polls and media circus hype, it had no way to tell 'Oh, dude X is probably going to lose' beyond likes.
A Famous Mortimer Production Quote 2010-11-04 04:54:03 PM
I don't have a facebook. I live off the grid and take dumps in libraries.
Ulyvanhammer Quote 2010-11-04 04:54:04 PM
I dont think this means anything. It seems to me that to most people, myself included, elections are a "more of the same" hassle. I vote. I dont vote for people I vote against them. Im pretty sure people like myself make up a pretty large demographic that don't get "like-happy" with the lesser of 2 evils.
INeedAName Quote 2010-11-04 04:55:15 PM
Grandmas Candy Dish: That's impossible. I clearly remember untagging myself from those gangbang photos.
After that statement, I have to assume your username is a Euphemism, and that makes me want to vomit.
highrye Quote 2010-11-04 04:55:19 PM
A Famous Mortimer Production: I don't have a facebook. I live off the grid and take dumps in libraries.
Mom?
URAPNIS