Live data from Hacker News

Twitter May Have Predicted the Election

technologyreview.com

31–35 of 35 posts

Re: Twitter May Have Predicted the Election

#31

Any global predictor that claims to have accurately determined Trump's victory should keep in mind that he lost the popular vote. Only a state-based predictor can be plausibly said to have been able to make such a prediction.

Could you elaborate on your point? Im not sure I understand.

Re: Twitter May Have Predicted the Election

#32
post #11

Earlier quoted context omitted.

My impression is that this was a choice of her campaign (differentiate from Bill Clinton? make her seem friendlier?). Jeb Bush and Bernard Sanders also had campaigns centered around their first names.

I find it odd that you refer to "Bernard" when he's almost universally referred to as Bernie, but simultaneously call John Bush "Jeb". I'm missing whatever the intended meaning of the way you wrote these names was.

I did not know that Jeb wasn't his real name, that's all there is to it.

Re: Twitter May Have Predicted the Election

#33

Any global predictor that claims to have accurately determined Trump's victory should keep in mind that he lost the popular vote. Only a state-based predictor can be plausibly said to have been able to make such a prediction.

Could you elaborate on your point? Im not sure I understand.

An accurate global predictor with no regional information should have predicted a Hillary win (since Hillary won the popular vote by a fair margin). So if you had a global predictor predicting a Trump win, either (1) it must have taken regional information into account (in order to understand that the distribution of votes was concentrated in such a way that the electoral college would not swing Hillary's way), or (2) it must not have been an accurate predictor. Most of the polls or methods that predicted a Trump win did not take regional information into account, which means they were just inaccurate predictors.

Re: Twitter May Have Predicted the Election

#34

Earlier quoted context omitted.

Could you elaborate on your point? Im not sure I understand.

An accurate global predictor with no regional information should have predicted a Hillary win (since Hillary won the popular vote by a fair margin). So if you had a global predictor predicting a Trump win, either (1) it must have taken regional information into account (in order to understand that the distribution of votes was concentrated in such a way that the electoral college would not swing Hillary's way), or (2…

Ah, I don't disagree.

But there might be "hidden" location information in datasets that dont appear to have it, such as Twitter.

Word choice carries regional and social information.

So it's conceivable Twitter data could correctly predict the election with ML techniques that essentially are estimating the electoral college without being trained on an explicit model of it.

Re: Twitter May Have Predicted the Election

#35
post #32

Earlier quoted context omitted.

I find it odd that you refer to "Bernard" when he's almost universally referred to as Bernie, but simultaneously call John Bush "Jeb". I'm missing whatever the intended meaning of the way you wrote these names was.

I did not know that Jeb wasn't his real name, that's all there is to it.

holy cow. mind blown. did not know that wasnt his real name either.
Post reply on HN