Comment: clarification

(See in situ)


clarification

Thanks for all your comments and activity. I'm the creator of GOP Unbound.

As stated on the support page, this forecast is made just by analyzing & aggregating Twitter data. It's well above 120,000 and sometimes even up to 200,000 tweets a day, that build the forecast, so kind of a significant data source.

The simple assumption is that Twitter will reflect the convention floor 'somehow'. So this is an experiment and it doesn't pay attention to pledged/unpledged, bound/unbound or super delegates, nor to the fact of how many state delegates actually have been chosen.

It is: a projection on the overall bias right on a virtual convention floor now and then in August, a thing that would become relevant after first ballot. The RP campaign may have even more delegates and Santorum might have fewer, but thats not the topic of this forecast. If the 'Santorum' token has been dropped on Twitter this heavy then this is what it will look like in top unbound. I don't separate 'positive' or 'negative' tweets, so Santorum may have more sarcastic press than others, I just track the PR regardless of bias.

What might be interesting : Sundays Gallup poll showed 40% Romney, 26% Santorum, 14% Gingrich backing from Republican party members. This is very close to what my forecast says! With one major difference: Gallup sees 8% for RP, I see 19%. ...so probably I'm not as good as Gallup on Ron Paul...

Watch the delegate race via our iOS app GOP Unbound