Hilary’s campaign data analytics

First, the campaign placed every congressional district by the odds that campaigning there could “ flip” a delegate into Clinton’s column. Every district has a number of delegates. These delegates are allocated proportionally. (For example, in Ohio, 12 districts had 4 delegates each while one had 17), this called for polling and modeling Clinton’s anticipated support level, gauging the persuadability of voters in a specific area and after that seeing how close Clinton was to a brink that would tip another delegate in her direction. (At the most fundamental level, for instance, districts with an even amount of delegates, say 4, are not as favorable terrain, as she and Bernie Sanders were likely depart them 2-2 unless among them attained 75 percent of the vote.)

That so-called “flippability score” was then stacked on top which media markets covered which seats. If multiple districts were affected by a media market with high “flippability” scores, it shot up the ranks. Then the algorithm took in pricing info, and what television programs it called the most “flippable” voters would be seeing, to discover what to buy.

More at politico, courtesy of MR.