Meet the Guys Who Are Trying to Blow Up Election Day in Colorado

It's something journalists and campaign insiders usually know on election day well before the public: outcomes of elections. Why, then, are the outcomes withheld?

“I think it's because one night in 1980, Congress flipped out and told the networks [releasing real-time results] was bad for democracy because there was some anecdotal evidence that folks were leaving lines believing Reagan had already won the race,” says Ken Smukler. 

Smukler is the president of InfoVoter Technologies, which provides election-day tools to nonprofits and the media. He and Sasha Issenberg, a journalist with Bloomberg, and Trevor Cornwell, an entrepreneur involved in the predictive-analytics game, founded VoteCastr, a startup that intends to release results throughout election day in swing states – like Colorado. Results will be published on Slate.

Westword: Why call a race early, and at what point will VoteCastr feel confident to put the results forward?

Ken Smukler: Technically, what we're doing is a play-by-play. So a play-by-play announcer will call a game, but they're not calling the outcome of the game, they're calling the game as they see the game, which is what we're doing. It's what the presidential campaigns will be doing, and networks will be grabbing data but not showing it to anybody. We're showing folks the game as it's being played.

We're confident that we can show the people the game from the very first moment we make it public, which will be in the morning on November 8. So we don't have any doubt in our confidence about numbers we're putting out at any time. We're actually going to poll locations on Election Day, grabbing a turnout number, and throwing that number through the logic we have built for that precinct to determine an outcome number. We'll aggregate outcome numbers up from the precinct to the county to the state.

What makes Colorado unique in this project?

Colorado is the only vote-by-mail battleground state, so we may, in fact, have the capability to actually call the outcome prior to election day. But in order for us to be confident in our numbers at the presidential level, one candidate would have to be outperforming another candidate by so much based upon our survey research and our collection-by-mail data. I don't know if that's going to happen or not. We won't know until we're well into the vote-by-mail count.

Let's go back to 1980, when Congress got involved in networks calling elections early.

Literally from that night in 1980 until November 8, 2016, nobody would have seen data on election day, which is a curious thing considering that many states – like Colorado, for example – are voting and they have all kinds of data that [the networks] are processing while people vote. What I argue is, if folks really have a problem with this kind of data while people are voting, why don't we just ban polling on the network side during the early-voting period?

What will election day look like at VoteCastr?

Polls open at 7 a.m. on the East Coast. We'll start grabbing data from eastern battleground states and processing that data. We'll be making our first call between 9 and 10 a.m. Numbers will stream in each battleground state throughout the day. You'll see in battleground states a map of what each candidate's numbers are in that state, and those numbers will change in real time based upon turnout input.

Since Colorado is the only vote-by-mail battleground state, it becomes a curiosity. You could theoretically call the race before the first pool opens on November 8. Whether we do that or not is yet to be seen, because we don't know what data we're going to be getting in during the vote-by-mail period.