Congressman Eric Cantor’s surprising defeat has left many pundits scratching their heads and asking themselves: How did this happen? Indeed, watching D.C. pundits lose their minds on Twitter, on multiple news sites, and on television has been quite entertaining. Nevertheless I think there’s some serious self-examination to be had about how pollsters and prognosticators got things so wrong. This upset was not without precedent; these things have happened before. However, Cantor’s internal polling had him up by as much as thirty-four points, and even the Daily Caller had him up by twelve points going into the election last night. 

They could not have been more wrong. Perhaps most shocking was how far off Cantor’s own team was. I asked a pollster friend of mine, who prefers to remain anonymous, about how they could have misfired so badly: 

Probably had more to do with [pollster] Mclaughlin being a moron. We weren’t in the field there so I can’t say for sure. Missing by five points raises an eyebrow, ten is a problem. You miss by 45 and you just don’t know what the heck you are doing…I honestly don’t know that I have ever seen a miss that big. That data is two weeks old. So it’s not like you are talking about ancient data and the environment moved out from under Cantor. That is just flat wrong. 

This raises a larger question: Do Republicans still have a data problem? After the 2008 election much was made of the Democrats’ huge data advantage. Shocks like this don’t inspire much confidence in the GOP’s electoral know-how. There are other factors at play here—open primaries that Democrats can vote in, low turnout in primaries, the seeming insurmountable lead that Cantor had. Nevertheless the GOP should think hard about their strategy come November.