Looking back at World War II, armchair historians might point to Allied valor, American industrial might, or the sheer evil of the Third Reich as essential enablers of the Allied victory. But us techies know better. It was the battle between Colossus and Enigma that decided the war. In other words, how we dealt with Big Data was the deciding factor.
Cut to the 2012 election. For all the buzz about demographics, the 47%, and “legitimate rape,” the battle once again came down to dealing with Big Data. In the Romney camp: ORCA, a tool for determining the likely outcome of the election. On Obama’s side: Narwhal, a tool designed to predict the “persuadability” of specific groups of voters. And of course, we all know who ended up winning.
So, was Narwhal really a better tool than ORCA? Arguably, yes — but that doesn’t tell the whole story. As with so many technology stories, the human side of the tools was the deciding factor.
Fundamentally, the Republicans used their tool with an expectation of victory. So ORCA gave them evidence of that victory — disastrously misleading evidence, as it turns out. In contrast, the Democrats used their tool with no expectations of the result. They allowed Narwhal to inform them about the nuggets of useful information it gleaned from the raw data, enabling the Obama team to take actions that would actually lead to success, even if they wouldn’t have guessed such actions would have been efficacious without such information.
The moral of this story: Big Data can mislead, especially if you have an expectation of the result. Such errors in judgment can lead to disaster. Instead, allow your data to inform you, even if it surprises you.