Data – Public, Private and Unrelated

 Aggregate of American Public Opinion: 1

Random Academy Voting Members: 3, the upset of the year, and proof predictions are still just that – predictions.

Courtesy of Microsoft Bing

Courtesy of Microsoft Bing

So who is a better predictor – your fingers when everyone is looking or when no one is? People everywhere are trying to predict the Grammys and while this is a notoriously difficult task, two data giants – Microsoft Bing and Facebook are giving it a shot. While the Facebook graphics are a great geographic representation of what various states are admitting to believing, Bing is taking all those searches you aren’t sharing with friends. Assuming the data sets are big enough, do we really change our opinion when others are looking? Do these data sets matter? Let’s take a look at two categories at this year’s #GRAMMYs.


Courtesy of and Facebook

Courtesy of and Facebook

Bing Prediction: Sam Smith

Facebook Prediction: Sam Smith

Winner: Sam Smith

This one wasn’t difficult. Anticipated to sweep this category, both Bing searches and Facebook around the country predicted Sam Smith to win. Nothing too surprising with the infographic on the right – outside of low data volume state like Maine, Alaska and North Dakota, not too many were seeing Mr. Smith as the loser. Interesting to note that music recording hub Nashville, Tennessee clearly thinks highly of Brandy Clark, worth tracking in the coming months!


Courtesy of and Facebook

Courtesy of and Facebook

Bing Prediction: Beyonce

Facebook Prediction: Sam Smith

Winner: Beck

This one was much closer and then it wasn’t. With all nominations winning various states on Facebook, Bing takes their realtime predictiveness to illustrate exactly the percentages of possible wins. Clearly this could have gone to any of the 3 frontrunners, but Beck came out with only a 3% chance according to Bing to take the win. A great lesson in correlation not equaling causation, the fact that big data couldn’t predict the voting patterns of a select few is a cautionary tale to finding patterns where it may not yield predictions. Perhaps we should be looking towards Montana to predict the next election, but there I go falling right back into the trap. Also, kudos to West Virginia for still rocking out to Pharrell’s Happy in 2015.

Courtesy of Bing Predict

Courtesy of Bing Predicts

Now it will be interesting to see if various aggregating search engine analytics such as Google Flu Trends can start to predict more of the empirical data sets that we have historically been not so accurate in predicting such as the 2016 election or the upcoming hurricane season. This issue of finding correlating statistics to support claims that might not have been caused by that particular set of correlations is a battle we are continuing to fight with big data. For now I suppose we just all have to stock up on more Apple Watches and Nest Thermostats – after all, Bing says thats what 2014 us wanted in 2015. Or whatever the Grammy voters want us to buy …

This entry was posted in Big Data, Grammys, Infographics, Microsoft. Bookmark the permalink.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s