The Google blog has a post about what search data say about tonight’s Academy Awards. IBM has its Oscar Senti-meter, a tool it developed with the L.A. Times and the University of Southern California Annenberg Innovation Lab.  Yahoo is predicting a winner based upon the bets people are making.

Google

Google says its predictions are based upon two factors:

• How the nominees fared in search results for the last four weeks of the prior year.
• For whatever reason, Google found that the searches people did in the state of New York were most accurate in predicting Academy Award winners.

Following that criteria, Google narrowed its choice to four picks: The Artist, Extremely Loud & Incredibly Close, Midnight in Paris and War Horse.

Google’s pick? Extremely Loud & Incredibly Close. In 2010 and 2011, underdogs won over popular picks. In each of those years the winners had the lowest search volumes. In 2010, The Hurt Locker won best film, beating out Inglorious Bastards. In 2011, The Kings Speech won best film over such favorites as The Social Network.

Google does not outright say that Extremely Loud & Incredibly Close will win.  If it were a popularity pick, The Artist and Midnight in Paris would be the favorites.

Google’s Angle: Showing off its search prowess for major events is standard for Google. Search is one of the most readily available data resources. Google is using that data to promote its Insights for Search tool which breaks down queries by category,

IBM

IBM is not making any predictions with its Oscar Senti-meter. It’s goal is to show how Twitter data reflects the sentiment  about the Oscars. The collaboration with USC and the LA Times reflects the analysis of millions of tweets for best actor, actress and picture categories. It “uses language-recognition technology, developed in collaboration with USC’s Signal Analysis and Interpretation Lab, to gauge positive, negative and neutral opinions shared in the messages. It also tracks the number of tweets.”

According to te LA Times, the Senti-Meter is designed to give insight into the vox pop surrounding Hollywood’s awards season. But it also reflects how we are using data to satisfy our own curiosity and work in ways that is more about figuring out a hunch. It’s a new discovery process that requires new skill sets in the data sciences. IBM’s Rod Smith said to me this week that the discoveries are easier to make, reflecting a shift in how we make business decisions.

“It’s more about the time to value,” Smith said. “You do not need tp plan it out.”

IBM Angle: IBM has invested billions in data analytics and has some pretty awesome tools such as Big Sheets. As a result, the company also has investments in fostering interest in the data sciences. But how do you make it sexy? The Oscars help.

Yahoo

Yahoo makes no hesitations about its choices as Google does with its predictions:

 Using the same methods to predict entertainment awards as we have done to predict political elections and sports games, we are extremely confident that The Artist will take home the Best Picture laurels; we give it an 89.7 percent likelihood of victory.

But interestingly, Yahoo does not use its own data to predict its winner. Instead, it relies on= sites such as Belfair and Intrade to make its predictions. David Rothschild, n economist at Yahoo! Research, says real-money markets are fairly accurate due to the fact that people  are actually making investments.

Social media, Rothschild argues, is too new to properly remove bias,  correlate it with outcomes, and fully understand its relative merit.

Services Angle

What does the Oscars have to with services? I think it all depends on your goals. IBM and Google are clear in their intentions.  Google is showing its search prowess and seeks to show its sophistication in using data to tell a story would not expect. IBM needs data scientists. And they want to show what companies can get out of sentiment analysis. Yahoo’s strategy is less clear. But who knows, maybe they will actually do the best job in picking a winner.