Oscar Analysis: How Did the Data Do?
Okay, so the 2014 Oscars have come and gone, and like any good analyst, I need to review the actual results against what the data was telling us leading into the event.
When viewed from a certain angle, the numbers actually mimic what played out last night's results.
If you recall, we analyzed a set of data that included facebook commentary, box office gross and Metascores leading up to the event in hopes of gaining some insights into which film might win Best Picture.
And as some of you pointed out, this exercise seemed a bit foolish because the decisions of Academy are anything but data driven. Best Actress winner Cate Blanchett pointed that out in her acceptance speech, when she called her win “random and subjective," seemingly as random as her shout out to social media telling Julia Roberts to “hashtag, suck it, you know what I mean?” We may never know what was behind that comment but what we can do is look deeper into the numbers we analyzed leading up to the big night.
Let's Look at the Numbers
Overall, the data gave Gravity a decisive edge in many of the quantitative metrics that we pulled. For example, it had the highest volume of facebook activity, box office gross, and metascore. And last night Gravity walked away with seven Oscars, the most of any of the films nominated for Best Picture. But it didn’t walk away with the coveted Best Picture Oscar; 12 Years a Slave did. In our analysis, 12 Years a Slave ranked fourth in total commentary on their Facebook page, but it ranked highest when we look behind the numbers and get into the actual text.
12 Years a Slave had the highest percentage of both positive commentary (56.2 percent) and negative commentary (15.3 percent) – an indication that people had strong opinions about the film one way or another. Sentiment analysis is a topic that, much like the Oscars, generates quite a bit of debate, but one thing is generally agreed upon - to generate a sentiment score you need feedback (textual) data. The film's impact and power compelled people to provide feedback, positive or negative, when the text was analyzed. 71.5 percent of the posts and comments about 12 Years a Slave registered with the sentiment engine. Compare that to Gravity’s 56.2% - a decisive victory with a more qualitative metric.
Harnessing the Feedback
Text analytic capabilities have advanced to a point where we can get value out of all feedback data we capture as an organization. Feedback exists across all industries, whether it be customers or partners commenting on goods or services, citizens providing feedback to a government agency, students reviewing a course or professor, patients detailing their experience at a hospital, and so on.
At Information Builders we are helping our customers realize the returns on all the feedback data they are capturing. In fact, you should check out this upcoming webinar I'm presenting on Feedback Analytics. Feedback or Textual Analytics provides us with a way to supplement all the quantitative analysis we do with qualitative analysis to help tell a more complete story, and to give us some more insight into what is going on behind the numbers.
For example, last night at the Oscars a twitter record was set for most retweeted post of all time, generating 634,000 retweets in just a half an hour and, at the time of this photo, recording a total of 2.5 million retweets. The previous record, according to twitter, was a picture of President Obama hugging his wife(779,000 retweets). However, if you look at what’s behind this tweet, there is a completely different story.
Poor Liza Minnelli, who was earlier mocked by Ellen, couldn’t even retaliate with an effective photo bomb. With a worldwide audience of nearly a billion viewers, when Ellen’s tweet was uploaded to Twitter so many people logged on to check out the epic photograph it overloaded its servers and brought the site down. Twitter even e-mailed Ellen to say she had just “broken Twitter”, which she announced at the awards ceremony. Real-time feedback in action.