Oscar predictions using search engine results? Not very accurate!?


WordTracker, which measures search queries, was used to predict tonights Oscar winners. Here’s the story from PRWEB. It looks like this approach could go down in flames based on Alan Arkin’s Best Supporting Actor win given that he had a fraction of the online queries of others, but still won.

Best supporting Actress in a supporting role: Jennifer Hudson! Hey, it worked for her.

Here are the numbers:

Wordtracker predictions:
Actor in a leading roll
Will Smith 8751
Leonardo DiCaprio 4485
Ryan Gosling 1507
Forest Whitaker 425
Peter O’Toole 100

Actor in a support role
Eddie Murphy 2670
Mark Wahlberg 2659
Jackie Earle Haley 656
Alan Arkin 236
Djimon Hounsou 167

Actress in a leading role
Penelope Cruz 10359
Kate Winslet 9077
Helen Mirren 5470
Meryl Streep 1155
Judi Dench 573

Actress in a support role
Jennifer Hudson 6439

Cate Blanchett 1716
Rinko Kikuchi 973
Abigail Breslin 416
Adriana Barraza 65

Best picture
Little Miss Sunshine 3121
Babel 2587
The Departed 2052
Letters from Iwo Jima 1317
The Queen 1112

Well, the results are in (bolded above) and only Supporting Actress was predicted by this approach.    Interenet people … must be stupid?

Advertisements

About JoeDuck

Internet Travel Guy, Father of 2, small town Oregon life. BS Botany from UW Madison Wisconsin, MS Social Sciences from Southern Oregon. Top interests outside of my family's well being are: Internet Technology, Online Travel, Globalization, China, Table Tennis, Real Estate, The Singularity.
This entry was posted in blogs, entertainment, movies. Bookmark the permalink.

3 Responses to Oscar predictions using search engine results? Not very accurate!?

  1. Pingback: Kate Hudson Blog » Kate Hudson February 25, 2007 11:05 pm

  2. Joe, nice reading your post. I actually had a similar idea, trying to see if social media online could predict the winner. I went the other way, though, looking at online sentiment analysis for each of the Best Pictures … http://searchistheos.blogspot.com/2007/02/internet-buzz-and-oscars.html

    there is way to much noise out there to get a good understanding of online buzz without spending more time digging into the comments at a more granular level using or not computational linguistic.

    -arnaud

  3. Abhay says:

    Arnaud is right. If you don’t really evaluate the results further, it’s just kitschy and amusing for about 10-15 minutes. I used your entry as an counterexample in my recent entry about another similar site.

Leave a Reply

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

WordPress.com Logo

You are commenting using your WordPress.com 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