i202 Fall 2009 School of Information, UC Berkeley

In the “ESP” game and similar image labeling games on the web, two people who don’t know each other assign text descriptions to images.  They get points if they agree on a label for an image without any discussion.  

• Why does Google encourage people to play this game? 
• In the game, scoring is lowest for abstract or very broad categories (“animal” gets a lower score than “bird’).  Why does this make sense given what you’ve learned in 202?
• Scoring is highest for multi-word or compound descriptions (“bald eagle” gets more points than “bird”).   Why does this make sense given what you’ve learned in 202?