A new study suggests that analyzing Twitter can help gamblers beat the spread for NFL games.
The study was conducted by computer scientists at Carnegie Mellon University and their full findings will be presented on September 27 at the Machine Learning and Data Mining for Sports Analytics conference in Prague, Czech Republic.
The group started their work as a class project and it blossomed into something bigger. Former student Kevin Gimpel, now a research assistant professor at Toyota Technological University at Chicago, eventually developed class project into a study of three NFL seasons (2010-2012). His research team then used automated tools to sort through Tweets at an average of 42 million messages per day.
Researchers combed through the tweets to choose hashtags associated with NFL teams such as #NewYorkGiants, #ChicagoBears, #SteelersNation, etc.
The group focused on Tweets sent at least 12 hours after the start of the teams previous game and one hour before the start of the new game.
After measuring fan and other engagement sentiment the group was able to determine with decent accuracy who would beat the spread based on volume of tweets and the distribution of positive and negative words in tweets.
What the study doesn’t determine is the winner of the game or the over/under score. The researchers at this point have been 55 percent accurate with NFL spreads and that could equal a huge advantage for gamblers.
The social scientists understands that sports books don’t just look at a team’s performance, home field advantage, and weather predications so there goal is to show how the psychological factors may play into the equation.
Twitter analysis is already working to help stock traders determine buy and sell plans for the NASDAQ, NYSE, and other exchanges, so why wouldn’t to also work to help hedge a bet during the NFL season.
Do you think Twitter and other social networks can really beat the odds on the spread?