France
Belgium
The tone of Twitter contributions can help predict the outcome of football matches when combined with betting market prices as a study has shown.
Twitter activities can predict if a team wins and football betting is wrong, according to University of East Anglia (UEA) researchers in the UK.
The study, which was published in the magazine Economic Inquiry, examined 13.8 million tweets during an English Premier League (EPL) - an average of 5.2 tweets per second.
These were compared with in-play winners, which are available at the same time on the online betting Betfair.
They found that when the combined tone of the tweets was positive in a second during a match-measured by a microblogging dictionary-the team more likely to win than the betting market prices implied.
Tweets were particularly informative according to the goals and red cards, suggesting that social media content is particularly useful in assessing the implications and importance of new information.
Social media are used by various companies and agencies as a forecasting tool, but the researchers wanted to find out how useful and accurate it is. They measured the overall tone of all tweets for each team in every second of 372 games played during the 2013/14 season.
"We find that Twitter activity predicts match results after controlling the market price," said Alasdair Brown from the UEAs School of Economics.
"Much of the predictive power of social media is presented shortly after important market events such as goals and red cards, where the tone of tweets can help in the interpretation of information.
"In short, social media activities are not just feelings or misinformation. If combined, they can improve the prognosis accuracy in combination with a forecast market," says Brown.
The researchers designed a series of betting strategies to determine the degree of misjudgment social media predicted.
Using conservative estimates of commission paid to Betfair and a strategy where tweets are positive in a team, a weather would have earned an average return of 2.28 percent from 903,821 bets.
This corresponds to an average return of 5.41 percent in the games under investigation.
