Computer models outperform human tipsters, but can’t guarantee riches.

On 4 July 2004, 57 minutes into the final of the European Football Championship in Lisbon, Greek striker Angelos Charisteas headed in a goal that was to win the tournament for his country.

It was one of the biggest upsets in international football. It was also a route to riches for a few lucky — or far-sighted — punters. A $100 bet on Greece at the start of the tournament, when the bookmakers gave them little chance, would have netted $15,000.

As the World Cup looms, hordes of gamblers are attempting to repeat the feat. In Britain alone, bookmakers expect to take £1 billion (US$1.86 billion) in bets. So is there a scientific way to spot this year’s Greece? Should punters back Trinidad and Tobago at up to 2000-1, or stick with Brazil, priced at a boring 11-4?

Read more here