If you know what you do, pay attention to tennis. There's quite some money in tennis predictions. I know people who don't really know a lot about data mining an

If you know what you do, pay attention to tennis. There's quite some money in tennis predictions. I know people who don't really know a lot about data mining and don't have a formal education in the field (and some of their 'math apparatus' sounds absolutely lame and laughable) who are making mid to high six figures in tennis. One guy even switched there from such an obviously attractive field as poker - a guy who already has an an excellent lifestyle from passive investments alone, and he's 32. If you can apply real data science there, and big data approach, mining a lot of data about tennis player, tournaments, surfaces, etc. etc., you can probably make millions. This isn't some temporary thing: bookies can't make better quality lines because they aren't offering odds reflecting real probabilities, they offer odds reflecting people's likelihood to bet on either side - and these are different things. And lines at betting exchanges are influenced by bookies, who make the bulk of the market, anyway.