Back in 2002 I was wrestling with financial modelling and as an experiment and learning aid, I built myself a World Cup model that endeavoured to 'predict' the winner. I didn't do very well. In 2006 I built an improved model and successfully forecast 66% of the winners and losers. I've decided to have a go again.
This is only a bit of fun ... most of the scores won't happen as forecasted ... but the model 'learns' about the various teams' capabilities as the competition progresses and it's interesting (if you like that sort of thing) to see the forecasted results changing as the competition progresses.
If anyone is interested, I've built all of this into some web pages and it can be found here ..... World Cup Forecast
As I've stated on the web site, DO NOT rush out and place bets on the basis on this. I know absolutely nothing about football (other than kicking the ball into the other teams net counts as a goal and kicking it into your own net makes your team mates shout angrily). This is just a bit of harmless fun and if the actual results bare any resemblance to the predicted ones, it will be more by luck than judgement.
Enjoy the footie!!!!