What you need to do is to take the co-efficient of the 120 minute scoreline, let's call this X, and provided there's no negative correlation N, invoke De Moivre's Theorem, whilst installing a vector V. Beware that there may be a variable C and that the dynamic could change with an inverse property applied. Assuming that the sum of the squares remain disproportionate to standard deviation, you should be OK.
I envisage these matches will be very tense and close which it makes them very tricky to predict, but I go for: France 1-1 Germany Brazil 2-1 Columbia
Too late for Germany/France (though obvious win for Germany). So can only offer second game... Brazil 2 - 0 Colombia Would like to see Colombia do better though
