Thank you for your response.

I'm aware of the discrepancy, I believe this is due to the way my approach uses binning. In my approach rank 1-10 are all in the same bin, favouring the chances of countries on the weak end of the bin, for example Portugal at rank 9. The opposite happens for countries at the strong of the bin, in this case Brazil with rank 1.

Smaller bins or an approach using the ratings directly will mitigate this.