originally posted by .sasha:
correct me please, but you have an underlying assumption that all games have a loss vs draw weights that are linearly predetermined by chances of victory, no?
You could argue , for example , that Chelsky has a low chance of beating Everton but also a low chance of losing.
You could make it as complex as you want, but the precision will be illusory.
Back of the envelope is about as predictive as something that I could spend 30 hours on in this case.