They are quite different indeed. It would be interesting to compare, at the end of the tournament, the likelihoods of both models on all matches. This would giv

They are quite different indeed. It would be interesting to compare, at the end of the tournament, the likelihoods of both models on all matches. This would give a pretty good idea on how good were the models.