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.