Pick one human for your "champion" and let me know who his or her picks are for each week and I will keep track. Provide who his or her favorite is and the predicted margin. Make sure it's done in a way I will notice because I do not attempt to read all the notifications I get.
BTW, Massey clearly states that his system is not designed to make predictions. And I've actually had e mail exchanges with him about that. My belief is that ratings should be based on comparing teams in terms of who would be predicted to win. He does not believe that. He believes that rankings should be based on the model that best explains past results. He does provide a formula for making predictions if you wish. But that's not the objective of his system.
Eat me John. I asked you probably 7 years running and you always declined. You've worn out the invite.
I have never declined. I just described what I thought would be necessary to do it and we never got around to it. I don't mind doing it at all. But to do it I think it fair to ask you to pick a human who you think is good and see how they do against a power rating system. Might as well be Sagarin as that's easily accessible.
Actually this would be a good time to start doing it as it should be about the time Sagarin has phased out of his Bayesian approach and gone to purely what has happened so far in this year's games.
You can do it yourself if you want. I think I remember you saying it's best to pick games involving top 25 teams. So you could start now through the end of the season picking all games involving top 25 FCS teams. Predict who will win each game and by how much. Then we can keep track of what the Sagarin system predicted for each of those games.
I mean, if you don't want to that's fine too. But I certainly don't mind seeing how it would turn out.