Curve-fitting is a model, not absolute reality (like any model). Yeah, perhaps the model's predictive power is weak, but I still find it interesting. It's been posted and now we can see how well it does through the end of the year. Perhaps someone out here in TOD-land might be able to make a better model for predicting future oil prices. I think that would be interesting, too. Of course, a well-functioning model would best be kept secret, I imagine. ;o)

-best

curve fitting is not a model. it's two completely different things.