Supposedly the committee is working pretty hard on coming up with a good way of using all metrics in the future. They have been in contact with people at several universities and with Sagarin, Pomeroy, and others about this.
One of the biggest issues seems to be finding a balance between the predictive and reflective metrics. Going 100% predictive would create the strongest field but the importance of wins and losses would be diminished. IMO, going
I imagine the goal is to be able to combine the different metrics into one number; or two numbers, one reflective and one predictive. Combining the different ratings isn’t trivial. You can’t simply average them.