I think I’m done using RealTimeRPI as my RPI source. They have not corrected the ASU result, and some other stuff looks off to me too. Looks like ESPN has basically done the same thing, replicating the RPI formula; I’ll be using them as my RPI source from now on:
