If you look at the difference through the evening, their call at one point was +10 Jones, when there was nothing to indicate that. It's no big deal when there's under 10% in, but at that point we were about 30-40 percent in. Way off.
Also, it wasn't until at 90+ that they were within 2%. Their prediction kept getting peeled down, peeled down, peeled down. It was almost as was predicted that it would come all the way back. That's not a sign of a good model, that's the sign of data aggregation.
That's not a great 'predictor'. Mine was within 2% at about the 20% mark. Had Moore up a bit, but nothing over 1% until I made the call at about 1.3% advantage of Moore over the swing.