Here's an interesting observation about the FBG "top down" method of stats prediction.
One easy stat to check is total yards across the whole league for a week (passing yards for QBs, rushing yards for RBs, and receiving yards for WRs, to make it simple.) This number isn't highly dependent on any one player's performance, so it should be statistically more consistent.
If we look at rushing yards for RBs, the final Week 1 FBG prediction was 3,398 yards. The actual number was 3,064, which is pretty close. For Week 2, the FBG prediction was 3,253 yards (maybe they dropped it because they were too high the previous week) and the actual number was 3,098, even closer.
It's a totally different story with passing yards for QBs. In Week 1, the final FBG prediction was 6,715 yards, and the actual number was 6,609 yards. Again very close. In Week 2, however, while the FBG prediction was 6,744 yards, the actual number was 7,555 yards. Passing yards shot up by almost 1,000 yards (roughly 15%) in Week 2, and FBG was way off because they were predicting a number consistent with the previous week. As you might expect, we saw a plethora of 300-yard passing games by QBs in Week 2.
For WRs in Week 1, FBG predicted 4,593 yards and the actual yardage was much smaller at 4,198. For Week 2, FBG predicted 4,444 yards, but the actual total jumped to 5,089---an increase of more than 20%.
I wonder what the reason for the huge increase in passing yards was this past week? Was it just random chance, or were the coaches all thinking the same way for some reason?