Statistics for past pre-seasons are hard to find, and do not seem to go back all that far. The following is what was available at NFL.com:
That is 14 rookie quarterbacks, over the past six years. No rookie quarterback pre-season statistics were available for 2010 or 2009. No player pre-season stats appear to be available before 2006.
Of the 14 quarterbacks we do have data for, 10 played significant portions of their rookie seasons. Those that have an N/A did not have any statistics to speak of in their rookie years. Six of the 10 rookies that played during the regular season improved their passer rating in the regular season, as compared to the pre-season. That belies the fact that there is still a slight negative correlation overall (-0.03), which means the higher the pre-season rating goes, the lower the regular season rating goes. We see evidence of this in cases like Trent Edwards (99.9 to 70.4), Jay Cutler (108.3 to 88.5), and Bruce Gradkowski (105.3 to 65.9).
One critical difference that makes it hard to compare with Wilson is that these quarterbacks were all playing against starting NFL defenses from the beginning of pre-season. Many of them struggled mightily early on. Cam Newton looked like a major project watching him last year before the real season started. Wilson got a soft landing against the 2nd and 3rd string players his first two games. That likely inflated his lofty pre-season rating (119.4), but may have also given him the confidence to do play wonderfully versus the Chiefs starters in game three.
It does not seem fair to Wilson to expect his rating would improve heading into the regular season the way Newton’s or Dalton’s did. A drop is far more realistic, but it needn’t be as tragic as Edwards or Gradkowski. One thing is clear, pre-season numbers are not great predictors of regular season numbers.