I don’t have a link, but I remember FO discussed the Manning effect a while back. It was from the early days, where they were explaining their methodoogy in on how teams with unusually good or bad DVOA on third downs tend to regress towards their 1st & 2nd down averages the next year. The exception to this was always the Indianapolis Colts offense, whose third-down DVOA was consistently higher than expected.

Thanks Bob. Fixed.