The Isaacson-Tarbell postulate (see my first post for an explanation) offers a simple formula for predicting the winning team throughout the NFL season: (1) the team with the better record wins, (2) if they have the same record, then the home team wins. In the 2010 season the formula was correct 58.6% of the time through the regular season. In 2009 the record was 61.3%.
New this season, using the postulate, we will project the final season records and standings each week. Since the formula is wrong roughly 40% of the time, over a 17-week season, this error will compound greatly. In fact, before the season, it will simply reflect which teams play at home most in the early weeks. We shamelessly crib Tuesday Morning Quarterback's guarantee here: all predictions wrong, or your money back.
No comments:
Post a Comment