Tuesday, December 25, 2007

Over/Under NFL Referee Rankings (after 2007 Week 16)

Here we analyze the total number of points scored (over/under) for each referee crew this season. Let's compare it to two items -- first, compare referees by the average total points scored in games they covered. Second, adjusted total points by comparing the average total points to what you would expect for the teams involved (using their average points per game season-long).

There is little change between the rankings of total points and adjusted total points at the top:
Gene Steratore (+7.0, 50.4)
Pete Morelli (+3.9, 47.7)
Scott Green (+3.6, 46.4)
Larry Nemmers (+2.7, 45.2)

The bottom is affected by using adjusted total points as opposed to total points -- Bill Leavy falls back (from third-lowest in total points to sixth-lowest in adjusted total points):
Terry McAulay (-8.3, 34.7)
Jerome Boger (-6.5, 37.5)
Walt Coleman (-2.1, 41.9)
Tony Corrente (-2.0, 41.1)
Bill Carollo (-1.3, 41.1)
Bill Leavy (-1.3, 39.8)

Bill Leavy had an unusually low expected total points this season after 16 weeks. At 41.1 total points expected, that was much lower than the next-lowest of 42.4 for Bill Carollo.

We also see an interesting correlation -- referees with large total points scored (or large adjusted total points) are in general worse for home teams! The correlation between total points per ref and home win-rate per ref is -0.535 and the correlation between adjusted total points and home win-rate is -0.540.

For example, Pete Morelli is the best for visiting teams through week 16 (home teams only win 29% of the games) and Pete also has a large total points per game (second largest at 47.7 per game). At the bottom, Terry McAulay is great for home teams (home teams win 80% of the games so far) and Terry has the lowest total points per game (lowest at 34.7 per game).

So, in general, if a referee tends to have low-scoring games, it's better for the home teams this season. What is the hypothesis there? We separately looked at how those who call more penalties tend to have home teams win more often.

Maybe the hypothesis is: if referees call penalties, there might be an unconscious bias to help out the home team such as perhaps from crowd pressure. And the way referees will affect the game with more penalties is to reduce the total number of points scored. So perhaps under this theory, less penalties means better for visiting teams and as a byproduct, more total points per game. On the other hand, more penalties means better for home teams and as a byproduct of the penalties, less total points per game. Post your comments to this speculative theory!



Post a Comment

<< Home