/cdn.vox-cdn.com/uploads/chorus_image/image/493675/old-computers-3.1349236693.jpg)
Whether you're a gamblin' man (or woman) or not, it's typically fun to take a look at the odds of the game to get a feel for how the so-called experts see the game playing out. However, FCS odds are typically difficult to come by, so we're turning elsewhere.
If you're a college football fan, you've probably heard of Jeff Sagarin and the computer rankings that he applies to various sports. Most notably, his computer rankings are used in determining the participants in the NCAA basketball tournament and in the BCS rankings.
If you're not familiar with the Sagarin ratings, I would suggest reading the ratings overview at the Sagarin website (the team-by-team ratings are also in there) but here's a quick overview for you if you're not into all that fancy schmancy talk.
For the first few weeks of the season, the ratings are weighted for each team (with the previous season serving as a basis). Once all the teams are what Sagarin refers to as "well-connected" then the starting weighted rankings are no longer used.
Each team is given three different ratings:
- Elo-Chess - only winning and losing matter, score margin does not matter.
- Pure Points/Predictor - only the score margin matters.
- Rating - a combination of Elo-Chess & Pure Points/Predictor.
While the Elo-Chess is used by the BCS, Sagarin actually believes the the Pure Points rating is the best indicator of future games.
Now on to making predictions. When predicting an upcoming game, one simply compares the ratings of the teams in question and assign the home team an additional amount of points according to the ratings you are using.
Here's a look at the current ratings for the CAA conference (HA = Home Advantage):
NCAA RANK | TEAM | RATING | ELO-CHESS | PREDICTOR |
---|---|---|---|---|
HA = 2.53 | HA = 2.48 | HA = 2.69 | ||
101 | James Madison | 62.93 | 61.69 | 62.94 |
106 | Old Dominion | 61.59 | 63.89 | 60.94 |
112 | Delaware | 61.07 | 55.76 | 61.86 |
113 | New Hampshire | 60.74 | 59.35 | 60.77 |
120 | Villanova | 59.46 | 59.13 | 59.30 |
139 | Richmond | 56.03 | 50.89 | 56.75 |
140 | Towson | 55.98 | 54.07 | 56.11 |
149 | William & Mary | 53.44 | 42.71 | 54.97 |
165 | Maine | 50.46 | 45.27 | 51.14 |
217 | Rhode Island | 35.99 | 41.11 | 34.55 |
227 | Georgia State | 32.27 | 31.74 | 32.14 |
As you can see, Villanova would be an approximate 5-point favorite when using the Pure Points/Predictor rating, and about a 6-point favorite when using the blended ratings.
The counter-argument to these ratings is the "weighted" concept. Villanova's weighted rating is most likely very low based on the 2011 season. Richmond's is likely a little bit underrated as well, but not as much as Villanova. The Sagarin ratings are still weighting last season a bit more heavily than the ideal.
What do you think of the Sagarin ratings and their validity in predicting the Villanova-Richmond game?
Loading comments...