You want to how your team stacks up against the rest of the 128 teams in college football. Or maybe you want to know whether the College Football Playoff committee got it right.
In 2016, we have a multitude of ways to rank college football teams. At the most simple level, computer rankings take margin of victory in games and adjust for strength of schedule. (Please, please, please avoid computer rankings that do not use margin of victory.)
Here at Football Study Hall, Bill Connelly has taken ranking football teams 10 steps further by digging into the play by play of each game. The S&P+ rankings take his four factors and make adjustments for schedule.
Even with this wealth of calculations, we're missing an important source of data for ranking teams: the betting markets.
Before a game starts, the markets close on a point spread that estimates the difference in strength of two teams. For example, Navy closed as a 4.5 point favorite over Army. Since the game was played on a neutral site, this spread means the markets thought Navy 4.5 points better than Army.
The closing spreads are a remarkable predictor of college football games. According to my calculations, the team favored by the closing spread has won 75.8% of games this season (640 of 844 games).
To rank teams, I take these closing spreads and adjust for strength of schedule based on an algorithm I developed from my Ph.D. research in applied math. This process is similar to my college football team rankings, which use margin of victory in games and then adjust for schedule.
More recent games get more weight than games from September, so these market rankings capture a more recent snapshot of team strength. Michael Beuoy, who uses the least squares method to do similar market rankings, also weights games in the same manner.
Let me point out two ways in which these market rankings differ from other quantitative measures of teams.
While analytics certainly affects the betting markets in 2016, the closing spreads also consider a vast array of subjective knowledge about football. People who invest in the betting markets watch football and make the appropriate adjustments.
This inclusion of subjective knowledge is most interesting in evaluating teams like Penn State who have greatly exceeded their preseason expectation. We'll see how the Nittany Lions rank below.
Slow to react
The market rankings change slowly. The markets can't change its mind on a team after setting on a closing spread. If the estimate is wrong by 43 points, teams rankings that use margin of victory will account for this performance while the market rankings will not.
Louisville provides a drastic example of this. The Cardinals entered their game at Houston as a 17 point favorite but lost by 26.
The markets could have adjusted on Louisville the next week, but they didn't. Heck, it was only one game. The Cardinals were a 28.5 point favorite over Kentucky to close out their regular season. Louisville lost by 3.
If Louisville seems a bit high in the market rankings below, remember the market rankings don't overreact to one or two games.
Let's look at the top 25 college football teams by the markets through the Army vs Navy game.
The top 25 college football teams by the markets
Big shocker that Nick Saban's team tops these rankings. The Crimson Tide is the clear favorite to win the playoff no matter what metric one uses.
However, Alabama has scored 14 touchdowns on special teams and defense, an insane rate that can't possibly continue into the playoff against better competition.
2. Ohio State
The market rankings had Ohio State only a point better than Michigan. The markets closed with Ohio State as a 4 point favorite over Michigan, which makes sense given the usual 3 points for home field advantage.
The markets deemed Michigan one of the top four teams in the country. The playoff committee just couldn't get over that second loss at Iowa.
As discussed earlier, Louisville's rank of 4th takes advantage of the slow changing nature of these market rankings. They have yet to account for how the Cardinals struggled in losses to Houston and Kentucky.
The final market rankings will see Louisville drop, as they're a 3.5 point underdog to LSU in the Citrus Bowl.
The markets started to respect the Huskies later in the season as they marched towards a Pac-12 championship. However, this opinion has changed abruptly, as Washington is a massive 16 point underdog to Alabama in their playoff semi-final.
The 6th ranked Tigers are the lowest ranked of the four playoff teams. Two factors play a role in this seemingly disrespectful rank.
First, Clemson was a 1 point underdog to Louisville at home, which implies the markets thought Louisville the better team by 4 points. At least this is what the markets thought back on October 1st, two weeks after the Louisville thumped Florida State by 43 points.
Second, the weakness of the ACC and rival South Carolina play a role in this low rank for Clemson. These spreads for the last 5 weeks show that the markets do like Clemson.
- Syracuse, 28 points
- Pittsburgh, 21.5 points
- at Wake Forest, 23.5 points
- South Carolina, 27 points
- vs Virginia Tech in ACC title game, 11 points
However, my schedule adjustments based on these spreads did not move Clemson higher than 6th in the market rankings.
USC's season changed when Sam Darnold took over as quarterback against Utah in late September. These market rankings should capture this changing team as games later in the season get more weight than those in September.
However, a rank of 9th suggests the markets think USC is far from a top 4 playoff team.
11. West Virginia
13. Florida State
14. Washington State
15. Boise State
18. Texas A&M
19. Virginia Tech
20. Penn State
James Franklin's team improved dramatically towards the end of the season, as Penn State won the Big Ten championship. While they finished at 5th in the playoff committee rankings, the markets don't give them nearly the same respect at 20th.
24. Oklahoma State
Check out The Power Rank's bowl predictions
These market rankings are one of many predictors that go into my bowl predictions at The Power Rank. To get a free sample of these predictions, click here.