An Explanation of Power Ratings
Strategies for Better Betting
Simply put, power ratings are a numerical assessment of a team’s overall talent level and ability. The best teams are ranked highest, the worst teams are ranked lowest. In my power ratings for college, I start with the top team ranked at 100 to start the season, while the lowest rated teams are often below 50. In the NFL, I use 20 as a benchmark figure for the top teams. But for the purposes of this essay, we’ll concentrate on college numbers.
Power ratings are the basis for pointspreads. The oddsmakers start with a difference between the two teams’ power ratings when they put up the opening numbers for every game each week. Then they factor in a number of other things. Homefield advantage – big difference between Boise State and Arizona, for example. Specific matchups on the two clubs – big play passing offense vs. a secondary that allows lots of big plays. Oddsmakers look at the trends — historical data as to how the teams play one another and how they tend to play games in each specific type of situation. Finally, they look at how the bettors support or don’t support the two teams. Basically, that is how the virgin lines are created each week. Again, it all starts with the power ratings.
Start your Power Ratings
The hardest part of making power ratings is simply to start doing it. I’ve seen some unbelievably complicated formulas in print, formulas that few, if any, have the time or inclination to produce. It’s much easier to do starting in August, before the first game has been played. I start with last year’s final numbers, both my own and those from reputable sources such as Phil Steele and The Gold Sheet. Then I adjust those numbers, often dramatically, to reflect offseason changes. A dramatic change in college football can be as much as 10 points.
Alabama-Birmingham lost almost all of their top notch defense from 2001, without a new group of top level recruits ready to step in. The Blazers were a team that I downgraded significantly. Meanwhile, UCLA fell apart down the stretch last year, but had a great talent base of premier athletes, a senior QB that looked ready to step up, and a fresh attitude for the new campaign. The Bruins are a good example of a team that I upgraded significantly to start the 2002 season.
The most time consuming part of the process, is examining each team’s roster and coaching changes in the offseason. Believe me when I say that this lengthy process gets me up to speed for the upcoming season – it’s an integral part of my handicapping, and gets me informed and comfortable with the personnel on each of the 117 Division 1 teams. But starting power rating numbers in August are only the beginning. The next step is adjusting those numbers on a weekly basis, to reflect how well or poorly each team is playing.
I use three basic factors to adjust the numbers each week.
First, I’ll use what I have seen. Generally speaking I watch about 15-20 college football games on any given Saturday, and incorporate what I’ve seen about the teams in my number. Specifically, I’m looking for the following. One offense, I’m trying to assess the overall play of the offensive line – do they dominate the line of scrimmage? Can the quarterback throw downfield, make good decisions and hit open receivers? Do the skill position players have big play ability? On defense, I look for similar things –did the front seven stop the run and put pressure on the quarterback? Was the secondary capable of making plays? Does the D have the size and speed to create turnovers? And of course, the special teams must be considered. Does the team have a decent kicker? Are they capable of making or giving up big plays on special teams?
Secondly, I go through every box score of every game. It doesn’t take as long as it sounds like it should, maybe three or four hours. I’m looking for how the game was played, not what the final score was. Which team controlled the clock? Who was able to run the football? Was there balance on offense between the run and the pass? How many times did each club have to punt? What was the defense able to do against the opposition? Were penalties, turnovers and special teams miscues a big factor in the game?
Third, I’ll check the injury reports following the games. When key players get hurt, the power ratings of their team must be adjusted downward. Key injuries can be anywhere on the team, not just the QB. If the big offensive playmakers are hurt, or key members of the offensive and defensive lines, it matters just as much as a quarterback injury, if not more.
Adjust Numbers For Each Teams
Once the hard work is done, I will adjust my numbers for each team. After a single strong or weak performance, I’ll adjust the numbers modestly, usually just a point or two, but after two or three consecutive similar performances, I don’t hesitate to adjust the numbers more dramatically, as high as four or five points. The amount I’ll adjust varies by time as well. In September, when the season is young, I want to adjust quickly, to be ahead of the linesmakers. The goal is to believe results, not potential – believe what I see, not what I have read prior to the season. For example, this year, Air Force showed that they were going to make some noise from the very first game, while Michigan State showed that they were likely to be a disappointing team early on as well.
In October, when conference play begins in earnest, I’m looking to adjust to how the teams are playing against competition of their own level. This is most important for the smaller schools, like those from the WAC and the MAC. It’s often difficult to gauge how good these teams really are in September because they don’t play many games against clubs of similar capability. Either they are overmatched, playing top notch competition on the road, or they get particularly weak foes at home. It’s not until October that I can get a feel for how these teams are going to play against teams of their own ilk.
November is a time to recognize teams that are peaking and those that have quit on their seasons. Smaller schools often don’t have the depth to play well in November when injuries really start to take their toll. And teams that had high expectations for the season that have not been realized often stop giving the same effort as the campaign begins to wind down. On the other hand, there are plenty of examples each year of teams that got off to poor starts, but have gotten better as the season has progresses, as well as top notch teams that really hit their stride at the most important time of the year.
Using power ratings to make smart betting decisions
Each week, after my power ratings have been adjusted to reflect the results of the previous week’s games, I’ll use them to make my own pointspreads for the next set of games for the upcoming week. It’s a simple formula – subtract Team A’s number from Team B’s number, factor in a homefield edge and voila – a pointspread is created. I’m confident in my numbers, so when I see a discrepancy between my line and the opening line, I’ll circle that game for close attention. Why is my number different from the oddsmakers number?
These are the games where I’ll be sure to check the local papers for information that I may have missed, look at the specific matchups between the two teams, and check how the two clubs tend to play games in similar situations. When I think that my number is better than the oddsmakers number for a particular game, it’s time to get a bet down.
The key to the whole process is having my own numbers to work with, making lines for every game on the board. That way, when the virgin lines come out every Sunday night, there are few big surprises – the lines, as a whole, start to make sense. When bettors learn to understand how the lines are created, and are capable of making decent lines of their own, the entire handicapping process is simplified. Making and adjusting power ratings is the only way to get that broad overview of the linesmakers process. Hopefully, this essay has shed some light on that subject, allowing the reader to use power ratings to improve their own betting results.