The Pi-Rate Ratings

June 10, 2018

The PiRates Are Hard At Work For Football Season

Hello Mates.

This is the head PiRate telling you that we have set sail and are on our way out to the green sea, otherwise known as the Football Field.

We have already begun to update our college football ratings for the 2018 season.  We have a unique way of updating our ratings from the end of the previous season to the beginning of the new season.  We have point values assigned to every starting and backup position on offense and defense plus an overall rating for the components of special teams.

For instance, as it should be obvious, the Quarterback is the most important position in college football.  So, we take special care in determining how many points better or worse each of the 130 teams is in week one compared to their final game of the year before by adding or subtracting points based on the QB.  We look at certain analytics data, just like baseball teams look at advanced statistics.  We actually have our own version of WAR for college football.  Instead of Batting, Fielding, and Pitching WAR, we have offense, defense, and special teams WAR, only instead of putting the final numbers into wins above average, we use points per game above average.  In the PiRate Ratings, a rating of 100.0 is average.  If a quarterback is worth 6.3 points better than an average QB, then before we look at the reserves, a team that is totally average at every other position would begin the season at 106.3 for this particular piece of the ratings’ puzzle.  If a quarterback has been determined to be 3.8 points below average, it would reduce the average team from 100.0 to 96.2.

On offense, the receiving corps is the next most important part of the equation, while the offensive line and running back corps are equal in importance as the last two parts.

Defensively, we look at all three units as they play against the run and the pass.  Stopping the pass is more important these days than stopping the run.  Interceptions, Passes Broken Up, Sacks, and Hurries are all parts of the Pass defense stats we look at.  We also have special ratings for nose guards that have the responsibility of taking up space and keeping blockers away from the linebackers, as these heroes of the trenches rarely show up in box scores, until one of the behemoths gets injured, and his replacement prevents the star will linebacker from making his usual 14 tackles with 2 for loss.

We have an intangible part of this equation as well.  Say a team has a change in coaches.  The old coach was a spread option or flexbone disciple, while the new coach is an air raid disciple.  It’s going to take 2-4 years to fully implement the changes, as that five foot nine inch triple option QB is not going to work as a dropback passer.

Another example in an intangible is the team that had 9 key injuries the year before, and 8 of the 9 return at 100% after missing parts or all of the year before.

In 2018, there will be a case where a potential all-star junior quarterback will transfer from one Power 5 school to another and be immediately eligible.  This will have major effects on his new team, much more than if he were the top incoming freshman in the nation.  This QB has already proven himself in a power conference, while the top incoming freshman only proved himself in Georgia high school action.  There is a difference as deep as the Grand Canyon in comparing the two, even though the incoming freshman might eventually become the next John Elway.  The seasoned starter showed what he could do when he passed for close to 350 yards against Auburn, while the freshman threw for 220 yards against Troup County High School.

This is just one piece of a 5-prong process.  We rate each team against every other team using 5 different processes.  Once we are done, we use three different algorithms to come up with a PiRate, Mean, and Bias rating for each of the 130 teams.  The PiRate and Bias differ by very little, so they will frequently be highly correlated, while the Mean rating is somewhat different than the other two, because it takes the equally-weighted average of the five ratings, while the other two apply heavier weights to some of the five components.  Because they share the exact same weighting of the prong we showed you above, these two tend to stay correlated.

We know we have a lot of advanced mathematicians reading our blog, and we are grateful for your patronage.  We try to make this more analytically dominated than you might get from other sports rating sites and even break down the fourth wall from time to time to show you what we do.

If you are not a math fan and just love football, then fear not.  We still hope to have computer ratings that are as accurate as possible, and we still will make fools of ourselves weekly to issue free picks that you can then take and gift to Las Vegas if you are foolish enough to not realize that these picks are worth exactly what you pay for them.

We have 10 weeks to go.  With a lot of late night work, we hope to have the ratings ready for you in 9 weeks max.


The Captain


Blog at

%d bloggers like this: