Welcome to Ray's Rankings!

My goal is to create a fully mathematical system that treats all teams exactly the same. There is no "eye test." There is no riding your reputation. There is no bias. Honestly, it's a pretty simple system, and as far as I know it is the only one that is 100% transparent (in that my math is out in the open).

People who do not understand the system will definitely not understand why it ranks Air Force over A&M some weeks (or a 3-3 BYU over a 7-0 Coastal Carolina). Truly, it is 100% data driven, and when we don't have enough data we can see some pretty wild swings. Usually that settles down around week 7 or 8.

Why Do It?

In any given season there are 125 to 130+ FBS schools. In general, the max number of games played by any school is 12. This means that any single team is going to play 9.6% of available opponents at best. Now factor in conferences and the fact that teams play 8-9 in-conference games. What you end up with is a relatively easy way to measure teams within the same conference and an almost impossible mission to measure teams across different conferences. The best you can hope for is common foes at similar times during the season with no injuries or other circumstances that might affect how the common foe plays. Most likely it will only be a singular opponent played at different times of the season with injuries and such making such a comparison almost irrelevant. 

What is the answer? Well, in high school football across the country there are usually far more than 130 schools vying for playoff spots in any given state. Similar to college, teams play almost solely within their conferences. How can they determine if a team from Sacramento should make the playoffs and a team from San Jose miss? Actually, California is a bad example as they use calpreps.com rankings are still "voted" on. 

Let's use Ohio (where I currently live). 

In Ohio they use a system that looks at who you play and who your opponents have played. They have different divisions like most states, and they're based on student enrollment numbers. Here's how Ohio works (straight from the OHSAA website):

Wins over a Division I school = 6.5 points 
Wins over a Division II school = 6.0 points 
Wins over a Division III school = 5.5 points 
Wins over a Division IV school = 5.0 points 
Wins over a Division V school = 4.5 points 
Wins over a Division VI school = 4.0 points
Wins over a Division VII school = 3.5 points

An example:

• Team A, a Division III school, is 3-0. 
• Each opponent Team A has defeated is Division III. 
• Opponents 1 and 2, which Team A has defeated, are each 2-1, with their victories coming over Division III schools. 
• Opponent 3, which Team A has defeated, is 1-1, with its victory coming over a Division III school. Opponent 3 has had one open date.

First level points - 16.5 divided by 3 (3 wins times 5.5 points for defeating Division III opponents divided by the number of games Team A has played thus far) = 5.5000 first level points.

Second level points -

• 11.0 from Opponent 1 (2 wins times 5.5 for defeating Division III opponents), 
• 11.0 from Opponent 2 (2 wins times 5.5 for defeating Division III opponents)
  
• 5.5 from Opponent 3 (1 win times 5.5 for defeating a Division III opponent)

is 27.5 points. These 27.5 points are then divided by 99 (since Opponent 3 thus far has had an open date) and multiplied by 10 (so that the points are in line with 1st level points) = 2.7778 second level points.

Total Points: 5.5000 (first level points) plus 2.7778 (second level points) = 8.2778

NOTE: Say during week four, Opponent 1 has an open date. Then the second level points divisor for Team A after week four becomes 98, since Opponent 3 has had one open date and now Opponent 1 has had one open date.

My Version

My system is based on the Ohio State High School Football playoff calculator as seen above (and I'm sure other states use the same or similar methods). There is a margin of victory modifier called the Simple Ranking System that I also use, and I penalize FBS teams for playing FCS schools.

• 1st Level - Teams earn points based on their current rank. There are 134 FBS schools in 2024, so each rank is worth 0.007462 pts.
  
• If you beat #1, you earn 1.0 point, #2 gets you 1-0.007462=0.992537 and so on.
• A win over an FCS school is 0 points, and losing to an FCS school costs you 2 points. Big boys shouldn't be playing the kiddos.
  
• 2nd Level - 1st level points of the opponents you beat (basically points from your opponents' opponents), average them with a modifier for off/bye weeks or shorter schedules. If the teams you beat play bad teams, you 2nd level points can actually go negative!
  
• SRS - Simple Ranking System scores. I apply a modifier to give this a smaller effect than wins.
  
• See here for the explanation of the SRS.

Add up the scores and the best is No.1.

Signature Wins and Quality Losses

I've wanted to find a simple way to account for the quality of competition from the start, but it needed to be 1) mathematical in nature, and 2) not screw up my way-too-complicated spreadsheets. So, using the point structure outlined above, I applied a reverse of it for losing. In other words, if you lose to #1, it costs you nothing. If you lose to #134 it'll cost you a full point. 

Not a Predictive System

My system is retrodictive, not predictive. It is designed to measure how teams have performed throughout the season, not to make a claim as to which team could beat another. That's what a playoff is for! I know that this can be somewhat counterintuitive when we look at rankings. For comparison, take a look at the AP poll. There are 134 teams, and 63 voters. The voters have no way to watch all the teams, so they put on their way-too-tight bias caps and vote for who they "think" would win between maybe 30 teams. When they're done, out pops a top 25. So the AP is a poorly created and not at all predictive tool. My system looks at what each team has done, and what the teams they played have done, and assigns arbitrary points to each win and loss (arbitrary, not random; in the sense that it could be 1 point or 10 points or 0.01 points).

 

Please see the right margin for links to my blog posts.

The current Ray's Rankings Top 25: