HOW THE RANKINGS WORK


I am not a computer programmer.  I am a college FOOTBALL FAN.  I put the FANATIC in the fan.  I’m also a CPA who uses spreadsheet programs daily.  I’ve come up with a model for determining the power ranking of the Division I teams for the entire season.  I did this for myself and for other football fans.  I’m not trying to impress anyone or score a bundle with the bookies.  I just wanted to develop a model that makes sense.

This is a “computer” power ranking because it was prepared using a computer spreadsheet program.  In theory, this could have been done on paper.  Like I mentioned earlier, I am not a computer programmer, so, if I used any boolean equations, it’s purely accidental.

If you’ve looked at these results in the past, you may notice a few changes. The changes include expanding the rankings to cover FBS and FCS teams, having a static strength of schedule (SOS) formula, expanding the formula for winning and losing games points, expanding the historical adjustment formulas, adding an additional source of points and changing the formulas for home/away and margin points are calculated.

The biggest change is that I’m now ranking all of the FCS and FBS teams. I didn’t want to do this, but Bama made me do it. Their schedule the past few years were so weak, that it simply threw off my results. Plus, with several teams jumping from FCS—I’m writing 1-A and 1-AA from this point forward, to 1-A from 2013-2015, I needed a way to rank them. To adequately accomplish this, I had to go all the way back to the beginning of the rankings and add all of the teams and their results. This is why I stopped publishing the results mid-season last year. Part of the reason I did this was with the old method, I tried to determine a general SOS factor weekly for 1-AA teams. Not only was it nearly impossible, it was impractical and not fair to the 1-A teams that played the relatively tough 1-AA teams. This required me to expand the current equations and enabled me to add the current SOS formula, which is a hell of a lot better than what the BCS said they used.

The SOS formula is calculated by taking a team’s opponents’ total game points from week 2 onward, subtracting the impact of that team’s game from the total, dividing it by the number of weeks that have passed in the season, and adding all of the amounts for the opponents played so far together.  Much better than looking at a team’s opponents’ winning percentage and combining it with their opponent’s opponents’ winning percentage like the BCS did. Especially when they were excluding the games played against 1-AA teams. The reason this is better is because many weaker schools play one or two strong teams and many weak teams, and they often play the same teams. These weak teams may have a decent record against their conference mates, but it doesn’t mean that they are a strong team for it. The result of a formula like this—Northern Illinois getting waxed by Florida State in a BCS bowl.

The new source of points is a formula that awards/subtracts points based on its opponent’s current ranking. The other formula of assigning points based on the teams’ relative ranking is also being used. Like all of the computer rankings, teams that win are rewarded and teams that lose are punished. Using these formulas in conjunction further rewards teams that beat better teams more than teams that beat sorry teams, and losing to better teams doesn’t cost as many points as losing to sorry teams.

I really did not want to add preseason rankings, but I realized that not doing so would mean that a win over Alabama State in week 1 would be worth as much as a win over Alabama, and I couldn’t have that. The preseason rankings are based on, but not exactly equal to, the historical composite rankings for each team over the past four seasons. Now, in these rankings, no team is protected by their early ranking. What I mean by this is that this year’s preseason number one will not stay in the top 10, or possibly the top 100 if it loses in Week 1. Even if it wins, another team or thirty teams could pass by them with better wins.

Formulas for home/away and margins are intertwined and depend on relative ranking of the teams. Teams will get points for winning away and/or winning by 17 points or greater. Losing at home and/or by more than 16 points will cost additional points. The amount of points depend on where this team and its opponent are ranked.

Another HUGE change is in the historical adjustment. I decided last season to switch from a per conference adjustment to a per team adjustment. This was also a huge amount of work, but I want to present the most accurate results to you as possible.

I hope you enjoy watching the results as the season unfolds.

 

Thanks,

Bill

3 thoughts on “HOW THE RANKINGS WORK

  1. Thanks for your effort. Enjoyable! Love the comments also.

    No preseason rankings makes so much sense and gave me hope for your process. Then, after claiming to be a college football fan, you make a statement like this:

    “I have not and will not rank lower division teams. Granted, we will probably see two to five Division I teams lose to lower division teams this season. This does not mean that the lower teams should be ranked. It means that the coaches and players of the Div I school had a really bad game”

    I don’t blame you for not ranking lower divisions, it would be difficult to draw realistic comparrisions. But really:

    “It means that the coaches and players of the Div I school had a really bad game”

    This shows that yoiu either don’t know as much about college football as I was ready to give you credit for or you have bought into the ESPN fueled mania and hype promoted by the “Big Boy” wannabees.

    There are a few FCS schols that could consistently play with many of the DI also rans. You should not dismiss those accomplished teams and the many talented athletes and coaches with such a statement because of the division they compete in.

    Like

    • Thanks for reading, Layne, and thanks for caring enough to comment. I do know that there are many talented skill players and linemen playing for lower level teams. I also realize that there are several talented coaches there that could be successful if given the same opportunities afforded to Mike Leach and Chip Kelly.

      This being said, the best of FCS can certainly compete with the worst FBS teams on a single-game basis, but they would have a difficult time competing with established, bad, FBS teams for an entire season due to their lack of depth. These schools can offer a maximum of 63 scholarships (spread to as many as 85 players) and are allowed 10 fewer walk-ons per season as FCS schools.

      Look at how Utah and TCU fared last season after stepping up their level of weekly competition in their new conferences. Though each team was extremely successful the previous ten years, most of their competition was horrible. North Dakota State may win its first couple of games this year if it were competing in the Sunbelt or MAC, but they probably could not handle the weekly pounding on their linemen. Injuries on the lines enable injuries to their skill players.

      Besides this issue, there are two other reasons I don’t rank FCS and lower teams. The first is that it would double the size of my model which would significantly increase the time required to get it ready each season. Secondly, I simply do not know enough about the FCS teams and conferences, and I don’t have the time to learn.

      If I can get this site to the point of having paid members, I may dedicate the time to add the FCS teams, but to do this in an accurate manner would require adding them to prior year models.

      Help spread the word about this site, and I’ll give this some serious consideration. I am trying to give as accurate an account of the games as possible.

      Thanks,
      Bill

      Like

  2. Hey Bill,
    I know one thing…you got way too much time on your hands! Just kidding, but I can’t imagine how much time you put into this theory and points system. I had to read it through a couple of times (that’s why I deal drugs and you work with numbers) to fully grasp the concept. I am definitely interested in how this will play out this season. Ray

    Like

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s