To modify a phase from Abraham Lincoln, “You can [please] all the people some of the time, and some of the people all the time, but you cannot [please] all the people all the time.”
That must currently be the reclassification committee’s world.
First, before looking at my updated region alignment projections, let’s note that I have the luxury of sitting in my cozy breakfast nook up here in Ohio nearly 600 miles from the GHSA headquarters and second guessing the arduous task of the reclassification committee – Were the cutoff sizes correct? Are the region alignments correct? Did they properly balance region size, travel distance, enrollment differences, school district boundaries, competitive balance, traditional rivalries, etc., etc., etc.? Were the concerns of football and non-football sports fairly addressed? Were the concerns of the public, private, and city schools fairly addressed? Were the individual concerns of all 455 GHSA schools fairly addressed?
Obviously the committee can’t win. There are simply too many competing interests and too many constraints to find the perfect solution from every perspective.
So, the reclassification committee must look for the only thing it can reasonably be expected to do find . . . a satisfactory solution.
In short, a solution good enough that it can’t be significantly improved upon with additional effort. The key is that “can’t be significantly improved upon” portion.
So, with that in mind, here's my renewed attempt at finding a satisfactory solution to this immense problem through a technique I outlined in my "waaaaaaay too early region projections".
As a summary, the objective of the program is to find the minimum Great Circle Distance (commonly referred to “as the crow flies” distance) between region members with the following exceptions:
- Pairings of schools more than 62.5 linear miles (approximately 100 driving miles) apart are allowed but discouraged
- Pairings of schools in the same district or currently in the same region are encouraged
- The minimum region size is four schools and the maximum is 1/3 more than the average region size for that class
Since there are so many teams in Class A, I did sixteen regions and leave it to the reader to pair them together to make each eight regions with two subregions each.
I only did football playing schools, although the exact same technique could be applied to the complete region alignments as well.
Some interesting notes (and perhaps discussion points):
- Moving to seven classifications and setting the Class A cutoff at 520 stretches the top six classes quite a bit. There are 48 schools in AAAAAAA for an average of six teams per region. For AAAAAA to AA, there’s about 56 schools per class for an average of seven teams per region. This region size sounds pretty ideal, except that Georgia’s uneven population distribution makes it difficult in some cases to make a region with a reasonable size AND reasonable travel distances for each of its members. As a result, each of the three largest classifications has one region with only four members.
- About 30 schools fall near or over the 100 mile “rule” for average region travel distances, although in the past it’s not been treated as a hard cutoff for playing down, especially in football. It’s not even clear (or at least not clear to me) how this distance is measured. Here I estimated it by multiplying the Great Circle Distance by 160%, but that simplified approach doesn’t take into account the specific routes a school would actually take.
- Camden County has the worst spot in the entire state. If we accept they must make an average three-hour drive only for football, then Camden County might be fine in AAAAAAA, but they should request to play down in other sports. However, that leaves Region 1 with only three schools for those other sports: Colquitt County, Lowndes, and Tift County. On the other hand, it could also be argued that Camden County can make a three-hour drive for all sports since there are only three region opponents . . .
- Gwinnett County has 17 schools in AAAAAAA, enough to make three regions of six, six, and five schools. But that would guarantee one school district with over a third of the 32 playoff spots. That might be fair given that Gwinnett County also makes up over a third of the classification, but I doubt it will be perceived as fair from all quarters. Another possibility is to make two regions with eight and nine schools. Eight seems reasonable since the average size is six, but nine seems a little large. Also, if Gwinnett can have a nine-team region, why couldn’t Cobb County, also with nine teams? That would put 26 schools in three regions and 22 schools for the other five. Chances are the committee sticks with the smaller and more balanced sizes.
- However, in the end, my conclusion for Class AAAAAAA is that aside from Region 1, practically anything could happen. Every school in the north is reasonably located near every other school, so travel distance may (and probably will) take second place to any number of other factors.
- Of course everyone is waiting to see if Valdosta opts to play up, but if travel is their concern they are pretty much looking at long drives either way. They are by far the longest in AAAAAA, but are the fifth largest school in the class. If they move up, they'll be the smallest school in the classifcaition and won't substantially improve their driving distance. Although school size matters less for large enrollments, that doesn't mean the officials at Valdosta think that. My guess is they elect to stick with AAAAAA.
- With a cutoff of 520 (89 football playing schools and 120 schools total), Class A could be wholly split into eight Public and eight Private regions. Another approach would be to split Class A further into Division I, Division II, and Division III. Suggestions on the cutoffs based on a simple model I made earlier would be 300 and 400. Jimmy Hughes, the head football coach at Dooly County, also suggested splitting Class A into two divisions with the cutoff at 300, but that would leave Division I with 62 football playing schools and Division II with 27. Splitting into three divisions with the 300 and 400 cutoffs would give each division between 27 and 33 football playing schools. Of course, the task of pairing up 30 schools spread across the state into regions with reasonable travel distances becomes more difficult if not impossible. In a trial run using those cutoffs with just four regions in each Class A "division", more than a third of schools would have an average estimated driving distance of 100 miles. The worst was Lincoln County, with 165.74 average miles. Not to worry, though, I've already proposed a solution for this as well – do away with Class A regions altogether.
- One final thought. Maybe the committee has done this, but for a problem of this magnitude I would highly recommend utilizing someone with combinatorial optimization experience. The below projections aren't perfect, but I would contend they are a good starting point. Also, other factors can be easily added in to meet other requirements of the reclassification. I can be easy found through the AJC :-)
Region projections
AvgEnrDiff = Average Enrollment Difference with other projected region members
AvgLnrDif = Average Linear Distance Difference with other projected region members
AvgDrvDif = Average Estimated Driving Distance Difference with other projected region members (AvgLnrDiff x 1.6)
All distances are one-way
“Group” designators are not intended to be the region designators, but are simply the numbering of each group going from south to north
AAAAAAA
AAAAAA
AAAAA
AAAA
AAA
AA
A
About the Author