American College Football: Pre-season vs. Post-season Rankings

The Game Room
1

Do you know of a service, a website, or a publication that has historical analysis of these measures of football programs’ strengths?

So far, the best I have found is Preseason Prediction Accuracy (1993-2009)

Last year’s accuracy table at Preseason Prediction Accuracy (2009) shows a much wider variety of ranking entities, so there may be some problems getting an overall evaluation of the ranking services as such.

My main point of curiosity is the idea of arranging the schedules each year to correspond with something like a Round Robin tournament within the Top 12, Next 12, etc., so that the BCS and a true playoff system could be approached using pre-season rankings as a way to set up the equivalent of “conferences.”

If there’s any way to show that the pre-season rankings are any way to judge teams’ strengths, it might be worth further discussion.

2

Good luck with that. I think preseason rankings are mostly a crap shoot. How else do you explain Clemson? :slight_smile:

I do think I understand what you want to acheive, but I don’t really know if it’s feasible under the current system. Now, football games are planned out years in advance, not a month or two before the season starts. The logistics involved with last minute scheduling would be mind-boggling.

One problem I see, people often rank teams in the preseason based on how good they think a team is, not the projected final record. A team might be ranked very low but have a cream-puff schedule. But with such an easy schedule, they finish with a good record and a higher ranking. Conversely, a team might be really good and is given a high preseason ranking but has a brutal schedule. They thus finish with a worse ranking. Preseason rankings (quality of team) are not the same thing as postseason rankings (actual record).

What you might want to do is use preseason bowl projections instead of rankings. It will give you a better understanding of how people think a team will perform record-wise. I spent a couple minutes on Google, but I didn’t see any good comparison sites.

Let’s say your system is adopted, what happens with injuries like Oklahoma from last year? All preseason rankings had them in the top 5 but a few stars were injured and the rest of the year was decent but not great and finished unranked. 2 starters were injured right before the season started and Bradford was injured in Game 1. Instead of being in a highly competitive Top 12, they finished the season unranked and would have been demolished in your fluctuating conference scenario.

There are a lot of other reasons to avoid this situation, but I’ll leave it there for now.

3

I was able to locate 2009-2010 College Football Season Final Congrove Computer Rankings Through games of: 1/7/10 where you can sort by various columns to see how rankings changed over the course of the season. Of course, this is only one of dozens of ranking systems or entities and some meaningful way of merging (or at least weighting) the results of the various ranking services would have to be addressed.

Wolverine, I recognize all the problems you allude to, and I was hoping to find some meaningful way to address such subjective issues as “strength of schedule” and “out of conference schedules” and all the other factors that keep rankings from meaning all that much. If there were a clearing house for scheduling the Round Robin games in the 12-team groupings based on some rational pre-season rankings, then the ordeal of bypassing schedules set in years past might be minimized in its complexity. A computer program could do the dirty work, since all the schools would know who they would be playing that season.

If we are ever to see something approaching a meaningful National Champion in college football, I believe the first thing we have to jettison is “the current system” which fails on too many levels. The biggest barrier I know of is the power vested in the conferences themselves. The inertia there makes any improvement that would satisfy the fans pretty well doomed to start with.

One simple approach to the issue might be something like The College Football Belt which appeals to me in its directness of facing the issue.

4

One thing you could do is institute a promotion and relegation system into “conferences” or pools. The winners of those pools would face a small playoff to determine a winner. Similar to the BCS, create 6 pools of 12 teams each in a similar geographical area (not everyone can afford to constantly travel around the country) and they play a round-robin schedule only within their pool. Take the 6 pool winners plus 2 at large teams and seed them to make a playoff.

The promotion and relegation system will reward teams that are doing well and punish the teams that are not. The bad teams are dropped from the pool and replaced with a good team from the same geographic area that would be next in line. For example Indiana is dropped from the Midwest pool (formerly Big Ten) and replaced with Central Michigan. Or drop Washington State from the West pool and add Utah or Boise State.

The high turn-over of student athletes every year makes promotion and relegation a little bit more tricky, but it might satisfy what you’re looking for.

5

i’ve gotta agree with wolverine. pre season is just a crap shoot to a great extent.

it’s kind of like the grading of nfl team’s drafts right after draft day(s) are done. just a bunch of talking head nonsense. the real grade of a team’s draft won’t be measurable until a couple of years out.

6

For sure. I suspect a major problem with the ranking systems in general is that turnover issue. Recruiting “experts” are being tasked with helping to evaluate teams before they ever play their first game. I see this as a pitfall to

as well as the injury component. Depth chart evaluations might become more of an issue if the ranking systems were to be taken more seriously.

I’d like to hear more of your “promotion and relegation system” which sounds smart enough just from its implications.

Let me insert that I realize the economics of the current conference mentality and also realize that upsetting that apple cart might be more of a task than rearranging the way that Congress works! But if the conferences, as such, could remain intact while this football oriented grouping for the sake of scheduling teams in roughly equal power brackets on a yearly basis, subject to change as the teams gain and lose whatever “power” is involved, would address the issues that the BCS is obviously doing only halfway – if that.

A true playoff system is just too unwieldy and subject to derailing because of injuries and other factors unique to the nature of football. No system other than some “on the field” solution will ever satisfy me, and I suspect a large percentage of football fans.

7

Here’s the Wiki article on promotion and relegation. Link

The basic concept is that the top leagues should eliminate the worst teams (relegate) by sending them down a division. You replace their slot in the league by bringing up the best team in the lower division (promotion). Western Kentucky, Eastern Michigan, and Miami (Ohio) had 1 win total between the 3. Send them back down and replace them with Appalachian State, Montana, and Villanova.

For your scenario, you would want to create even more divisions than FBS and FCS. However, with the fluctuation of talent changing drastically every year, the top division of only 12 teams is too narrow. In 2003, LSU was 14 in the AP preseason poll and 15 in the coaches poll. In 2002, Ohio State was 13 and 12 respectively. In 2000, Oklahoma was 19 in the AP preseason poll. In 1997, Michigan was 14 in the AP. It would be better to expand your field a little.

8

Thanks for that link, Wolverine. I was unaware of the concept before. Makes great sense to me.

As for the number of divisions, I was hoping to have the Round Robin format determine the outcomes within divisions, so that the number of members would be determined by the number of games in the season. Twelve just seemed about right, but it could be moved up to 16 (or some other number) if the season were expanded accordingly.

And I would be content to throw all Division I teams into one stack and let something like Jeff Sagarin’s College Football Ratings determine the component teams by “division” since he has them all in the same stack.

Having some ultimate playoff of “division” winners would lead toward some National Champion, and the mechanics of that system would have to be worked out as equitably as possible within the 3- or 4-month football season. I’m not pretending to solve the entire problem, just the regular season games where heretofore so many questionable matchings are a big issue.

Having each “division” be more or less equally matched in strength would make for exciting games, if nothing else, and might make the cupcake schedule no longer a point of contention.

9

I wanted to come back to this issue because of an amazingly clever article in WSJ that my brother pointed out to me. I’d love feedback in this thread from anybody with an opinion on its merits and viability.

Hurry to see it because it may lapse as a live link soon.

Here’s How to Fix College Football

10

I saw it linked from Fark.com. (Despite that) I thought it was very clever. No idea if it would work, but it’s damn clever.

Biggest problem is that American sports fans are so damn adverse to the idea of promotion and relegation that I just don’t see it flying with them. And the ADs would say “Wait, so if we have a rebuilding year we could be stuck playing Akron and Appalachian State next year?” So, chances of this happening are zero, but it’s interesting.

11

Sadly, I agree with you.

I would just like to see a system where perennial dogs (like Vandy, my school) are provoked to shit or get off the pot, so to speak.

12

I really like the system they developed. It’s very well thought out and balanced. My favorite part is the unspoken assumption that the 4 winners of each conference will face each other in a playoff for the national champion. They completely avoid that 7 letter word altogether so that some people won’t automatically get upset.

13

I’ve taken the liberty of sorting the data at the bottom of the article on the dollars column. To avoid an outright copyright infringement, (I hope), here are the Top 25 in that measure:



School           Current Conference  New Conference  Conference Type  Football Revenue in 2009*  Point Total

Texas                    Big-12         Southern         Super             87,583,986                13 
Ohio State               Big Ten        Big North        Super             68,196,195                11     
Florida                  SEC            Atlantic         Super             66,150,063                13 
Georgia                  SEC            Atlantic         Super             65,218,406                 8 
Alabama                  SEC            Southern         Super             64,606,392                14 

LSU                      SEC            Southern         Super             61,868,953                 9 
Penn State               Big Ten        Big North        Super             61,767,717                11 
Auburn                   SEC            Southern         Super             58,618,819                 8 
South Carolina           SEC            Atlantic         Super             57,118,519                 6 
Nebraska                 Big-12/Big Ten Big North        Super             55,226,605                10 

Michigan                 Big Ten        Big North        Super             52,246,025                 5 
Michigan State           Big Ten        Big North        Super             43,506,725                 5 
Tennessee                SEC            Southern         Super             42,805,360                 7 
Oklahoma                 Big 12         Southern         Super             42,638,431                 8 
Wisconsin                Big Ten        Big North        Super             40,005,517                10 

Iowa                     Big Ten        Big North        Super             38,896,234                11 
Arkansas                 SEC            Southern         Super             38,630,265                 8 
Texas A&M                Big-12         Southern         Super             38,358,422                 6
Clemson                  ACC            Atlantic         Super             35,205,168                 8 
USC                      Pac-10         Pacific          Super             35,203,483                 9 

Washington               Pac-10         Pacific          Super             34,177,030                 5 
Oregon State             Pac-10         Pacific          Super             30,874,059                 8 
Arizona State            Pac-10         Pacific          Super             29,857,334                 4 
West Virginia            Big East       Big North        Super             28,952,873                 9
Colorado                 Big 12/Pac-10  Pacific          Super             27,827,286                 1

The whole list reveals many wonderful things. And Houston is by far the best ROI going!