Thursday, January 22

Fisking the ESPN Prestige Rankings

ESPN has put together a "prestige ranking" of all the I-AA college football programs. "ESPN's Prestige Rankings are a numerical method of ranking the best FBS college football programs since the 1936 season." Obviously, this is more of an exercise that is just for fun and not to be taken too seriously. But at times, I like to take frivolous things more seriously. It's the nerdy statistician in me. So let's take a look at what they did and where they may have erred.

Their system:
National Title: 25 points
Major bowl berth: 10
Major bowl win: 10
Conference Championship: 10
AP Top 5 Finish: 10
AP 6-10 Finish: 6
AP 11-25 Finish: 4
Heisman Winner: 8
Bowl Appearance: 3
Bowl Win: 3
10-win Season: 2
Week as AP #1: 2
Win over AP #1: 1
Each All-American: 1
First Round NFL Pick (since 1970): 1
Losing season: -2
TV Ban year: -1
Bowl Ban year: -2
Probation year: -1
Financial-aid Penalty year: -1
Recruiting Penalties year: -1
Each penalty of "show cause action:" -2

What they definitely got right:
  • 25 points for a National Title is dead on.
  • They handled the conference championship perfectly, and have the right amount of points on it (see their page for details or just trust me).
  • 8 points for a Heisman trophy is probably right. Perhaps it could go to 10, but it is a major aspect of the sport, even if you disagree with it.
  • The AP finish rankings are pretty much perfect.
  • When it comes to one point for each All-American, you could debate that this isn't important and that team success should supersede everything. But everywhere I went, the great players were integral to the thoughts and feelings that everyone expressed to me. At LSU, I heard no less than 40 renditions of Billy Cannon's punt return vs. Ole Miss. Even the youngest Dawgs wanted to discuss Herschel Walker's amazing freshman season. Tommie Frazier could run for governor of Nebraska one day and win in a landslide. When talking about prestige, certainly the great players who made the great plays matter. Do offensive linemen really belong in this category? Maybe not, but to keep the methodology consistent, you have to include them.
What they got wrong:
  • Including points for a 10-win season is very problematic. It inherently means less than it used to, thanks to the additional games on the schedule and the fact that those additional games are almost exclusively facing the weakest competition available. I don't think that beating Georgia Southern really improved Georgia's prestige any, but it did get them to 10 wins this season. Furthermore, it shouldn't be included because it is redundant with all the other measures. In statistics, it's akin to what we'd call an overspecified model.
  • Including each week ranked as the AP#1 and giving it two points, yet counting a victory over the AP#1 with only one point seems totally backwards to me. Alabama held their #1 ranking the week they beat Mississippi State. Did that really add to the program's prestige in any way? The sport of College Football is inherently about the big moments. And there is no bigger moment in the sport than taking down the #1 team in the country. This should be worth 10 points on its own. Just ask Texas Tech fans. Crabtree's touchdown against #1 Texas trumps any prestigious moment in the history of the program and will until they win the conference (and even then, there will be some who argue that this was bigger).
  • Strength of Schedule is basically absent outside of the one point for taking down the AP #1. It's a part of the game (well, it used to be anyway). I realize it would be hard to include this, but the "big games" are important. Perhaps including nationally televised broadcasts would have been good (though quite unfairly skewed as well). Or at least a point for every win over an AP ranked team.
  • First round NFL Pick doesn't belong here. Does Ryan Leaf's NFL experience add to the prestige of Washington State? He was chosen 3rd overall. Also, since it only goes back to 1970, it will skew the data to more recent success. ESPN still has trouble understanding that the two sports are not the same. Also it's relatively redundant with the All Americans.
What's missing?
  • Part of a program's prestige has always been wrapped up in its all-time great coaches. Can you think of Ohio State without Woody Hayes? Or Alabama without Bear Bryant? It would not have been hard to include points for any coach that was with a school for at least ten years and had at least a .700 winning percentage. For Penn State, give Paterno credit for each decade of tenure.
  • Because their system goes season by season, there is nothing for all-time records. For most fans, this certainly plays into the argument for the prestige of their school. It would not have been hard to include a certain number of points for every 100 wins all time. Yes, there would be some redundancy, so perhaps this is not needed.
  • Nothing for attendance? Shouldn't Miami should be penalized for their lousy fan base? Fans are a part of the game, too, and certainly affect the prestige of a program.
1936?
They actually address this issue, saying:
The AP poll was introduced that season, making it the first time the longest-standing news organization in the United States began ranking teams and crowning a national champion. Starting in 1937, the NCAA began recognizing "major college programs" (now known as the FBS). To accrue points, a program had to be recognized as one of these major programs by the NCAA.
This is a curious decision to me. When we are talking about "prestige", that doesn't precisely equate with "history", but they're certainly related. To begin in 1936 is to begin after the death of Knute Rockne. In these rankings, we have no Four Horsemen, no Red Grange, none of the "point a minute" Fielding Yost teams, and none of the great Army teams from that era. However, all these seasons matter to many fans today, and are talked about specifically in terms of prestige. Go to South Bend sometime and see if nobody quotes Grantland Rice to you. I promise you it's not possible.

The omission of all early seasons will knock down Michigan and Notre Dame at the least. Perhaps this is by design? As I learned on the road last season, Notre Dame is universally resented around the nation. Michigan is coming off a down year. They have two of the most active internet fan bases and will surely be irate about a lower ranking, causing them to talk about this endeavor that much more. Uhh... kinda like I'm doing right now. It seems like their rationale here doesn't really meet with the overall goal of the project, so I find myself skeptical about the decision. I mean, shouldn't Michigan beating Stanford 45-0 in the very first bowl game add to its prestige a bit? We'll see how things net out as they reveal 1-10 over the next two days. So far, #s 11-119 feel pretty accurate.

Overall, these rankings make a lot more sense in the pre-BCS, pre-horrendous scheduling era. If anything, they point out how much of the tradition has eroded in the last ten years. If I had to bet, my money would be on Nebraska or Oklahoma to take the title. Those "bonus" scholarship players and years of beating up on the other 6 teams in their conference are going to pile up the points.

No comments:

Popular Posts