YourLinx Ratings, Superlist, Interviews

Home ] Up ] The Retrodiction Question ] [ Carl Banks ] Todd Beck ] Jeff Bihl ] Richard Billingsley ] Bill Born ] Lee Burdorf ] Warren Claassen ] Tim Clavette ] John Coffey ] James Howell ]

 

 

Up

The Retrodiction Question
Carl Banks
Todd Beck
Jeff Bihl
Richard Billingsley
Bill Born
Lee Burdorf
Warren Claassen
Tim Clavette
John Coffey
James Howell

An Interview with

Carl Banks
Daniel College Football Ratings
NCAA Football
2002

YL: What are the most important factors in rating teams?

Carl: Good question. What factors are important in rating teams depends on what your purpose is in rating them. If your intent is to predict winners, then any factor that can be correlated to winning can be an important factor. If your intent is to make a lot of money by gambling, then your system should include not only on-the-field factors, but also off-the-field factors such as alumni tendency to bet on their own team. If your intent is to design a system to make your favorite team look as good as possible, the important factors are the things your team is good at. If your intent is to do it just for fun (like mine), then it doesn't matter.

But if your purpose is to decide what teams are to continue into the postseason, then the only factor that should be considered is wins and losses. You see, the purpose of the postseason is to to reward teams who win, not teams who score a lot of points or rack up a lot of yards. No one feels the need to use such stats to determine playoff teams in professional sports, or conference champions in college sports, except as a deep tiebreaker.

And there's no reason that college football should be any different. The problem with college football is that we cannot simply compare records, because of vastly disparate schedule strength. Thus, we need to take into account not only winning percentage, but the quality of the wins and losses.

YL: I always thought the conference setup was good until they began to have so many teams in conferences that they couldn't all play each other. If there were sixteen conferences and a playoff system, wouldn't the world be great? Since that's not going to happen, what role will computer ratings have in deciding who plays whom in the post season in the future? Will the BCS survive after this year's debacle?

Carl: The BCS has to survive until 2006, thanks to contracts, but something Roy Kramer was reported to say a couple years ago makes me suspect that it won't continue past that. Kramer was asked if there would be a playoff after 2006. It was a perfect opportunity to feed us more rhetoric about how, thanks to the BCS, a playoff isn't necessary. But instead, he said something like, "I never rule anything out." Hmm.

Concerning the role of computer rankings in college football, I think that their current popularity and official use is mostly a result of the tragic system we have in place. The bowl system settles nothing, and so we turn to computer systems in hope that they can resolve the lingering question of who was the best. A playoff would answer that question, however (well, at least it would answer the question of who accomplished the most).

Thus, the advent of the playoff will likely result in a decline in computer polls' popularity, and an end to their official use. Still, some sports, tennis for example, use computer rankings, so it is possible they will be retained in some way.

YL: Why did you begin doing ratings?

Carl: Curiosity. In a quasi-college-football-related chat room I hang out on, it's common to taunt fans of perennially strong teams, especially Florida State, with a long cascade of victories to "prove" that a certain lower division team was better than the perennially strong team. For example (about ten teams omitted):

Buffalo > Ball State > ... > Washington > Miami > Florida State

It was obvious to me that a long chain such as this was meaningless. I wondered whether a shortness of path criterion could be used to rate teams, and so I developed a system that awarded teams points based on the length of the shortest path, with the points decreasing exponentially as the path increases. Considering how ad hoc the system is (not to mention its mathematical flaws), I'm surprised how well it works.

YL: It is very interesting that it appears to do the job so well. I have looked at your final ratings for the year and they are very similar to most of the well-established formulas. How many weeks into the year does it take to start making sense?

Carl: The rankings start to settle down after most teams have played seven games.

YL: When did you begin to rate sports teams?

Carl: In 2000.

YL: Then you've come a long way in a short time. With such success in this area, what is your chosen profession - would it be, perhaps, something to do with mathematics?

Carl: Well, I major in Aerospace Engineering, and my research is on Flight Control. I suppose that involves a lot of math. Whether engineering is my chosen profession, however, is not certain.

YL: We all know about those "chosen professions." What teams or sports do you like best?

Carl: College football is the only sport I follow avidly enough to know how all the teams I don't root for are doing. I used to avidly follow Major League Baseball, but the recent labor troubles and the tampering by Bud Selig, the baseball Antichrist, have turned me away from it.

Still, having grown up in The 'Burgh (it's not hip to say Pittsburgh these days; gotta say The 'Burgh), I will always love the Pirates. I never was much of a Steelers fanatic, though. Considering how bad I felt about their recent AFC Championship Game loss, I can't even imagine how bad the really crazy Steelers fans felt.

I call Penn State my alma mater, and I root for the Nittany Lions and worship Joe Paterno. My first college football memory was the greatest moment in Penn State history: the 1987 Fiesta Bowl. Now that I attend Virginia Tech, I root for the Hokies, too.

But my favorite team ever is the 1999 National Champion Penn State Women's Volleyball team. I absolutely loved them because they were always happy and they kicked butt.

YL: I like women's sports. Talk show hosts annoy me to no end criticizing women's sports and any new sports group that comes along. Then, if it folds, they say: See, nobody wants to see it. Women's basketball is the best to me. I've grown quite weary of the slam dunk. Get rid of it and make those guys shoot the ball like the girls have to.

But all that aside, what other interests do you have?

Carl: I have a lot of intellectual interests. I guess the main ones are linguistics and computer science. I enjoy literature. Other than that, I'm pretty boring.

YL: Well, I don't think you're boring and I think your ratings are a rather interesting concept. Where do you call home?

Carl: Thanks. Although I'm attending college in Blacksburg, Virginia, I consider Pittsburgh my home and I'm remain a Pennsylvania resident. I'd like to continue living there, but alas, it's probably easier to get a job as an oceanographer in Pittsburgh than as an aerospace engineer.

YL: What else would you like for your fans to know about Carl Banks?

Carl: I have more information about me than anyone wants to know on my web page.

We would like to thank Carl Banks for the interview. You can visit his site at
http://www.aerojockey.com/danielratings/

Here are some other places you will want to visit:

 

Copyright 1998-2002 Nutshell Sports
Webmaster