Tuesday, August 16, 2016

What Does it Take to be Ranked?

One of my favorite aspects of college football is the poll system. It is a fun, flawed way to look at how those who (allegedly) closely follow the sport view certain teams. Many professional sports websites have their own ‘power polls’ that track who a given columnist believes is the best team. However, there is no ‘official’ body that conducts a poll of the best professional sports teams for a given season. College basketball also has a poll, but polls are more valued in college football. A top-ten ranking in college basketball means little to a team that flames out early in March Madness. Meanwhile, a top-ten ranking in college football is still worn as a badge of honor even if a team loses their bowl game. With this in mind, I wanted to take a semi-statistical analysis overview of polls in college football. More specifically, I wanted to answer the question: What exactly does it take for a team to be ranked? The next paragraph will outline the methodology if you are interested. Otherwise, skip on down.

I looked at the final AP Poll data for the eleven year period of 2005-2015 and identified the number of losses by each team ranked in the final poll. I used the AP Poll for no other reason than because it is easily available at sports-reference.com/cfb. I used the final AP Poll because it was much easier to look at one poll per year for eleven years than a weekly poll in the same period. I used the period beginning with 2005 because that is when the first era of modern realignment began. The ACC hosted their first title game and the Big East began life in its final form after raiding Conference USA. In addition, college football schedules have been pretty uniform (in length if not quality) since 2005. Schools played eleven regular season games in 2005 and have played twelve each season beginning with 2006.

Here are the number of teams from each of the Power 5/BCS conferences that have finished ranked with a certain number of losses in the past eleven seasons.
If you play in a Power 5 or BCS conference and you finish with two or fewer losses, congrats. You will finish in the final poll even if you lack a great football history. If you finish with three losses, there is a pretty good chance you will finish ranked. Lose four games and you are still better than 2:1 odds to finish in the final poll. However, lose one additional game and you can pretty much kiss your chances of finishing ranked goodbye. Speaking of, who was the highest ranked team of the last eleven years to finish with five losses? That would be a tie between South Carolina (2010) and Auburn (2014) at 22nd.

As you can see, I included each iteration of the Big East in this analysis because they were a BCS conference despite being diminished compared to their previous incarnation that included Miami and Virginia Tech. I also included the first season of the American Athletic Conference as they retained the Big East’s automatic BCS berth. While that year’s AAC champ actually won their BCS game and finished in the top-ten, there is a pretty good argument to be made they did not have the strength of a typical BCS conference. Here is the overall version of the previous table with the Big East and American Athletic Conference removed.
The results are a little different. Only a single team from a Power 5/BCS conference has not finished ranked with three of fewer losses. Care to guess who it was? It happened just last season and they play in what is widely regarded as the toughest conference. The Georgia Bulldogs. Home of The Hedges, Vince Dooley, and Herschel Walker apparently don’t get the benefit of the doubt from the national media. Once the Big East and American are removed, you still have a nearly eight in ten chance of finishing ranked with four losses. The percentage with five losses is relatively unchanged.

Here are the number of teams from each of the non-BCS/Group of 5 conferences that have finished ranked with a certain number of losses in the past eleven seasons. The two most recent seasons of the American, when they have not had access to an automatic BCS bowl berth are included. You should recognize the other five conferences and lets hold a moment of silence for the WAC.
If you do not play in a BCS/Power 5 conference and you want to guarantee you finish the season in the final poll, you have little margin for error. If you finish unbeaten or have just a single loss, a spot in the final poll is reserved for you. However, with two losses, there is a one in four chance you will not have a shiny number beside your name at the end of the season. A closer inspection of the table will show that as long as you don’t play in one of the very bottom leagues (MAC or Sun Belt), you should be alright with two losses. The only other team that failed to finish ranked with a pair of defeats was Fresno State (and they were walloped in their bowl game). Three losses means you should probably give up the dream of a ranked finish and if you lose four or more games, you can pretty much forget it.

Obviously this analysis failed to include some important variables: Teams that win their bowl game probably tend to finish ranked more often than teams that lose their bowl game despite having the same number of losses. Teams with a better scoring margin (more blowout wins and fewer blowout losses) probably tend to finish ranked more often than teams with a poorer scoring margin despite having the same number of losses. Teams that lose in a conference title game probably tend to finish ranked more often than teams that don’t reach their conference title game despite having the same number of losses. Those are just three that come to mind. I’m sure there are others. That being said, I was still a little disappointed, being a mid-major apologist and all, that non-BCS/Group of 5 teams had such a small margin for error in being ranked. Even if they are not ‘really’ one of the 25 best teams in the country, a poll ranking should be a ‘Gold Star’ congratulating them on a fine season despite systematic disadvantages. Here’s hoping Appalachian State or Georgia Southern can break the glass ceiling in 2016 and become the first ever Sun Belt conference member to finish the season ranked.

No comments: