BCS Mess 11/16/2012

By Michael Gertz
Friday, November 16, 2012

We all know the BCS has its issues. Apart from the lack of a playoff system quite yet, even the rankings themselves often fail to effectively decifer a team's true quality. Let's take a closer look at a few systematic failures of the rankings and how they affect the rankings this season, and analyze the odds behind who will make and win the BCS Championship. See Our Full Ratings

Outcomes Over Quality at the Top

The idea behind the BCS rankings is to rank the teams according to how good they are. But essentially what the BCS does is a little different, because it has to still heavily value wins and losses. Losing by just 1 point at the top team in the country is more impressive than winning by just 1 point against a bad team at home, but has to be judged by the BCS as less impressive, otherwise the rankings start to feel wrong.

As a result, most oddsmakers and sensible fans would agree that #4 Alabama is better than #3 Notre Dame and would be favored if they met in a bowl game, but Notre Dame must be ranked higher because they have not lost. In fact, our ratings of true quality of teams still have Alabama rated 1st overall by a slight margin, and Notre Dame just 10th. This is because Alabama has handily beaten many good teams while barely losing to 1 great team, while Notre Dame has just barely defeated several decent or even mediocre teams.

This same effect has impacted a few other teams near the top as well. #8 Texas A&M and #12 Oklahoma are both underrated because they happened to lose close games to top tier opponents. Our ratings have them 4th and 5th respectively. Meanwhile, though ineligible for the BCS, the AP Poll has Ohio State ranked 6th. Our ratings have them just 27th because they have faced an extremely easy schedule and even still barely won several games. The gambling community certainly agress with this assessment, as this week Ohio State are underdogs at unranked Wisconsin.

BCS Missing the Mark on Lesser Conferences

Every week it seems the BCS voters attempt to identify a small conference leader as a legitimate contender and fail. Because they likely do not know much about these conferences, they simply pick the team with the best record. Often, those teams are simply the ones that have had the weakest opponents so far and even then have barely managed wins, while others barely losing to stronger non-conference opponents and dominating conference foes are more deserving.

At the moment, this phenomenon is most apparent in the rankings of #19 Louisville, #20 Louisiana Tech, and #22 Rutgers. Our ratings have these teams respectively just 60th, 43rd, and 52nd in the country. More reasonable candidates include our 20th rated Utah State, 21st Fresno State, and 23rd Brigham Young. Once again oddsmakers back us up, with Utah State favored at Louisiana Tech this week, and Rutgers big underdogs at our 30th rated Cincinnati.

The AP Poll identifies a few other vastly overrated small conference teams with a chance to enter the BCS rankings. Teams receiving AP votes include their 25th ranked Kent State and Toledo. Our ratings have these teams 70th and 67th, and Kent State has yet to even play a team in our top 50.

The Percentages Behind the BCS Championship

The concensus opinion seems to be that #1 Kansas State and #2 Oregon will meet in the BCS Championship if both stay undefeated, regardless of who is ranked #1. Our ratings give Kansas State about a 92% chance of winning each of their final 2 games, which without a conference championship game gives them an 84% chance of staying undefeated. Oregon faces a little tougher road, vs #13 Stanford (96%), at #16 Oregon State (86%), and then a home conference championship game against the winner of this week's USC @ UCLA game. Both of those teams are rated close to Stanford, so Oregon would likely win that game about 95% of the time. This totals to a 78% chance of Oregon being undefeated, giving a 66% chance of a Kansas State and Oregon championship matchup.

Of the remaining 34% of scenarios, about half would likely result in Notre Dame taking part, as they have about a 50% chance of remaining undefeated. Most of the other half would probably involve Alabama, with a slim chance of Georgia should they upset Alabama in the SEC Championship. Using our current ratings, Oregon would be a slight favorite over Kansas State, Notre Dame would be a heavy underdog either way if they got in, and Alabama would even be slight favorites over any opponent if they snuck back in.