If you’ve never looked at football.stassen.com, you ought to take a look. It has a lot of great resources for college football research.
One of the best things on the site is the compilation of the preseason consensus. It allows you to see what everyone was thinking before the season starts, and it goes back to 1993.
I decided to go back and look to see how good the preseason consensus was at predicting BCS participants.
By Preseason Conference Rankings
I first looked to see where the BCS conference participants were ranked in the preseason standings.
As it turns out, there’s almost always one wild card in the BCS. Every year except two—1998 and 2003—had a team ranked fifth or below in its conference/division make the BCS. The 2003 season was anomalous in that the BCS had six first place picks and two second place picks in it. The 1998 season had five first place picks, two second place picks, and one third place pick.
The lowest-picked teams to make the BCS were 1999 Stanford and 2002 Iowa who both were picked eighth in their conferences.
Only three seasons had more than four first place conference/division picks make the BCS. I’ve already mentioned ‘98 and ‘03, and the third was 2001 with five. However rather than being totally orderly, ‘01 also had two seventh place picks make it in Maryland and Illinois.
Here is a table showing the frequency of making the BCS for each preseason conference/division ranking.
|Conf./Div. Ranking||Teams in BCS||Teams per Year|
First Place Picks
Here is a table of how well the teams that were picked first in their conferences or divisions did at getting into the BCS.
“1s in BCS” refers to how many first place picks from the conference made the BCS. “BCS Teams” tells how many total BCS teams the conference has fielded. “Tot. 1 Picks” tells how many total first place picks the conference had in the BCS era. For conferences without divisions, it’s one a year; for conferences with divisions, it’s two a year.
|Conference||1s in BCS||BCS Teams||Pct.||Tot. 1 Picks||Pct. in BCS|
We see a couple things in this.
The Big Ten’s first place picks make a very small percentage of its total BCS participants, thanks to having the smallest number of first place picks make the BCS but the largest number of BCS teams.
The Big 12 and especially the SEC had trouble putting their preseason first place picks in the BCS. In the case of the Big 12, part of it was guessing incorrectly as to whether OU or Texas would win the South. The SEC was a little messier, and I’ll explain more about that later.
I finally took a look at consensus teams. In this case, I defined a “consensus team” as a team that received three or fewer rankings of below first for its conference/division.
Surprisingly, only 29 of the 47 (61.7 percent) consensus teams made the BCS. Here is the number of consensus teams by year:
It’s interesting that from 1998 to 2003, there were never more than five consensus teams, and that only happened once. During that time, 69.57 percent of the consensus teams made the BCS.
In 2004 and 2005, there were six consensus teams, and in 2006 there were eight consensus teams. During those years where everyone suddenly agreed more often, just 45 percent of consensus teams made the BCS. Things toned down a bit in 2007, when all four consensus teams got to BCS bowls.
The Big Ten has only had five consensus teams, but all five have made the BCS. It appears that when everyone agrees on the Big Ten, its preseason champ makes the BCS; when everyone doesn’t necessarily agree, it’s preseason champ doesn’t make the BCS.
The Pac-10 has also had just five consensus teams. Only 1999 Arizona failed to make the BCS; the other four were USC from 2003-07. That’s an 80 percent success rate.
Next up in accuracy was the Big 12, which had nine consensus division winners in the BCS era. Seven of them made the BCS for a 77.8 percent accuracy rate. The two that missed it were 1999 Texas A&M and 2006 Nebraska.
After the Big 12 came the ACC. It had ten conference/division consensus teams, and six of them made the BCS (60 percent). Three of the failed four were divisional picks—VT in 2005 and both FSU and Miami in 2006—with only 2001 FSU falling short among pre-divisional play consensus teams (all of which were of course FSU as well).
We fall below .500 in accuracy with consensus Big East picks. Only two of the five consensus teams made the BCS, but interestingly the misses came from 2004-06. It shows that people were pretty sure they knew what was going to happen post-ACC raid, but we can see in hindsight that really no one knew what was going to happen on many levels.
Finally, we see that the SEC was the most difficult to predict. It had 13 consensus teams and at least one a year, but only five of them made the BCS. That’s a 38.5 percent success rate.
Funnily enough, every time Florida has been a consensus team (’98, ‘01, and ‘06), it has made the BCS. The other two hits were ‘99 Tennessee and ‘07 LSU. Eastern division consensus teams were 4-2 in making the BCS; Western division consensus teams were 1-6 in making the BCS. This seems to further confirm the fact that the SEC West is generally more wide open than the SEC East, which has only ever been won by Florida, Georgia, and Tennessee.
What about 2008?
As of writing this, only eight of the preseason rankings the site uses have been published for 2008. For comparison, Stassen.com used 18 in 2007.
What is amazing is the uniformity of the picks. Clemson and VT are unanimous in the ACC. Missouri and Oklahoma are unanimous in the Big 12. West Virginia is unanimous in the Big East (and USF is the only pick for second in the conference). Ohio State is the only one picked to win the Big Ten so far, just like USC is the only one picked to win the Pac-10.
Everyone is in complete agreement about the selection of seven of the nine conference/divisional top dogs. The only discrepancy? That fickle old SEC, where Florida and Georgia are split even in the East and LSU has a five picks to three advantage over Auburn.
We haven’t seen this much agreement since 2006 when only the Big 12 South was disputed. The success rate in that year was only three out of eight correct.
There is still time left and many more rankings to go, so we may not end up with quite this much consensus when the season starts. Just remember that if your BCS picks all fit with the conventional wisdom, history says you’ll only be about half right.