Sunday, December 12, 2010

Auburn and Oregon in Context: Strength of Schedule

Following up on my post on Stewart Mandel's point regarding Auburn's and Oregon's schedules, I decided to look at their schedules in context with every other team that has played in a BCS title game.  Rather than using opponents’ records as the barometer, I used the three computer rankings that go back to 1998: Sagarin, SRS, and Sorenson.  (I must have a thing for rankings that start with S.  I blame my kids’ consumption of Sesame Street.  Or maybe the fact that they were fighting over Steamer & Samuel this morning.  Whatever, it’s the little ones’ fault.  Definitely.)  Using opponents’ records is a simple, but weak way to evaluate strength of schedule because it rewards teams that play in conferences with smaller numbers of conference games and/or whose members tend to play tomato cans outside of the league.  Here are the ranks of each of the 26 teams that made the title game, sorted by the average of the three rankings:

Sagarin Sorenson SRS Average
Florida 08 4 4 5 4.33
Florida State 98 5 3 5 4.33
Bama 09 2 10 2 4.67
USC 04 7 5 5 5.67
Oklahoma 08 7 9 2 6.00
Florida 06 8 10 4 7.33
Florida State 00 11 2 11 8.00
USC 05 8 16 9 11.00
Florida State 99 11 9 14 11.33
Oklahoma 04 15 10 11 12.00
LSU 07 11 12 14 12.33
Oklahoma 00 14 13 14 13.67
Texas 05 13 15 23 17.00
Auburn 10 15 24 14 17.67
Tennessee 98 24 22 26 24.00
Miami 01 27 27 22 25.33
Nebraska 01 29 26 29 28.00
Ohio State 02 30 42 21 31.00
Oregon 10 19 38 38 31.67
Texas 09 38 31 26 31.67
Oklahoma 03 39 29 32 33.33
Ohio State 06 38 35 34 35.67
LSU 03 28 38 41 35.67
Miami 02 37 34 37 36.00
Virginia Tech 99 43 35 58 45.33
Ohio State 07 53 51 57 53.67

The first point to make about this chart is that it underrates Auburn and Oregon because it compares their pre-championship game schedules against 24 other teams’ post-championship game slates.  After the title game, both Auburn and Oregon are going to find themselves about five spots higher in the rankings, so Oregon will be in the middle of the chart and Auburn will be right around tenth.  The conclusion to be drawn is that the e-mail that led Mandel to create his table – a message claiming that Auburn had an extremely hard schedule en route to Glendale and Oregon had an extremely easy one – is totally wrong according to the computers that aren’t castrated by the BCS.  Auburn played a tougher schedule, but not by a wide margin and neither team is an outlier in the 26-team sample.

It is worth noting, however, that the general point stands that strength of schedule has predictive power in title games.  The top six on the list all won the title game, with the exceptions of Oklahoma ‘08 (they were playing a team above them on the list) and Florida State ‘98 (they were starting the Rooster instead of Chris Weinke).  The bottom seven on the list all lost with the exception of ‘03 LSU, who played a fellow member of the bottom quartile. 

Also, the list gives us more reason to heap scorn on the Big Ten.  Leaving aside the fact that only one Big Ten team has made the title game over a 13-year period (joining the ACC as the only conference about which that statement can be made), Ohio State has two schedules in the bottom five of the list.  The best of the three Buckeye teams to make the title game had the 18th best schedule of the 26 title game participants and will likely be 19th after Oregon meets Auburn.  This isn’t a jibe at Ohio State, as the Bucks can’t be faulted for the state of their conference rivals.  This is a jibe at the rest of Jim Delany’s conference, which produced neither title contenders outside of Ohio State nor a depth of good opponents to challenge the Bucks in the three season in which they made the title game.

1 comment: said...

Thanks for this article, really useful piece of writing.