Northern Illinois beat Notre Dame 16-14 this past Saturday. The Huskies were four-touchdown underdogs and won the game outright.
Through the first few weeks of the 2024 college football season, there have been 50 teams favored by 10 or more points. Eight of those favorites have now lost outright1.
If you compare this to the first two or three weeks of last season, five teams won outright as a double-digit underdog out of a possible 52 games2.
The problem is the Irish have been here before. A loss when you’re expected to win by 10 or more points is no good.
Over the past three seasons, seven teams have lost three games as a double-digit favorite: Notre Dame, Miami, Liberty, North Carolina, Air Force, Rice, and Clemson.
Notre Dame, Miami, and Rice are the only teams to have lost all three of their games at home as double-digit favorites. The table below shows the results and spread for each one of these games for these seven teams. All data from TeamRankings BetIQ Custom Trends tool.
Yet another drive chart
Notre Dame put together only two scoring drives in the loss to Northern Illinois. The Irish gained 50% or more of available yards on only three of its 11 total drives.
Northern Illinois capitalized on useful field position on three of its four total scoring drives. The Huskies marched 98 yards down the field on its the opening possession of the game to score a touchdown too.
The full drive chart3 highlighting yards per play and available yards gained per drive is shown below. All data from cfbfastR.
Monday morning quarterback
It’s clear the Irish did not get quality quarterback play in the loss over the weekend. After watching a full weekend of football, the variance of quarterback play is everywhere.
From Notre Dame’s Riley Leonard to NC State’s Grayson McCall to Syracuse’s Kyle McCord or Vanderbilt’s Diego Pavia, there are a wide range of outcomes in performance from a transfer quarterback. The sample size is also quite small.
We're in an era of major college athletics where continuity or returning production might not mean what it used to mean for a team's success. And how could it?
If players are free to move season to season, more and more will move, and the more that move, the more unpredictable the results.
Head-to-head conference records
One sign of maturity is taking the right things seriously. Media and college football fans take the conference pride or conference versus conference records too seriously. Fans are fanatical though.
David Hale has an insightful thread on the head-to-head conference records. The overall records don’t tell you too much, especially with so many teams in each league.
What does it mean when Virginia Tech loses on the road as a 13-point favorite at Vanderbilt or when Cal4 wins at Auburn as a 12-point underdog?
Or if Miami blows out Florida in Gainesville or Tennessee whomps NC State in Charlotte?
There are 33 combined teams in the ACC and SEC.
Some teams are good, some teams are bad.
You can find the code for both of these charts and others here. If you’re looking for more recommendations like last week, here is another plug for TeamRankings.
TeamRankings Bet IQ and the Custom Trends tool is super handy to search for old point spreads or totals. The database goes back to the 1996 college football season. TeamRankings also has a ton of other data sources.
UL Monroe 32, UAB 6 (UAB Monroe -11)
Cal 21, Auburn 14 (Auburn -12)
Northern Illinois 16, Notre Dame 14 (Notre Dame -27.5)
BYU 18, SMU 15 (SMU - 12)
Sam Houston State 34, Rice 14 (Rice -10)
Vanderbilt 34, Virginia Tech 27 (VT -13)
Boston College 28, Florida State 13 (FSU -16.5)
Georgia Tech 24, Florida State 21 (FSU -10.5)
Texas State 42, Baylor 31 (Baylor -26.5)
Duke 28, Clemson 7 (Clemson -21)
FIU 46, North Texas 39 (North Texas -10.5)
Colorado 45, TCU 42 (TCU -21)
Wyoming 35, Texas Tech 33 (Texas Tech -13.5)
NIU head coach Thomas Hammock on the Dan Patrick Show today is a good listen. Hammock is a great quote.
This drive chart is a bit different than last week as it focuses on available yards. I still think including third down conversions would be useful too, but that would involve some more elbow grease to scrape and join together the play-by-play data.