Of course any team is vulnerable to a letdown after a big win. It happens even to the greatest of coaches and teams.
But more than either Davie or Holtz, Willingham was vulnerable to sequences of games that were just impossible to explain on any rational level. After drilling Florida State in Tallahassee in 2002, the Irish lost to a thoroughly unremarkable Boston College team.
In 2003, after apparently righting the ship a bit by winning three games in a row, including a 50-point road win at Stanford, ND was whipped by a then-5-6 Syracuse team.
In 2004, after looking dreadful on the road in the opener against BYU, ND completely dominated Michigan in the second half in route to a win. Then after winning three in a row, concluding with a five TD victory over Washington, ND looked like a weak Pop Warner team losing 41-16 to a Purdue team that would finish the year at 7-5.
Later, with things apparently going well again with a road win over Tennessee to get to 6-3, ND suffered defensive breakdown after breakdown to lose at home to a Pittsburgh team that had to go to overtime, at home, to beat I-AA Furman.
To try to measure the level of inconsistency, I went to Howell's end-of-the-year ratings for teams from 1986 through 2004. I broke teams down into those that finished 1-20, those that finished 21-40 and those that finished 41st or worse.
Let's look at how each coach performed against teams at these levels (for these purposes I counted the Insight Bowl loss to Oregon State as a Willingham game, even though the record is technically assign it to Baer).
HoltzTop 20: 28-22-2 (.558)
21-40: 20-7 (.741)
41+: 52-1 (.981)
DavieTop 20: 4-12 (.250)
21-40: 5-8 (.385)
41+: 26-5 (.839)
WillinghamTop 20: 6-7 (.462)
21-40: 3-6 (.333)
41+: 12-3 (.800)
Of course, Holtz was by far the best coach, but his teams were predictable in the sense that they essentially never lost to a team outside the top 40. (His only loss of this sort was to Air Force in 1996, a team that finished 6-5 and ranked 45th). His teams were an excellent bet to beat teams in the second 20 and well above .500 against the top 20.
Davie's teams were predictably mediocre. They were a good (but not sure) bet to beat teams outside the top 40. They were under 50% to beat a team in the second 20 and a longshot to knock off a team in the top 20.
Willingham's distribution, however, is just plain odd. His teams actually were nearly .500 against the top 20, but much worse against the second 20. His win rate against teams outside the top 40 was worse than either Holtz or Davie. When ND fans point out that the Irish suffered many more blowout losses under TW than any other coach, it's not hard to understand given the wild inconsistency in the play of his teams.
On a related note, Davie (for all of his incessant moaning about the difficulty of the schedule) had easier schedules than either Holtz or Willingham. Davie played 26 percent of his games against the top 20, 22 percent against the second 20 and 52 percent against teams outside the top 40. He was the only ND coach of the 3 with the luxury of playing more than half his games against fairly weak teams.
For Holtz the distribution was 39-21-40 percent and for Willingham it was 35-24-41 percent. So Holtz had the hardest schedules with Willingham in between the two, but actually closer to Holtz than to Davie.
So what, if anything, does this portend for the upcoming year?
Many of the pre-season magazines will have half or so of ND's schedule in the top 20. It probably won't work out that way. Every year since 1989, three to five teams on ND's schedule have finished in Howell's top 20. The lone exception is 1996 when only two ND opponents finished in the top 20.
The Irish faced four top 20 opponents in each of 2004 and 2003 and five in 2002. Davie faced as many as four only once (1997). In 1998 Davie had the benefit of playing seven teams that finished outside the top 40.
So, assuming ND goes to a bowl, a typical distribution among the three groups would probably be something like 4-3-5 or 4-4-4. Let's assume for the sake of discussion that the distribution is 4-4-4, which is exactly what it was in 2004.
If ND is going to be a good team, it needs to take care of business against the teams outside the top 40 and go 4-0. Right now, I'd say the four best bets to wind up in this group are Navy, BYU, Washington and Syracuse. We'll also include Army in this category as they'll likely fit here as well. ND played three of these teams last year and went 2-1. An inexplicable loss like BYU in 2004 or Syracuse in 2003 absolutely must be avoided.
Against the middle four, 3-1 is the needed mark. I'd peg Stanford, Pittsburgh, Purdue and Michigan State as the most likely entrants into this group. Of course, Purdue might finish in the top 20 and Stanford might fall into the bottom group, but historically those teams are a pretty safe bet to wind up in the middle of the pack. ND played all four of these teams last year, going 2-2. ND needs to avoid a repeat of the epic choke against Pittsburgh.
Against the top four, 2-2 is probably the needed mark here. Those teams might be the bowl opponent, Michigan, USC and Tennessee. Note that these four actually were the top 20 last year (Oregon State, ND's bowl opponent, finished 20th in Howell's ratings) and ND finished 2-2 against these teams.
A 4-0, 3-1 and 2-2 distribution would result in a 9-3 record that would probably hearten most ND fans.
The good news, I suppose, is that the Irish have played well enough against the good teams to have decent or better years. But the Irish haven't been able avoid season-killing losses to weak teams. If ND starts to beat the teams it should, things will look up in a hurry.