Many posters (including me) have spent a lot of time focusing on what the record will be next year. Ultimately, of course, it comes down to wins and losses, but one year is too short of a period, in many cases, to make a meaningful evaluation simply from the record.

A more reliable way is to look at four key statistics: yards per rush, yards per pass, and the defensive numbers on these two statistics. For all of Davies's blather about the kicking game (not that I'm suggesting that anyone really believed him), in order to win consistently you've got to be able to run the ball more effectively than your opponents and pass it more efficiently. That doesn't necessarily mean run it more or pass it more (total yards numbers are very deceiving), but more effectively.

Trying to count on turnovers and special teams play to always win games is a futile strategy in the long run. You need to play well in those areas, not give games away, and hopefully win a game or two a season on them, but ultimately there just aren't enough of these plays to allow you to consistently win games. In any given season a college football team will run and defend roughly 1500 plays on offense and defense combined, but fewer than 200 plays will involve turnovers or be special teams plays (e.g., punt and kickoff returns). So, you may occasionally snatch victory from the jaws of defeat this way, but over the long haul you just can't win consistently unless you can run the ball, stop the run, pass the ball and stop the pass.

A problem with focusing too much on first year records of coaches is that in any one season you may get a lucky win or two or an unlucky loss or two that will make a record look better or worse than it should be. Holtz's 5-6 first year is a fine example of this. But, if you focus on these four statistics you can usually see the seeds of improvement or disaster, as the case may be. The differentials are all reported so that "+" means improvement and "-" means worsening.

Look at N.D. across coaching changes since Ara.

1963 (Devore)
Yards per rush: 3.2
Yards per pass: 5.2
Yards vs. rush: 3.7
Yards vs. pass: 6.6

1964 (Ara)
Yards per rush: 4.2 (+1.0)
Yards per pass: 9.5 (+4.3)
Yards vs. rush: 2.0 (+1.7)
Yards vs. pass: 6.0 (+0.6)

As we can see, Ara's performance was nothing short of spectacular, as one would anticipate from his spectacular career. Ara's improvement also, of course, showed up in the record, which went from 2-7 to 9-1. Now consider the switch from Ara to Devine:

1974 (Ara)
Yards per rush: 4.6
Yards per pass: 7.1
Yards vs. rush: 2.3
Yards vs. pass: 5.0

1975 (Devine)
Yards per rush: 4.2 (-0.4)
Yards per pass: 5.6 (-1.5)
Yards vs. rush: 3.5 (-1.2)
Yards vs. pass: 5.8 (-0.8)

As we can see, in every category N.D. got worse the first year under Devine, which signaled that Devine was not as good of a coach as Ara. The program, however, was functioning at a high level and the falloff was nothing like the horrifying drop in Davie's first year (below), and the decline from 10-2 to 8-3 in the record was probably about indicative of where the team stood. Now to the Devine to Faust handoff.

1980 (Devine)
Yards per rush: 4.2
Yards per pass: 4.9
Yards vs. rush: 2.8
Yards vs. pass: 4.9

1981 (Faust)
Yards per rush: 4.0 (-0.2)
Yards per pass: 6.4 (+1.5)
Yards vs. rush: 3.6 (-0.8)
Yards vs. pass: 5.8 (-0.9)

In Faust's case, three of the four indicators dropped, which was a worrisome sign, along with the obvious falloff in the record, though probably the decline in that year wasn't quite as great as the drop from a 9-2-1 record to 5-6.

Devine's 1980 team was very strong defensively and had a reasonably good running attack, but could not throw the ball with any effectiveness, meaning that by the end of the season they were almost helpless offensively. Consider that their last 5 games were a 3-3 tie with a weak Georgia Tech team, a 7-0 win over Alabama, a 24-10 win at home over a weak Air Force team, a 20-3 loss to U.S.C. and a 17-10 loss to Georgia in the Sugar Bowl. Even with the 24 points against Air Force, they averaged less than 10 ppg over this stretch, meaning that the only hope they had of winning games against good teams was to get into a defensive struggle. Faust's 1981 team was also a little snakebit (though some of this was Faust's poor situational game coaching) losing by 1 at Purdue, by 6 at home to F.S.U., by 7 at home to U.S.C. and by 3 on the road at Penn St. None of the games that they won were close; every victory was by at least 13 points. So, if N.D. had been 2-2 in those close games, they would've been 7-4, which probably would've been more reflective of the team's true state, but Faust clearly had the team pointed in the wrong direction.

Now for the Faust to Holtz handoff:

1985 (Faust)
Yards per rush: 3.7
Yards per pass: 6.8
Yards vs. rush: 3.5
Yards vs. pass: 6.7

1986 (Holtz)
Yards per rush: 3.8 (+0.1)
Yards per pass: 8.4 (+1.6)
Yards vs. rush: 0.4 (+0.4)
Yards vs. pass: 6.9 (-0.2)

Note that Holtz had three of the four indicators going in the right direction and the falloff in pass defense was very small. This indicated that even though the record remained constant that Holtz had the team pointed in the right direction, an observation confirmed by the team's obviously improved play and their 1-5 record games decided by 5 points or fewer. Had they gone 3-3 in those games they would've been a 7-4 team, which probably would've been more reflective of their actual performance.

Now the Holtz to Davie handoff:

1996 (Holtz)
Yards per rush: 5.2
Yards per pass: 8.5
Yards vs. rush: 3.1
Yards vs. pass: 5.7

1997 (Davie)
Yards per rush: 4.1 (-1.1)
Yards per pass: 6.8 (-1.7)
Yards vs. rush: 4.5 (-1.4)
Yards vs. pass: 6.4 (-0.7)

The dropoffs in all four areas (and large dropoffs at that) were certain indicators that disaster had arrived, and it had. Davie very nearly managed to duplicate Ara's impact, but in the other direction. But really, the drop from an 8-3 to a 7-6 record considerably understated Davie's negative impact. In truth, Holtz's 1996 team was probably better than its 8-3 record and Davie's 1997 7-6 team was worse than its record.

Let's consider recent coaching changes at a couple of other schools. Stoops took over Oklahoma in 1999 and went 7-5 his first year, an improvement from the 5-6 the previous year. Here are how the indicators stack up:

Oklahoma

1998
Yards per rush: 3.8
Yards per pass: 6.7
Yards vs. rush: 3.1
Yards vs. pass: 7.1

1999
Yards per rush: 3.9 (+0.1)
Yards per pass: 6.9 (+0.2)
Yards vs. rush: 3.1 (0.0)
Yards vs. pass: 6.9 (+0.2)

Although the improvement was slight, Stoops had OU pointed in the right direction even without a tremendous improvement in the record -- that came the next year.

Now, how about U.S.C., which switched from the much-maligned Paul Hackett to Pete Carroll and saw its record improve from 5-7 to 6-6, sparking optimism among the Trojan faithful.

2000 (Hackett)
Yards per rush: 4.1
Yards per pass: 7.1
Yards vs. rush: 3.1
Yards vs. pass: 7.3

2001 (Carroll)
Yards per rush: 2.6 (-1.5)
Yards per pass: 7.2 (+0.1)
Yards vs. rush: 3.9 (-0.8)
Yards vs. pass: 6.2 (+1.1)

Hmmm. Carroll's record is mixed, but a couple of things would really worry me if I were a Trojan fan. One is that both of the rushing numbers (offense and defense) took gigantic slides in the wrong direction. Now, I know that U.S.C.'s best back was hurt for much of the year, but even when he wasn't hurt, U.S.C.'s rushing numbers were anemic. The passing game stayed pretty good (though 7.2 per pass is only slightly above average; Stanford averaged 8.3 last year and Florida led the nation at 9.1) and the pass defense clearly improved.

So what accounted for U.S.C.'s slight improvement in the record? Well, consider our friend turnovers. U.S.C. in Hackett's last year was -19 on the turnover differential and in Carroll's first year was +16. Now, some of this may well be coaching, but some of it, frankly, is luck. In general, every +5 on turnover differential gives a team another victory. Consider that N.D.'s 9-2 2000 team was +14 on turnovers and its 2001 5-6 team was only +3. The 2000 team probably would've been in the 6-5 or 7-4 range had it been close to even on turnovers. In truth, the 2000 team really wasn't much better than the 2001 team. It's simply impossible, over the long haul, to generate consistently huge turnover margins. Consider that Holtz, N.D.'s best coach ever on turnovers, averaged +6 for his career; Ara and Devine both averaged +1. U.S.C., had it been close to even on turnovers, could well have been a 3-8 or 4-7 team, which is the fate usually reserved for teams that average only two and a half yards per carry.

Consider what happened to U.S.C. in the bowl game with Utah. This was not a particularly good Utah team. They were 7-4 overall and 4-3 in the Mountain West conference and the Utes' big win was probably a 28-26 road victory over Indiana. Even with the win over U.S.C., Howell ranked Utah only 30th and they faced one of the easiest schedules of any team in a bowl game. Sagarin had Utah at 36th and ranked the Utes' schedule as the 62nd most difficult. The Utes allowed 3.5 yards per carry during the regular season, which is a reasonably good figure, but they weren't exactly playing the strongest running teams in the nation. Nevertheless, U.S.C. managed just one yard on the ground in 25 carries (they were actually negative 50 or so at the half). Even winning the turnover battle 2-0, U.S.C. still lost 10-6. So, while it seems to be a trendy projection that U.S.C. is on the road to recovery under Carroll, I am far from being convinced.