Friday, July 20, 2007

Leaving Lollipop Lane For The Mean Streets

Although he never said it explicitly, the argument underlying Phil Steele's positive take on the 2007 NC State football team was plenty apparent: they'll regress back to the mean. Bad luck becomes average luck and there you go: instant improvement.

SMQ points out that Wake Forest and Maryland are due for some regression themselves--but this ain't the good kind. NC State bumbled its way to a 2-5 record in close games while Wake and Maryland were a combined 12-0 in the tight ones. SMQ examines the yardage, first down, and turnover differentials in each team's close games, and the results are enlightening.

Wake Forest: out-gained in five of six, out-first-downed in all six. Record: 6-0.
Maryland: out-gained in all six, out-first-downed in four of six. Record: 6-0.
NC State: out-gained in three of seven, out-first-downed in three of seven. Record: 2-5

In baseball, you can often find an over- or underachiever by looking at records in one-run games. Since those are toss ups (regardless of whatever talent gap exists between the two competing teams), teams that are significantly above or below .500 in those contests tend to be teams with achievements that don't accurately reflect their performance. It's why run differential (or point differential, in the case of football) is generally more telling than won-lost record--it allows you to wade through this kind of luck.

Back in 2005, the Nationals went on a huge first half run that vaulted them into first place. But they were never legitimate contenders: they'd managed a 16-3 record in one-run games during the hot streak, and despite their shiny overall record, they'd allowed more runs on the season than they'd scored. When the good luck vanished in the second half, so did the Nats' winning record. When it was all over, Washington was what its run differential had said it was all along: a .500 team.

I liken the 2006 versions of Wake and Maryland to that Nationals team. It's just that the Deacs and the Terps didn't have enough time to expose themselves.

SMQ on Wake Forest:

I watched Wake Forest last year, I’m looking at its performance in black and white, and I still refuse to believe that team was any good, and even more adamantly refuse to believe it has any hope of beginning to think about threatening a repeat. The collaboration of circumstance, turnovers and clutch kicking in that championship run is not sustainable. In fact, Wake is a textbook first-to-worst candidate regardless of personnel changes (which also happens to work against them, minus four of their best defensive starters) because the Deacs’ every-down play was still among the worst in the league, as it’s always been.

On Maryland:

There’s struggling with Florida International, and there’s beating Florida International with an interception inside your own ten with a few seconds to play, but the two other visits from the Sunshine State really stand out here: Florida State outgained Maryland by 235 yards, more than doubled up the Terps in total offense, held them to 37 yards the entire second half, and somehow, incredibly, with but a single turnover – on its first possession of the game at that – wound up losing on a missed field goal that would have tied with a few seconds left; and Miami forced UMD into seven three-and-outs on its last eight possessions, but gave up the two early bombs I linked to Wednesday to Darrius Heyward-Bey, two plays that accounted for nearly two-thirds of Maryland’s entire offensive production, and couldn’t overcome those lapses. Virginia blew a 20-7 lead in part by allowing a fourth quarter interception return for a touchdown and still could have tied on a late two-point conversion (the pass failed, obviously); the Terps finished 88th in total offense, 84th in total defense, and just a couple slots outside the final polls.

I don't understand why these sentiments aren't more prevalent as we get closer to the season. Okay, sure I do: people worry less about how a team reached its record than about what that record is. Wake won 11 games last year. Maryland won 9. Both return a reasonable amount of talent, therefore both will be successful in 2007. That's a natural and oft-repeated conclusion.

CFN's Wake Forest preview makes note of the Deacs' good fortune in 2006, yet in predicting that a "down" year is most likely (or at least that a modest record would be a success), CFN reasons that this is because "everyone's going to take them seriously this year." How insightful. They had the answer, but rather than treating 2006 as an indictment, they wondered rhetorically if Wake could catch all the breaks again.

The numbers presented by SMQ speak for themselves, but I want to make a couple of additional notes. Below is each school's national rank in total offense, scoring offense, total defense, and scoring defense in 2006.
       TotO  ScoO  TotD  ScoD
Wake 96 78 45 12
UMD 88 74 84 51
NCSU 97 101 36 50
It's tempting to credit Maryland and Wake Forest for their perceived efficiency, but knowing what we already know about the teams, that's credit they don't deserve. Particularly interesting is the scoring defense column; both Wake and UMd ranked 30+ slots higher in scoring d than in total d. That was enough, in Maryland's case, to make its scoring defense on par with NC State's despite a fifty-slot gap in their respective total defense ranks (Maryland allowed 43 more yards per game).

Finally, there's SDPI: teams are ranked based on the points they score and allow relative to the league average and standard deviation. SDPI suggests Wake Forest was only the 5th best team in the ACC last season; Maryland rates 9th.

One of these two teams is going to finish in the bottom third of the Atlantic Division. Sportswriters will wonder where it went wrong. What happened to all of that promise? Nothing--it wasn't there to begin with.