Finally, The Old Man and the Hyperbole, Woody Paige, wrote:
Hannibal never endured such a demanding march, or October — road games against the Patriots and the Chargers and a home game with the Saints. Guess what? The final eight games are no bargain. The Broncos do get the Bucs and the Browns at home, and conclude the regular season at SAF at Mile High, as they did last year, against those pesky Chiefs. But they must play at Carolina — ever heard of Cam Newton? — and K.C., Oakland and Baltimore.
Woo. I've been covering the Broncos since 1974, and there hasn't been a schedule this grueling in any season since then — or, certainly, before.
This, of course, settles the issue, since 1974 was a watershed year. It brought us "Jungle Boogie" from Kool and the Gang, The Texas Chainsaw Massacre (warning, nightmares will ensue), and Woody Paige, Denver's own cuddly serial killer of football knowledge.
Reading all of this, one might wonder why Peyton Manning decided to sign with the Denver Broncos in the first place. Didn't he realize the Broncos were entering a world of pain?
Of course he did, and yet, he signed anyway. That's because Peyton Manning knows a little secret that Woody Paige (and his merry band of men) hasn't figure out: strength of schedule is bovine excrement.
That's right. All of this writhing and hand wringing about the Broncos' upcoming schedule might be good mental compost, but at the end of the day, it's still crap. The difficulty of one's starting schedule has nothing to do with how many games a team will win.
How do we know? Rather than use some shoddy stats (the previous year's winning percentage of a team's opponents) and apply it to the future, we can use some simple linear regression. If what the pundits say is true, one would expect some relationship between a team's strength of schedule and its ability to win games.
But in fact, there is no relationship at all--none. We looked at five seasons of data from 2006-2010 and regressed a team's beginning strength of schedule against its winning percentage at the end of that same year. The correlation coefficient was the lowest we've seen in a long time: 0.0285. In fact, it's hard to get a much weaker correlation. In other words, strength of schedule has no relevance other than providing for mental masturbation.
We've been saying this for awhile now, but it's worth repeating yet again: an NFL season (and winning football, for that matter) is too complex to model with any single variable. This is true with respect to quarterbacks, and it's certainly true when it comes to last year's results. There are just too many variables. Coaching, injuries, free agency, the draft, scheme changes, and random luck are just a few.
Don't fall for it. Ignore your cognitive bias. Just because the Bengals and 49ers were good last year, do not make the assumuption that this relationship (them not sucking) will continue into 2012. This is the sort of bias that burned buy-and-hold investors during the last financial crisis. Don't make the same mistake--unless, of course, you enjoy cerebral titillation like Woody Paige.
(Note: Regression resulted in a P value < .05, you sun-drepraved statistical nerds)