clock menu more-arrow no yes

Filed under:

Advanced stats hate Iowa, UNC and Washington State, but don't let that bother you

New, comments

The S&P+ numbers (and other computer rankings) aren't particularly fond of three of college football's most enjoyable stories. Here's why (and why it doesn't matter).

Jeffrey Becker-USA TODAY Sports

Numbers can be killjoys. They aren't impressed with dramatic finishes. "I see it as a tossup game," they say, not really caring who won. They don't get starry-eyed by gaudy performances against mediocre teams. And they aren't inclined to like a team simply because it is doing well in the win column.

When new sets of S&P+ and F/+ rankings come out each week, I get stock reactions -- "faulty metric" and "that's clearly wrong; you have Team A ahead of Team B, but Team B beat Team A," and the like.

But as each season progresses, I pick up friends and enemies based simply on who the numbers like or don't like. Clemson and Michigan fans haven't had much problem with S&P+, and Oklahoma fans are coming around to the concept. But three fan bases have decided my metrics and, therefore, my own brain are faulty. (Who am I to say they're wrong?)

  • Iowa is 10-0 for the first time ever and in the AP top six for the first time since 2002. If the Hawkeyes win out (beating Purdue, Nebraska and the Big Ten East champion), they are without a doubt in the College Football Playoff. Considering the "Eh, they'll be decent, but ..." predictions from the preseason, this is a wonderful story.
  • North Carolina's only loss came via red zone failure in the season opener against South Carolina. The Tar Heels are smoking hot at 9-1 and on a collision course to face Clemson for the ACC title. Their offense has scored 48 or more in five of 10 games, and their defense has allowed 21 or fewer in eight. Average score against ACC foes: UNC 44, Opponent 22.
  • Washington State has won seven of nine games since a Week 1 dud, with losses to only Cal (when the Golden Bears were smoking hot) and Stanford (ditto) by a combined eight points. They have won three straight conference road games (against Oregon, Arizona and UCLA, no less), and if a field goal had drifted against Stanford, the Cougars would be in the Pac-12 North lead.

These are great stories. Three fan bases who haven't had a ton of reason to be jazzed up are jazzed up. This is why you talk yourself into your team every preseason. There are just enough random successes in college football to think you might be one. Iowa City and Chapel Hill were scenes of joy on Saturday, and Pullman would have been if the Cougs hadn't been beating UCLA in the Rose Bowl instead.

Iowa, UNC and Washington State are sixth, 12th and 24th in the polls, respectively, and are 28th, 30th and 71st in S&P+.

SIGN UP TO GET THIS IN YOUR INBOX

Get one roundup of college football stories, rumors, game breakdowns, and Jim Harbaugh oddity in your inbox every morning.

Faulty metrics, right? If you want to believe the numbers have been programmed to hate your team, there's nothing I can say to stop you. But out of curiosity, I wanted to look into each and figure out why the numbers hate them so much.

Iowa

Iowa's the easiest to figure out: the Hawkeyes have mastered the art of close wins. (That, by the way, could be an art, and could be sheer randomness. Just ask Auburn.) But if we believe the rankings of the teams they've played, it gets easy to believe Iowa might be more of a No. 28 than a No. 5.

  • No. 41 Minnesota: Iowa led by five with two minutes left before a touchdown sealed the win, then won by five.
  • No. 42 Pitt: Iowa won by a last-second field goal at home.
  • No. 44 Illinois: Iowa led by three points with four minutes left before turnovers and field goals increased that margin to nine.
  • No. 65 Indiana: Iowa led by one with 10 minutes left before pulling away.
  • No. 66 Iowa State: A tie in the fourth quarter before the Hawkeyes pulled away late.

Combine that with the fact that Iowa's most impressive performance (against Northwestern) is less shiny thanks to the Wildcats' statistical fade (two duds followed by three decent-at-best performances), and you've got a profile of a team that is good and plays some of its best football late. It's also a team that's only actually played like a top-10 team maybe once or twice.

North Carolina

Because of scoring margins, UNC is harder to explain. How is a team that has killed all comers in the ACC barely cracking the top 30?

Well, for one thing, they haven't killed all comers. As incredible as they looked in crushing Wake Forest, Duke and Miami (combined: 175-66), they beat No. 70 Georgia Tech by seven, No. 42 Pitt by seven and No. 87 Virginia by 13. Not games that scream, "Elite!" Plus, their raw numbers have been boosted by knocking around two FCS opponents and not a single team that ranks better than 42nd.

And ... well ... losing to South Carolina is pretty awful. Even though it happened 10 weeks ago, it still counts.

UNC is rising. The Heels were 46th four weeks ago. And if the last two murderous performances are the new standard, S&P+ will say, "OK, fine, you're great," in the coming weeks.

Washington State

Washington State's in the same boat. First, that Week 1 dud came against Portland State. And of the Cougars' seven wins, four have come by seven or fewer points, and five have come against teams ranked 69th or worse.

And the Cougars' two most impressive wins were less impressive in the box score. Oregon gained 7.2 yards per play to Wazzu's 6.1 in WSU's overtime win. Saturday, UCLA gained 6.2 to WSU's 5.3, but the Cougs recovered all three of the game's fumbles; a muffed UCLA punt set WSU up on a short field (and resulted in a TD), and WSU's Luke Falk recovered his own fumble after a sack on what eventually became another TD drive. If fumbles luck flips, WSU probably loses.

Both UNC's and WSU's ratings are lower than I would have anticipated. I would have expected the Cougars to crack the top 60 and the Heels the top 20 by now. But considering UNC has yet to play a top-40 opponent and WSU was taken to the wire (or beaten) by three opponents that rank 79th or worse (Arizona, Rutgers, Portland State), their ratings are not that surprising.

Also: who cares?

Even if the numbers are right about all three teams, Wazzu is projected to finish 8-4, amazing to consider when thinking back to the first-week loss to PSU. North Carolina is projected to finish 10-2 (with a projected loss to rival NC State, which has both a higher rating and, yes, more losses than the Heels) with a spot in the ACC title game. Think back to the SC loss and the preseason "This is the year they breakthrough? Yeah, right" scoffing, and tell me that's not awesome.

Oh yeah, and Iowa is projected to pummel Purdue and win a tight one over Nebraska. The Hawkeyes aren't going to be given a favorable shot in the Big Ten title game, especially if the opponent is Ohio State. But who cares? A 12-1 record would be an amazing turnaround. And the tight finishes have given this even more of a storybook run.

Yeah, so the numbers hate your team. So what?

More college football for you

Facebook
Twitter

* * *

SB Nation presents: Which teams should be in the Playoff right now?