The independent news organization of Duke University


A polling problem

<p>Georgia Tech was the preseason favorite to win the ACC Coastal Division but has yet to pick up a win in conference play, an example representative of a national trend in which teams deemed to be preseason powerhouses have fallen far short of expectations.</p>

Georgia Tech was the preseason favorite to win the ACC Coastal Division but has yet to pick up a win in conference play, an example representative of a national trend in which teams deemed to be preseason powerhouses have fallen far short of expectations.

The media—like Jon Snow—still knows nothing.

Okay, we know some things, but the notion of asking anyone—even national and local beat writers who are paid handsomely (just kidding)—to predict a top 25 before even seeing any team take the field is incredibly disturbing and, looking back, highly entertaining.

Here’s a link to the original preseason top 25, and a look at the conference and division winners that each respective conference’s media predicted prior to the season:

-Big 12: TCU

-ACC: Clemson and Georgia Tech

-Pac-12: Oregon and USC

-Big Ten: Ohio State and Wisconsin

-SEC: Auburn (overall), Alabama and Georgia (divisions)

That seems pretty solid, right? Let’s analyze what the media was thinking two months ago.

The Big Ten and Big 12 were fairly straightforward: pick between Ohio State and Michigan State or TCU and Baylor, then just let the rest fall where they may, because there’s really no way any other team in either of those conferences could usurp their current kings.

Then there is the ACC, where the media thinks Clemson will be good, if not very good. It has an inkling of how good a Jameis Winston-less Florida State can be, and it thinks that a Georgia Tech team that absolutely crushed Mississippi State in its bowl game and returns its stud quarterback will dominate the Coastal. Unfortunately, the Yellow Jackets are a hot mess and Duke is okay, which, in the ACC Coastal, pretty much makes you the ‘76 Dolphins.

The SEC media, which is notoriously bad at predictions—they’re 5-of-23 since 1992—really took a leap when it selected Auburn to win it all; the reward was a team that can’t score much and wilts under the bright lights. It also selected Georgia to win the East thinking Nick Chubb is a deity, but the guy can’t play quarterback and running back, and now everyone looks dumb and Florida is good again.

Finally, the Pac-12, which had the hardest job of any media group, in my opinion. The South is loaded with four teams that could presumably win a spot in the playoff, but so much has to break right for every team to win that they were really just shooting in the dark when trying to pick between UCLA, USC, Arizona and Oregon to win the conference.

Think of how much it must burn every Pac-12 writer to know that after countless hours of film research and roster reviews, their predictions were blown to hell by Utah. Swing and a miss.

But as much as I like to have fun with this, it’s an issue that persists in every college sport, especially college football and basketball, which routinely rely on rankings to determine postseason destinations. The biggest issue with the annual AP Preseason Top 25s is that they’re based completely on speculation and actually harm polling for the rest of the season.

Take, for example, the fact that despite its top spot in the current poll, Ohio State is not the best football team in the nation, nor has it played that way. Utah and LSU are the only teams in the top five to have faced another team currently ranked in the top 25 so far and are perfect against ranked opponents. TCU, Baylor and the Buckeyes are a combined 0-0, and yet Ohio State raked in 28 first-place votes last week.

By the logic of polling and the way the system is set up, the Utes have absolutely earned the No. 1 spot. To this point, they’ve defeated Michigan—unranked at the time—Oregon and California. But because the Buckeyes won the title last year and everyone kept talking about how good they would be all offseason, voters gave and continue to give them the top slot, despite the fact that they have two close calls against Northern Illinois and Indiana and haven’t come close to dominating an easy schedule the way Baylor and TCU have.

One of the knocks against college football, and college sports in general, is that there are far too many teams, and far too few of them are actually any good. After all, even a bad game in the NFL can’t come close to replicating the 6-3 double-overtime ‘thriller’ Virginia Tech and Wake Forest produced last year.

The argument is that the truly talented teams not in the SEC or Pac-12 really don’t have to play anyone for most of the regular season, and then we get surprised when Florida State gets thrashed by an embarrassingly-superior Oregon team and TCU has to fume all offseason.

But, to play devil’s advocate, it’s that range that makes college football so exciting. After all, if someone told me they had Temple, Houston, Iowa, Toledo and a James Conner-less Pittsburgh in their preseason top 25 and LSU and Utah in the top five, I’d probably call them a liar or Raven or God or something, because that’s nuts. 

All things considered, the polling system isn’t broken, it’s just hindered by a flaw that can be easily fixed: kill preseason polling and wait until Week 1 to judge. The only people who have any clue as to which teams will be good are the lucky fans who believe every year is their team’s year. If anyone else, even a tried-and-true reporter, tries to tell you what’s up, just point them to this year's and every other preseason poll ever.


Share and discuss “A polling problem” on social media.