Every year, college football fans view opening weekend like a holiday. This year is no exception, as fans across the country eagerly anticipate the beginning of the 2018 – 2019 college football season after a long (not to mention turbulent) offseason. If last weekend’s “Week Zero” games were an appetizer, this weekend will be a true feast of college football action.

I’m very much looking forward to conducting the MaxDiff Poll this season because (in my admittedly biased opinion) this is one of the most valid college football fan polls with participation open to the public on the Internet. Each week this poll’s primary objective is to answer one simple question: Which teams do the average college football fan feel are actually the best in the nation at the present time? This poll doesn’t claim to be predictive of future games or anything like that – it simply aims to quantify what fans believe as of right now. If you are interested, I’ve provided some more info at the bottom of this article about what this poll has over other public college football fan polls on the Internet.

I will be conducting the MaxDiff Poll each week of the 2018 – 2019 season similarly to last season. For the next thirteen weeks, this poll will launch every Monday at 10 AM Eastern / 9 AM Central. If you would like to participate, simply check Reddit CFB each Monday for the poll link. You can also receive regular updates letting you know when the poll opens by following Fanjuicer on Twitter or joining our online panel. Results are posted to FanJuicer.com every Tuesday or Wednesday morning before 10 AM Eastern / 9 AM Central.

This Week’s Poll Specs

When was this fielded?: August 21-23, 2018
Number of respondents: 1,175
Research technique: Online quantitative survey using the MaxDiff Market Research Technique.

Interpreting the Scores: The MaxDiff scores for each team are indexed at 100. For example, Michigan State’s MaxDiff score of 129 means that team was ranked 29% higher than the average team in the exercise. It is important to note that these scores should not be interpreted as “votes.” See our explanation of MaxDiff if you would like more information.

FanJuicer’s “Big Six” Takeaways From This Week’s Poll

  1. Fans agree with The AP Top 10 (for now).The MaxDiff Poll’s Top Ten this week aligned perfectly with the AP’s if you just swap the order of Miami and Penn State.
  2. The usual suspects own Tier I.Each of the three members of Tier I (Alabama, Clemson, and Georgia) made last year’s Playoff. Two of those three have made The Playoff the last three seasons. One of those three has made it the last four. Yawn.
  3. Fans think OKLAHOMA STATE is the most UNDERRATED team in the AP.The Cowboys barely registered in the AP after clocking in at #29. By comparison, fans in the MaxDiff poll put Mike Gundy and company in the top 20 at #19! I triple checked this one to make sure there was not an error in my calculations. There wasn’t; fans simply have a much more optimistic view of Oklahoma State’s chances than the AP voters (for now). This is the biggest discrepancy I’ve encountered thus far in any of the weekly MaxDiff Polls against The AP. It’s worth noting that it isn’t just this particular week’s MaxDiff poll in which fans voiced a more favorable view of Oklahoma State than the media establishment. In FanJuicer’s Big 12s Preseason Power Ranking Poll that I conducted a couple of months ago, fans put Oklahoma State in a tier above Texas, who the AP ranked #23 this week! This leads me to my next point. . .
  4. Fans think TEXAS is the most OVERRATED team in the AP.If Texas and Oklahoma State were to play tomorrow on a neutral field, who do you think would win outright? This week’s MaxDiff Poll suggests the media would choose the Longhorns while the average college football fan would lean toward the Cowboys.
  5. The winner of this weekend’s Michigan vs. Notre Dame clash will enter Tier III next week. The winner of Washington vs. Auburn will enter Tier II.This weekend features two heavyweight clashes that will have a major impact on this season’s Playoff picture. The winners of these two games will drastically increase their chances of making The Playoff.
  6. Who will be this year’s top G5 Team (besides UCF)?Writers say Boise State. Fans say Boise State or Florida Atlantic. Fans ranked Boise State (#26) and FAU (#27) similarly in our poll, but in the AP there was a more sizable gap between the Broncos (#22) and Owls (#30).

The MaxDiff Poll Versus the Spread?

Last season, I noticed a peculiar trend in the MaxDiff poll that I didn’t mention in any of the weekly results.

  • A majority of the time, teams who fans felt were OVERRATED in the AP tended to win versus the spread the following week.
  • Teams fans felt were UNDERRATED in the AP tended to lose against spread!

This pattern was particularly hot in the first half of last season. It could very well just be random noise. However, just for fun, I’d like to follow it this season. Using this pattern, here’s this week’s picks against the spread:

  • NOTRE DAME (-1.5) over MICHIGAN
  • MIAMI (-3.5) over LSU
  • TEXAS (-13.5) over Maryland
  • BOISE STATE (-10) over Troy
  • OKLAHOMA (-21) over Florida Atlantic
  • UTAH (-29) over Weber State
  • OKLAHOMA STATE (currently no spread) over Missouri State

Keep in mind that this pattern last season may just have been random noise, so I wouldn’t bank your Pick Em Pool on this just yet.

How Do Fans Feel About The Current Playoff Format?

Most fans favor the current four-team format to the old BCS system, but would like to expand The Playoff to include more teams.

This research suggests an eight-team playoff system is the sweet spot for most fans.

More About The MaxDiff Poll’s Validity

Here’s what the MaxDiff poll has over many of the other public Internet college football fan polls out there:

  • A sample that is reasonably representative of college football fans nationwide (at least in terms of the teams participants follow). In many well-known polls, such as Gallup’s Presidential Approval Poll, demographic variables such as age and gender have been shown to have a strong impact on how people are likely feel. It’s easy to see the importance of ensuring your sample represents your defined population in terms of the particular variables that are likely to impact opinions on the topic of study. However, unlike the Presidential Approval Poll, this poll is a ranking of football teams, and in this context, I believe it reasonable to think one variable – the team a voter calls “his” team – is likely to impact his perceptions far more than any demographic variable such as age or gender. In other words, a 40 year-old female Michigan Wolverines fan will vote more like a 24 year-old male Wolverines fan than a 40 year-old female Buckeyes fan. Fans who participate in the weekly poll indicate their team affiliation. I then compare the sample’s composition of fans to the best known estimates of fan base sizes such as Nate Silver’s 2011 estimates and others to see if things match up. They usually do, but if something is off from the usual, I weight the data to correct for any discrepancies. Yes, I’m aware this poll’s sample is not perfect, and it isn’t perfectly representative. However, neither is the ESPN fan poll or any of the others for that matter. This poll’s sample is of as high a quality as you can get without spending hundreds of thousands of dollars.
  • A relatively large sample size. The poll usually aims for at least 1,200 weekly respondents. Ever wonder why those polls you see on TV have about 1,200 or so participants? That’s usually enough.
  • Highly engaged participants. The vast majority of the fans who participate in this research are engaged and knowledgeable about college football and usually give thoughtful, informed opinions. Not only do they self-report a high level of knowledge about college football, but they demonstrate it in the voting exercise. A reasonable degree of logical consistency in the MaxDiff exercise allows me to identify and throw out the small minority of participants who don’t have informed enough opinions on college football to be meaningful.
  • MaxDiff is just a better technique for ranking 25+ items. Most people can’t handle a traditional ranking task of more than six or seven items. Not only is it difficult, it’s extremely boring. If I gave you a list of 25 movies and asked you to put a “1” beside your favorite movie, a “2” beside your second favorite, and so on, that would probably be a time-consuming task and it also probably wouldn’t be too much fun. Not only that, when you finished your rankings, would you ask yourself if you would rather watch the movie you ranked 21st more than the one you ranked 17th? The MaxDiff Technique gets around this and forces voters to make these critical contrasts.
  • In most fan polls, voters start with that week’s AP or CFP poll and just move a few teams around. The format of the MaxDiff Poll discourages this to an extent, but doesn’t completely prevent it.
  • The MaxDiff Poll makes Tiers, which give a little more insight into how fans feel about the teams.
  • This poll is done by a professional.