clock menu more-arrow no yes

Filed under:

3 overrated college football teams for 2018 by preseason analytics

New, 2 comments

Preseason polls are more accurate and useful than you might think, but there are always some misses.

NCAA Football: Orange Bowl-Wisconsin vs Miami Jasen Vinlove-USA TODAY Sports

It’s the time for optimism. The humid air of August brings hope for college football fans, as they dream of a national championship or playoff spot.

Then the air turns crisp with the coming of fall. In October, that college football playoff contender drops a conference game. Then they drop another, making a playoff berth all but impossible.

We start to wonder… maybe that team was overrated this preseason.

Preseason hopes can be irrational. It’s like thinking Connor McGregor starts meditating for anger management.

Analytics can help you avoid this trap.

How to predict overrated teams

To gauge expectation, we use the preseason AP poll. This might seem like too easy a target. The poll is amusement until the season starts, right?

However, the preseason AP poll is a surprising predictor of team strength. Consider the following:

  • Since 2005, the higher ranked team has won 59.9% of bowl games with no access to information from the regular season.
  • In the college football playoff era (2014-2017), half of the top 5 teams have made the playoff.

To overcome this challenge of predicting overrated teams, I’ll use preseason analytics. My regression model at The Power Rank uses team strength the past 4 years, turnovers and returning starters to predict 2018.

These analytics rely on linear regression, the same data science method used by Bill Connelly in the S&P+ preseason numbers. In a recent article, I explained the basics behind this model and why it excels at picking off overrated teams.

I’ve used these numbers to write this article on overrated teams in 2015, 2016 and 2017. I’ve been correct on 7 of 9 teams. The other two teams were Penn State and Oklahoma in 2017, who finished the season ranked lower in the final than preseason AP poll. However, both teams were better than I expected, so I call these two predictions a push.

Let’s look at 3 overrated teams for 2018.

1. Wisconsin

4th AP, 8th The Power Rank, 12th S&P+

The Badgers had an exceptional season in 2017. Beyond their 13-1 record, they excelled on both sides of the ball.

In my adjusted yards per play, the offense ranked 19th. They bring back every starter on an offensive line that probably features multiple NFL players. The defense ranked 12th in my adjusted yards per play.

In addition, Wisconsin plays in the Big Ten West, by far the weaker division in the conference. The AP poll has 4 teams from the Big Ten East in the top 15, while only Wisconsin joins them from the West.

However, there are concerns the Badgers can’t live up to their preseason rank of 4th. These start at the quarterback position, as they bring back Alex Hornibrook. While the junior can make throws, he’s not type of player you trust to win big games.

Wisconsin can win games despite Hornibrook because of their defense. However, this unit lost 4 of their top 6 players in the secondary. In addition, they lost 3 starters in the front 7, which suggests a decline.

Moreover, the schedule doesn’t favor Wisconsin, as they make road trips to Michigan and Penn State in inter-division games. They also travel to their 3 best opponents from the West (Iowa, Northwestern, Purdue).

To justify 4th in the preseason AP poll, Wisconsin has to make the college football playoff. My numbers give a 30% chance for this to happen, so I’ll take the 70% chance they’re overrated.

2. Miami

8th AP, 17th The Power Rank, 13th S&P+

The Hurricanes got everyone excited with a 10-0 start to the 2017 season. In his second year, Mark Richt had his program together, as they thumped Notre Dame 41-8.

Then Miami finished the season 0-3, which took some of the luster off a strong season. This didn’t prevent the AP pollsters from putting the Hurricanes 8th to start 2018.

To make the overrated argument, let’s look at the quarterback position. Malik Rosier had his ups and downs in 2017, but his 54% completion percentage doesn’t inspire confidence (60% college football average). He will have weapons at the skill positions, especially if NFL caliber receiver Ahmmon Richards can stay healthy.

However, there are concerns on the defense. Coordinator Manny Diaz had this unit at 11th in my adjusted yards per play in 2017. But they lose an astounding 42 tackles for loss among 4 players on the defensive line. No matter how well Miami recruits, it’s difficult to replace that type of production.

My model predicts 8.2 wins for Miami. If they fall on the low side of that number, Miami won’t be considered a top 10 team by season’s end.

3. West Virginia

17th AP, 43rd The Power Rank, 43rd S&P+

I almost wrote about the Mountaineers in last year’s article on overrated teams. They were 22nd in the preseason AP poll, well below the 63rd in my preseason model. The only reason to think they could crack the top 25 was transfer QB Will Grier.

It almost seemed too easy to pick West Virginia as overrated. In addition, Bill Connelly hammered this point home in his preview, so I wrote about other teams.

Through ten games, West Virginia made an overrated call look stupid, as they posted a 7-3 record. Will Grier led the offense to the 10th best pass offense by my adjusted yards per play.

Then Grier got hurt against Texas, and the wheels fell off in a 0-3 finish to the season. Still, the promise of another season with Grier has West Virginia ranked 17th in the preseason AP poll.

This is too high. After their 7-3 start in 2017, they ranked 23rd as a team when combining offense and defense by my adjusted yards per play. Even during their best stretch, West Virginia didn’t crack the top 20.

The offense under Grier should be fine. But the defense ended last season 68th in my adjusted yards per play, and they lose enough players on all three position groups to think they don’t improve.

In addition, the schedule is difficult. West Virginia will be the underdog in all 5 of their road games against Power 5 opponents (four Big 12 teams plus North Carolina State). Besides Kansas and Youngstown State, there are no mostly certain wins.

My analytics project 6.1 wins, which makes it difficult to think of West Virginia as a top 20 team.


The 2018 Preseason Market Report

I have a little confession to make.

I wasn’t only using the preseason regression model to make these predictions. This off season, I’ve developed a method to rank teams through market win totals.

These market rankings agree on these 3 overrated teams:

  • Wisconsin: 4th AP, 8th market
  • Miami: 8th AP, 14th market
  • West Virginia: 17th AP, 26th market

These market rankings provide a wisdom of crowds alternative to preseason analytics. In The 2018 Preseason Football Market report, I post the rank and rating of the top 25 college football teams.

To get this free report, check out thepowerrank.com.