College Football: The Analytics Suggest Tamping Down the Hype on Houston

Advanced metrics offer a contrarian view on one of the offseason's most buzz-worthy teams, the Houston Cougars.

It’s the unofficial start to the college football season that every causal or serious fan celebrates: the first preseason magazine landing on the shelves of your local grocery or drug store.

Athlon Sports (late May) and Phil Steele (late June) provide bookends to the preseason publication releases. By mid-summer podcasts, message boards, the Twittersphere, and national news sites are abuzz with preseason favorites, value picks, and busts.

In recent years, the offseason has seen the rise of analysis by the advance metrics crowd, which does yeomen’s work in power rating all 128 teams. SB Nation’s Bill Connelly, Brian Fremeau, McIllece Sports, and Ed Feng, among others, publish numbers-based preseason rankings that often provide an alternative viewpoint to team ratings based more on narrative, traditional counting statistics, and the eyeball test.

For those assessing Vegas’ recently released season win totals, games of the year spreads, and futures odds, a helpful exercise is seeking disagreement between the projections of the preseason magazines and numbers-based systems. Discrepancies between the magazines (who I will affectionately call The Gurus) and the metrics experts (who I will even more affectionately call The Geeks) can tell us a lot about overvalued and undervalued teams heading into the season.

Over a few weeks, I conducted a comparative analysis using four preseason college football publications -- Athlon Sports, Lindy’s, Phil Steele, and The Sporting News -- to develop a consensus top 25 among The Gurus. I then compared those composite rankings with four advanced metrics systems released by The Geeks (Fremeau’s FEI, ESPN’s FPI, Connelly’s S&P+, and Feng’s The Power Rank).

Here is a closer look at one trendy team, the Houston Cougars, where a chasm exists between the magazines’ and metrics’ rankings.

Hype Around Houston

In coach Tom Herman’s first year, the Cougars went on a magical 13-1 run, which included a Peach Bowl win over touchdown-favorite Florida State. numberFire’s own nERD efficiency rankings rated Houston (16.15) as the 15th-best team in the country at the end of 2015, tops among all Group of 5 schools.

Despite his name being linked to Power 5 job openings in the offseason, Herman is back. So is his prolific quarterback, Greg Ward Jr., who has established himself as a Heisman hopeful. Calling his 2015 campaign a breakout would do it little justice. Ward racked up 38 total touchdowns, many of them in spectacular fashion.

The defense will once be led by the front seven, which returns experienced playmakers and could be even better against the run this year.

Throw in some blockbuster recruiting wins this winter and spring (remember Ed Oliver), and it’s easy to see why The Gurus have collectively ranked Houston, on average, comfortably inside the top 25. Optimism surrounding Houston is at an all-time -- and rightfully so.

But recency bias can sometimes skew preseason rankings, not to mention Vegas betting lines. That’s one of the primary reasons why the metrics are much more tepid on Houston, giving them an average national rank of 41.75. The 24-spot difference is largest among the top 25 teams analyzed and second only to UTEP in the entire 128-team FBS.

Team Metrics Avg. Rank Mags Avg. Rank Difference
Houston 41.75 18 -23.75
Iowa 35.25 18 -17.25
Michigan St 25 15.75 -9.25

Sustained Success

There are a couple of factors at play. First, the metrics generally lean on sustained success -- both from an efficiency and talent standpoint. The recent boost from Herman has been stunning, but Houston must prove that the way they won in 2015 is achievable over a longer period of time.

Using ESPN's efficiency ratings, FEI, and S&P+ to get annual averages, Houston has never been ranked in the top 20 nationally and only twice -- 2011 and 2015 -- has it cracked an average top 30 ranking. In between those two memorable seasons, the Cougars struggled to maintain even mediocrity, averaging 87th, 44th, and 76th, respectively.

Houston five-year efficiency

Even if we agree that a program’s short-term track record matters less when it brings in a new, dynamic coach, talent is still king. And Houston has simply not kept pace with other teams inside The Guru’s top 25, according the 247Sports composite recruiting rankings. A look at the two teams ranked ahead of Tom Herman’s squad in the average magazine rankings -- Michigan State and USC -- and the two teams ranked immediately behind Houston -- Iowa and TCU -- reveals a canyon between the upstart Cougars and their more established peers.

On only two occasions in the last year years did Houston’s recruiting class top Iowa’s (by two spots in 2013 and by 14 spots in 2016), and the Cougars failed to best TCU, USC, or Michigan State in a single class.

247 Sports Composite Recruiting Ranks


Bill Connelly also warns of some regression in luck that may be heading Houston’s way, despite the effectiveness of the defense's front seven. In his terrific Houston season preview, Connelly summarizes, “This was a flawed team experiencing a wonderful season, and while they were incredibly fun to watch, they may not have been using a replicable formula for victory.”

That’s because recovering 65 percent of fumbles (second in the nation) and opponents picking off only 15 percent of defended passes (7 percentage points below the national average) are outcomes not likely repeated. Phil Steele -- who expressed more reservations about Houston than the other magazines, ranking them 30th -- highlighted that Houston had a plus-21 turnover margin, good for second in the nation. According to Steele, since 1991, 77 percent teams with a plus-11 turnover margin or more were weaker or the same the following year. In the last four years, using a plus-14 turnover margin, 87 percent of teams regressed or remained stable the next season.

Connelly also notes that Houston prevailed in close games -- at Louisville and hosting Cincinnati and Memphis -- that, based on the statistics, they should have been expected to lose.

Combined, these factors kept Houston’s overall efficiency ratings depressed last year. While the AP and Coaches polls had the Cougars ranked 8th to finish the year, the average ranking from The Geeks was just 27th.


Houston’s schedule will do the Cougars no favors. Even if Houston remains the class of the American Athletic Conference, their non-conference schedule may force some regression, on the stats sheet and in the win column.

Date Opponent S&P+ Win Probability FPI Win Probability
Sept. 3 Oklahoma 21% 10%
Sept. 15 at Cincinnati 49% 52%
Oct. 8 at Navy 46% 62%
Nov. 17 Louisville 35% 31%
Nov. 25 at Memphis 52% 66%

The South Point in Vegas lined Houston’s season win total at 9.5, with juice on the over. FPI projects Houston to win 8.4 games but gives them only a 0.1 percent chance to go undefeated. S&P+ is even more pessimistic, giving the Cougars a win projection of 7.7.

Both S&P+ and FPI projects Houston to lose their neutral site game against 10-point favorite Oklahoma to start the year and to Louisville at home in mid-November. The Cougars will also get tests in conference play on the road against Cincinnati, Navy, and Memphis.

Tom Herman and his staff deserve credit for getting the Houston program on a path to becoming elite, but judging by two key preseason metrics -- efficiency and the accumulation of talent over time -- Houston is more of a top-40 than a top-20 team. And some inevitable luck regression and difficult non-conference and road slates won’t help.

In college football, the margin for error is razor thin, especially for non-Power 5 schools, so if Houston takes even a slight step back in 2016, it could mean the difference between a New Year’s Six bowl game and one played in the third week of December.