NCAAF teams
Seth Walder, ESPN Analytics 7y

Where FPI and Vegas disagree on college football win totals

College Football, USC Trojans

Sam Darnold hasn't even become a member of the Cleveland Browns yet and already FPI is recommending fading him.

It's only natural to want to compare ESPN's college football FPI projections against the prognosticators in Las Vegas. While FPI's preseason projections and Vegas' win totals are roughly in line on many team's individual forecasts, there are disagreements. And one of the biggest ones this season is the aforementioned Darnold and his USC Trojans.

So let's look at some of the teams where Vegas' win totals and FPI diverge, and find out exactly why FPI thinks certain teams are overrated or underrated on the betting market. All win totals are courtesy of Westgate Las Vegas SuperBook unless otherwise specified.

USC Trojans

Vegas win total: 10.5 (-150 under) | FPI projected wins: 8.92

FPI sees USC as the 12th-best team in the country, including what it considers the 17th-strongest offense. At first glance, that seems like a pessimistic take on the offense, given that last season Darnold had the third-best Total QBR in the nation at 84.8. Additionally, had the Trojans' offense played the entire season as well as it did after Darnold took over, USC would have had the second-most efficient offense in college football (it was the 12th-most efficient offense in 2016 when accounting for the entire season).

But Darnold is not the entirety of the offense. Only five starters return on that side of the ball for USC and, most notably, Darnold will have inexperience in front of him after three members of his offensive line departed for the NFL. (However, it is worth noting that the Trojans do have multiple four-star recruits waiting to potentially take those roles.) While FPI does not specifically take that into account, it does recognize that over the past four years combined ESPN, Scout.com and Rivals.com ranked the Trojans fifth, fifth and sixth in the nation in recruiting, respectively -- and yet still predicts a step back on offense.

FPI sees the Trojans as favorites in 11 of their 12 games (they are narrow underdogs at Notre Dame in Week 8), but they are only slight favorites over Stanford in Week 2 and at Washington State in Week 5, which brings down their expected win total.

Ultimately, part of the difference between FPI and Vegas might be caused by Darnold's reputation. The former freshman phenom is already being touted as the next great NFL quarterback, so perhaps that is prompting -- or is expected to prompt -- favorable bets on the Trojans and driving up the total. That Southern California isn't all that far away from Nevada probably doesn't hurt, either.

TCU Horned Frogs

Vegas win total: 7.5 (-125 over) | FPI projected wins: 8.57

FPI has a big ol' crush on TCU.

There's just no other way to put it. This isn't an indictment of the rest of the Big 12 or about an easy schedule. FPI simply projects a drastic improvement for the Horned Frogs in 2017. It doesn't have them returning to quite the prominence they reached in 2015 -- which the model hasn't forgotten about, by the way -- but it does foresee a possible return to relevance.

One of the reasons for this is that last season TCU might not have been as bad as six wins sounds. Keep in mind: The Horned Frogs played two games into double overtime (vs. Arkansas and Texas Tech) last season and lost both. Based mostly on their (and their opponents') final FPI rating, TCU was expected to win 7.68 games last season, rather than the six they won. That makes 8.57 projected wins in 2017 seem within reason.

It is worth noting that FPI sees a dramatic rise for TCU's offense this season. The model has the Horned Frogs all the way up to 11th in the FBS on that side of the ball, which might come as a surprise. Consider, however, that TCU has 10 starters back on offense, including quarterback Kenny Hill. Though Hill wasn't great last season -- he finished 58th in the FBS with a Total QBR of 61.7 -- he did have some success in 2014 with Texas A&M when he posted a QBR of 80.3 ... but he was benched during a 59-0 loss to Alabama in October and didn't return to the field again for the rest of the season. Interestingly, the model is already optimistic about the offense despite not factoring in Hill's play from his days in College Station.

Utah Utes

Vegas win total: 6.5 (-150 under) | FPI projected wins: 4.68

Between USC and Utah, it might seem like FPI has a problem with the Pac-12.

Instead, FPI's issue with the Utes -- first and foremost -- has to do with their schedule. When the 57th-best team in the country faces the sixth-hardest schedule, that isn't generally a recipe for many wins. Utah has what should be two cupcake wins on its schedule -- North Dakota and San Jose State -- but the closest it comes to another surefire victory is almost a straight toss-up (52.5 percent chance to win) at home against Arizona State.

For the over to cover against the Vegas line, Utah needs -- by FPI's count -- at least four upsets.

Of course, Vegas obviously views Utah as a better team than FPI does, and presumably would quibble with those individual game odds.

But what the model sees is a team that is returning only nine starters. Sure, the quarterback is one of them, but Troy Williams was 81st in Total QBR last season and FPI actually envisions Utah's offense taking a step backward.

Nebraska Cornhuskers

Vegas win total: 7 (-125 under) | FPI projected wins: 5.65

Don't be fooled by the Cornhuskers' 2016. It was nothing special.

Nebraska had the single-worst FPI among Power 5 schools with at least nine wins last season, which gives you a pretty good sense of how unimpressive the season actually was. The Cornhuskers finished the season with an FPI of 7.0, 48th in FBS, getting edged out by fellow nine-win Big Ten school Minnesota (which FPI also rates lower relative to Vegas this season).

So the reality is that Mike Riley enters this season with two disappointing campaigns under his belt. And FPI doesn't see a reason for it to turn around this season, either.

Despite four-year composite recruiting rankings of either 20 or 21, the Cornhuskers' recent history (they have finished 37th, 28th, 39th and 48th in FPI from 2013 to '16, respectively) is a detriment to their projection. On top of that, Nebraska is returning only 10 starters.

If we assume -- which we shouldn't, but let's do it anyway for simplicity -- that Nebraska will win every game that FPI gives them a 75 percent or higher chance to do so and lose every game in which they have less than a 25 percent chance, the Cornhuskers would be 4-4 with four other games -- at Illinois (64 percent chance of victory), vs. Northwestern (38 percent), at Minnesota (51 percent) and vs. Iowa (47 percent) -- remaining. To hit the Vegas over in this scenario, Nebraska would have sweep those four games.

Syracuse Orange

Vegas win total: 4.5 (-130)* | FPI projected wins: 5.65

Syracuse is a team that FPI is planting its flag on as ... well, not a good team, but a good team relative to Vegas' expectation.

Dino Babers is in the second season of a rebuild and coming off a 4-8 campaign in his first season. Vegas sees just a half-game of improvement in 2017. Unquestionably a factor in that 4.5-win line is the Orange's relatively tough schedule (34th most difficult, per FPI).

But FPI sees reason for optimism for Syracuse. For starters, the Orange are returning a remarkable 19 starters, including junior quarterback Eric Dungey. Dungey was solid last season -- ranking 47th in the nation in Total QBR -- before a head injury cut his season short after nine games.

While Dungey returning to the field should be a plus for the Orange, FPI sees the biggest uptick in play coming from the continuity on defense, where they are bringing back 10 starters. Last season Syracuse had the 100th-most efficient FBS defense, but, partially thanks to the school bringing back so many players, FPI envisions the Orange leaping up to 52nd this season.

*Syracuse's win total line is courtesy of South Point Sports Book.

For more from ESPN Analytics, visit the ESPN Analytics Index. 

^ Back to Top ^