The offseason is a time for improvement and that’s exactly what we strive for at BSN ATS. While you were watching baseball or playing with your kids this summer we’ve been going through different metric variations to see what worked best for us last year in betting on college football while trying to apply NFL style metrics like YPP to the collegiate game.
There are some obvious challenges such in trying to do this, as the differences in competition level and ever changing rosters from year to year make these numbers much more unpredictable.
In going back and trying to figure it all out we landed on three different factors that can help us in calculating our own spread: yards per play, returning points, and strength of schedule. These three factors can help us find baselines between all these teams and try to make sense of it all. Our results in 2016 are fairly encouraging with two different variations (which come down to the weight given to one of the three above mentioned metrics versus the other) came out to do pretty well last year doing around or better than a 55-percent clip against the Las Vegas Spread.
Of course, the variances from one metric to another can make a big difference even offering different favorites in certain games. One example is Friday’s Rocky Mountain Showdown, where our metrics differ on who should be favored. This comes down to how you weigh YPP where CSU is rated higher while CU is aided by a great strength of schedule from 2016 which puts their numbers in perspective evening things out - another interesting stat, CSU ’s offense ended last season on a tear leading the nation in YPP average in their last three games with a 9.3 average. All this to say, finding consensus picks amongst the metrics is ideal.
There are a few caveats worth applying here; spreads where Vegas had a favorite of 25 points or more were unreliable so we try to stay away from those. Also, the first two weeks of the season where we applied the 2015 figures didn’t do as well so this first month will get tricky - which is also why we’re not going in great detail on the numbers and formulas here.
Of course, ultimately the numbers are just a guide to narrow down what games may have an incorrect line pushed by public perception.
With that, we’ve identified a few upsets and a few favorites that pass both the eye and the analytical test as we dive into another experimental season with NCAA spreads - which will always be more volatile given the nature of college football and the youth and inexperience these kids have.
As always we don’t shy away from predicting big games either as you’ll see here below, even if that means going against the numbers (and our best judgment).
What this really means is football is back! so here we go.
A big thank you goes out to Andrea Cornaglia who's been essential in plugging in all the different variables and spear heading this experiment in trying to best predict NCAA metrics.
Dogs the numbers love
Wyoming +11.5 at Iowa
No matter how you split it, our metrics say the Hawkeyes shouldn’t be favored by more than a touchdown here, even if they're playing at home. Both teams return 14 starters after winning seasons in 2016, but there’s only one returning quarterback here and he should make a big difference. Wyoming’s Josh Allen is a potential top NFL pick and he’ll have to play big against a veteran-laden Iowa defense for this upset to come to fruition.
Wyoming’s defense wasn’t very impressive a season ago but with eight returning starters they should be improved. Iowa’s offense is the big question mark here as they’ll be heavily reliant on their running game and their top back, Akrum Wadley, isn’t a bell cow.
The Cowboys look poised to keep this close. I might be skeptical at -7 or -10 but -11.5 is right in the sweet spot. Go Pokes!
Temple +17.5 at Notre Dame
Look, Temple’s been a good program and they were a much better team than the Irish last year, at least record wise. So even if this is at Notre Dame, 17.5 (it’s 18.5 in some places) is just too many points.
That said the Owls won’t be the same team we’ve seen the past couple seasons. They’ll be replacing a three-year starter at quarterback and only return four starters on a defense that’s been one of the best in the country.
Notre Dame is the better team on paper and should be favored at home according to our metrics. But that’s just too big a number for a Temple team that’ll come in wanting to prove something, especially as they go through a coaching change.
While talented, the Irish are young at key spots, most importantly quarterback. After a terrible 4-and-8 season the pressure is high to have a bounce back season. Shaking that pressure might not be so easy against a worthy opponent in the opener. Give us the Owls and the points.
Favorites the numbers love
Virginia Tech -4 versus West Virginia
In a top 25 matchup played at FedExField the home of the Washington Redskins on Sunday night, this should be one of the big games of the weekend.
Both teams come into the season with the promise of being surprise top-25 candidates and a lot of it will come down to quarterback play where the Mountaineers have the apparent edge with former Florida QB Will Grier. Tech will start freshman Josh Jackson so there’s a lot more uncertainty with them.
The Hokies advantage is on defense where they return seven starters and have NFL talent with great size all over their back seven. That’ll be a big test for Grier and the West Virginia offense. WVU is only returning seven starters on either offense or defense so it might take a while to get clicking which is why the numbers like Va Tech here and so do we.
Louisville -24.5 at Purdue
This is close as our numbers advise caution but it’s also rare to see a favorite of this large a spread even be considered. Another neutral site game here played in Lucas Oil Stadium, Louisville is the obvious favorite coming off a big 2016 season while Purdue was 3-and-9.
The big difference here lies in the quality of returning starters, Purdue brings back their starting quarterback in David Blough and 10 other starters with him. But that’s nothing compared to the reigning Heisman in Lamar Jackson, some high-level defensive playmakers at cornerback and their rush linebackers where James Hearns and Trevon Young should have big games with only two returning starters on the Boiler Makers o-line.
The other factor here is that the Cardinals laid the smack down on opponents last year winning games by outlandish margins especially early where they won by an average of 41 points in September including a 63-20 throbbing of Florida State. If Louisville comes out the gate with the same sort of fire in 2017 as they did in 2016 this will be the easiest money you’ve made all week.
With Purdue allowing 38.3 points per game last season on defense, good for 117th in the country, I like our chances.
Big game hunting
Alabama -7 against Florida State
Ok first things first, the numbers advise caution here with the seven point spread. All metrics favor the Tide, but the margins are in the three-to-five point range. But it’s rare to get a Nick Saban team in single digits so we might as well pounce.
Last season Alabama was 10-and-5 against the spread and only won one regular season game by less than seven points last year, against Ole Miss in a shootout 48-43.
Here’s the thing about Saban he’s had all offseason to prepare for the Noles and he’s going up against a man he knows well in Jimbo Fisher. He’s has a returning starter at quarterback for the first time in the last three seasons.
The big factor here should be FSU’s two returning offensive lineman a big mismatch against Alabama’s alway talented front. Which exactly what makes the Tide so hard to play close, let alone beat. The run game and QB play should be good enough to put up points and if FSU loses pace at all they’re done, as that’s when Bama can put on the pressure and let their secondary talent shine. They can also finish you off with their run game. The Seminoles will need a big game from their talented defense, which could happen, but I’m just not sure it’ll be enough if the offense can’t put up 30 plus.
Give us Bama in one of the few chances you’ll have to bet on them at a reasonable number.
Michigan -4 against Florida
Part of big game hunting is going against the numbers if need be and this is a classic case. Florida is favored by one metric so there’s not a consensus towards Michigan but the numbers don’t take into account the losses Florida’s suffered through recent suspensions.
Now 10 Gators will be suspended including their most essential playmakers on offense, receiver Antonio Callaway and running back Jordan Scarlett. Those are massive losses for a Gators team that needs all the help they can get. Especially since they’ll be starting freshman Feleipe Franks at quarterback.
In a neutral field site that’s a big ask of a freshman QB in a nationally televised game. Michigan’s lost lots of talent from a year ago, but they return reliable starting quarterback Wilton Speight. While Michigan's lost lots of defensive talents to the pros Jim Harbaugh’s recruited really well and now's the time to show it on the field.
Michigan likely won’t be as good as last year but they should still be plenty good to beat this Florida team with all the uncertainties on offense.