It’s that hard time of the year for college football fans, stuck between spring practice and the fall season.
In a little over two months, stadiums will fill and voices will be hoarse once more. But until then, let’s cure some of the long summer’s wait by talking about some enticing season-long bets.
We’re talking about over/unders on team win totals, and the perfect formula for wagering on them.
For those who don’t know, many of the top sportsbooks allow you to bet on these win totals, and it’s a very simple wager to make.
For example, if Georgia’s win total is set at 10 ½, the bettor decides whether the Bulldogs will win more or less than that figure. In the case of Alabama, whose over/under is set at 10 ½ or 11 (depending on the book), there’s a chance of the bettor gaining back his wagered money and nothing else if the Tide end up with exactly 11 wins.
We all know injuries can ruin a seasonal bet. Look no further than Florida State last year for proof of that. There’s also the factor of having money tied up all regular season. After all, teams must play the full 12 games before a payout.
But these seasonal bets are some of the most confident ones to make. They can often be made on better grounds than single game bets, where a couple fluke plays could make the difference between bottle-popping and remote-throwing.
These are wagers that can be made far more accurately with specific data, schedule analysis based on said data, and good old-fashioned roster breakdowns. That’s what we’re here to bring you today.
What’s the magical stat formula for predicting college football over/unders? Well, a lot of it has to do with a team’s immediate past.
The S&P rating has been around for over a decade, a formula that judges teams based on efficiency, field position, ability to finish drives, and explosive plays. The formula has gained a lot of steam over the years, for good reason.
It is especially useful for picking over or under on season win totals the following year. If a team only had the 40th-best record in college football, but had a top-15 S&P, that’s a good inclination that that team would reap the benefits the next year.
Obviously, departing draftees, graduates, and transfers will need to be taken note of, as will personnel changes and the strength of a team’s schedule. But even through all of this, the S&P has been a solid measuring stick for win-loss trends.
Unfortunately, the sharp minds in Vegas have taken note of this. They’ve been using it to set the over/unders for quite some time. So, there needed to be something more specific that could point bettors to success, something that molds with the S&P but still delivers results in a more digestible manner.
And that advanced stat is called Second-Order Wins (or 2ndO Wins). And while there’s been plenty said about this stat, there’s a more precise way of using it.
No, it’s not a sexy name, but it may just be the under-the-radar stat that wins people some serious money.
Football Study Hall put together a great article and graphic a few seasons back diving into the concept of 2ndO Wins. In a nut shell, it’s an advanced statistic that determines how many games a team should’ve won over a given season.
So, using the four factors we talked about for S&P and adding in turnovers, we get to see what teams were a little luckier (or unluckier) than the rest.
If a team wins eight games, they could’ve had a 2ndO win total of 9.7, 7.2, exactly 8.0, etc.
From research, it looks as if the more extreme the difference between actual wins and 2ndO ones, the more likely the team will go in the direction of whatever the 2ndO numbers are.
When looking at this graph provided by the wonderful folks at Football Outsiders, we can see how all the team’s stack up. This includes S&P ratings for offense, defense, and special teams, as well as the overall S&P values and—most importantly—the easy on the eyes statistic known as 2ndO wins.
The stat itself almost seems too simple to be true. It’s hard to see Northwestern at 10-3 last year and just assume they’ll be worse this year because the previous year they had 2ndO of 7.9. But despite returning their starting quarterback and a dozen other starters, Vegas has the Wildcats over/under set way down at 6.5. And that has a lot to do with the S&P and 2ndO.
So, how accurate is 2ndO at picking a team’s direction the following season? Well, out of the numbers it gave for 128 teams in 2016’s FBS, 69 of them were picked to correctly rise or fall from their previous record. 59 teams were not, because they whether went in the other direction, or their record didn’t change at all.
This almost makes the 2ndO look like just another statistic. That is until you look at the more severe differentials, like Northwestern from last season.
In 2016, 25 teams “should’ve” won 1.0 or more games than they did. That’s a big number, especially when most teams fall into the decibel differences (Iowa minus 0.7, for example).
Out of those 25 teams that 2ndO thought should’ve been better, 19 of them ended up having more wins the next season.
23 other teams should’ve had 1.0 or more wins taken away from their record, according to 2ndO. And 19 of those teams had less wins in 2017. That’s a 38-10 mark overall between the two figures.
2ndO has gotten better as time has worn on too, but it has a healthy history. Over the past five years, the accuracy mark of 1.0 more or less is a whopping 151-79 (.657).
One small drawl back to checking 2ndO wins at the end of the year is the fact bowl games or a conference championship have the ability to add to the win total. Again, wagers will be placed on regular season wins, not total wins. This makes it a bit harder to judge how many wins a team will have during a 12-game schedule.
But the win differential is all that matters, not the total amount. And the only significant differential that can be made is if a team plays three postseason games. The only FBS team to do that last year was Georgia.
With only a bowl game to play, the numbers at the end of the regular season and after the postseason shouldn’t change much at all.
So, the best thing to do in this situation is to find a team whose 2ndO numbers from 2017 add up close to the over/under Vegas has set for their season. That is, only teams that had a 2ndO of 1.0/-1.0 or greater.
For example, take Arizona. They went 7-6 last year and had all seven wins in the regular season. Their 2ndO was at plus-1.0, meaning they should’ve won another game. Going by last year’s 1.0 or greater numbers, the Wildcats have nearly an 80% chance of winning more games this year.
What makes this great is the fact their over/under is set at 7.5. Arizona only needs to be a game better from their 2017 regular season win-total. It makes for a wager with much more potential even before going into scheduling and roster analysis (which we’ll get to in a bit).
While Arizona was plus-1 in terms of 2ndO wins, some other teams had far more drastic numbers.
At the end of 2016, Notre Dame had four wins, but possessed a 7.2 2ndO. That signified a huge potential jump for the team, and that’s exactly what they got, reeling off 10 wins in 2017.
As the differentials get bigger, so do the possibilities for 2ndO accuracy. Six teams had a plus or minus of 2.0 or more last season. Five of those were correctly predicted, not only in terms of the teams getting better or worse, but also getting better or worse by multiple games.
Over the last five years, teams in the plus/minus 2.0 or more category were predicted correctly at a 24-4 clip. Those teams also had multiple wins/losses more 21 out of 28 times.
Given those numbers, a team like Miami-Ohio in the MAC Conference could be the unlikely team to watch this year. They were 5-7 last season and have a 5 ½ over/under this season but boast an impressive plus-2.8 2ndO. It may give people a reason to actually watch some “MACtion” this year. Maybe. But more on this team and others later.
So, even with the wonderful accuracy of the 2.0 figure, it’s hard to tell whether these numbers tip-toe around oddsmakers.
All six “2.0 teams” (as we’ll call them) from 2016 were given over/unders that reflected their 2ndO. That is except for Ohio State, who was at -2.0 after an 11-win regular season. Vegas gave them a 11.0 over/under instead of something lower, and the Buckeyes won exactly 11 regular season games.
But before anyone considers the Vegas oddsmakers to be impenetrable, it’s worth noting some of the teams they were predicting to go up and down. Notre Dame was coming off a 4-win season. They’d only had one other season under five wins since 1963. So, it was quite obvious they’d win more than four.
Same goes for Michigan State, who was clearly headed back up after losing nine games for the first time since 1982.
But some other predictions by Vegas were incredibly accurate. They nearly predicted the five-game drop of Idaho from 2016 to 2017. They put the over/under on seven wins for West Virginia, and the Mountaineers had exactly that despite winning 10 the year before. They even got it right despite likely needing QB Will Grier’s injury to pull it off.
But while this is all impressive, there’s still outliers to be found.
After all, they missed Virginia’s over/under by a full two games in 2016, and Michigan State’s by 2 ½.
So, the best thing to do is still find the teams with the biggest 2ndO differentials, and hope Vegas hasn’t leaned too far in that direction. And after that, select teams on the 1.0-1.9 scale that are being held stable by Vegas.
There are six teams that fall under the plus/minus 2.0 category this season. Those are Army (-2.8) Troy (-2.2), Akron (-2.2), Northwestern (-2.1), Baylor (+2.1) and previously mentioned Miami-Ohio (+2.8).
As for the teams in the other category, there are a few that could follow the 2ndO pattern and win money just by improving by one regular season win. Those would be Arizona, Eastern Michigan, and Ole Miss on the plus side, and Air Force, Boise State, Kansas State, Louisiana-Lafayette Miss State, Old Dominion, and South Carolina on the other.
These are the teams we’ll be exploring further as we head into the next phase of our search.
Finally, we consider the X’s and O’s. While most of a team’s strengths can be found in the five components of S&P, some of those numbers can blossom with personnel and coaching changes.
For example, Arizona (see a theme here?) has a new head coach, Kevin Sumlin. He will be implementing a version of the spread offense that’s better suited for his mobile quarterback, Khalil Tate. These are the types of things to consider in addition to the hard numbers.
Those figures will do most of the work. But knowing where every potential roster deficiency and advantage is will go a long way in deciding whether to commit on a season-long wager.
When blending the 2ndO numbers together with strength-of-schedule (SOS), we can find an extra edge to these teams. We can also concoct these figures with a more practical overview of each team’s opponents.
First off, we looked at the 2ndO and S&P percentile numbers of every one of these teams’ foes. Next, we sprinkled in the consideration of how good the team was the last season in general.
Just because BYU (+1.3 2ndO) is on Arizona’s schedule, doesn’t make them a better opponent than Houston, who only had 0.6 last season. Houston had three more wins than BYU and should be considered the stronger opponent.
Lastly, we look at the current roster of opponents, how much production they’re returning, and how much of it has departed. Any coaching changes, whether it be of the head coach or coordinators, should also be of note.
And after weighing the roster impact with our 2ndO knowledge and the scheduling difficulties, a few different teams stood out to us.
Arizona: Over 7.5
Might as well go ahead and get this one out of the way.
As we said before, this is a team that meets the standard 1.0 2ndO mark, which would’ve given a team a 79.2% chance of improvement last year, and a 75.2% over the last three seasons.
And seeing as they only need to be one win better this regular season as opposed to last, it sets up beautifully for prospective bettors.
As we said before, having Sumlin taking over the offense benefits Tate and his running abilities. Using him outside the tackles as a duel-threat will exploit many weak defensive teams in the Pac-12.
He was impressive as a sophomore in 2017, ending up in the top-15 in passing efficiency. Oh, and he also ran for over 1,400 yards, 327 of them coming in one game.
Tate is the electric playmaker who Sumlin can turn into a college star. Just look what the head coach did with Johnny Manziel.
Meanwhile, Arizona’s defense had all sorts of struggles last year. Their defensive S&P ranked 115th in the country, as they were 123rd against the pass and 108th in scoring on that side of the ball. But they get a break due to having five freshmen starting last season. They’ll have nine returners on defense alone, so things can’t get worse on that side of the ball.
They have too many explosive offensive options not to be a threat for Pac-12 this year. They have a top-5 quarterback, a thin but underrated receiver corp., and four of five starters back on a line that was top-25 in sacks allowed. Throw in two backs in J.J. Taylor and Nathan Tilford who compliment each other well, and it’s easy to see how loaded the Wildcats are.
And luckily for them, the Pac-12 is even more depleted this season.
Arizona matches up well with the majority of their opponents. The FBS adversaries had a combined record of 114-78 last season. ESPN has that ranked as only the 63rd-ranked schedule. The best team that they’ll play from last year, aside from USC, is Washington State.
The Cougars were 9-4 in 2017 but are expected to fall off after returning only four offensive starters. They have no college experience at quarterback and were already a -1.4 in 2ndO last season. Vegas has their over/under at 5 ½, and they certainly could fall that far.
Using our projections, Arizona looks to have easy wins over FCS school Southern Utah and Oregon State (one win in 2017).
Their hated rivals, Arizona State, are projected to take a step back this year. Between the awkward fit of Herm Edwards as coach and the defensive blunders that await them, they’re behind Arizona despite defeating them last season.
They ranked 104th in points allowed, and only have three returning starters on defense. They also were a bit lucky, as they played to a -1.6 in 2ndO. They have a very talented receiver in N’Keal Harris but have a long way to go to replace holes in the run-game.
Between S&P forecasts (ranked 93rd) and player progression, Colorado seems stagnant. The same could easily be said for Cal. Those were both wins for the Wildcats last year and should be easier W’s this season.
Opening at home with BYU won’t be a cakewalk. The Cougars weren’t as dreadful in the second-half after an abysmal start to the season. Their 2ndO puts them in position for a bowl game, but they don’t have near enough punch on offense to keep up with Tate and Co.
With these games in-hand, Arizona would only have to win one game in between Houston, USC, Utah, UCLA, and Oregon to go over 7.5. This means they avoid both Stanford and Pac-12 favorite, Washington.
Depending on how the many newbies on the Utah defensive front mesh, they could be in a world of trouble against Arizona’s stable of rushers. And who knows if the Josh Rosen-less UCLA Bruins will be all that better in Chip Kelly’s first season.
Their two toughest games will come against USC and Oregon, Although they’re both at home. The Trojans will probably be too much for them, but it’s worth noting the USC will still be doing some searching on offense in late September after losing starting QB Sam Darnold to the draft.
Though ESPN’s Football Power Index (FPI) isn’t the greatest measuring stick for which games will be won and lost, it tells an interesting story about those remaining five games.
Arizona has at least a 46% chance of winning each of the Oregon, UCLA, and Houston games, and are favored (57%) in the latter.
South Carolina: Under 7.5
The Gamecocks used a lot of lucky turnovers to help themselves pull out games in 2017. It’s part of the reason they were sitting at -1.5 2ndO when it was all over with.
They were the 110th-best rushing team, barely in the top-100 in scoring, and 95th in special teams S&P. But they were plus-11 in turnovers, and that propelled them to nine wins.
Their over/under is set at 7.5 this season. Taking away the bowl win over Michigan last season, they had eight wins during the main schedule. So, they’d only have to drop one extra game to fall under.