Sport playbook

College football playbook

A huge, uneven market, and what to check first · 8 min read

College football is a huge, uneven market: more than 130 FBS teams, dozens of games on a fall Saturday, and a gap between the best and worst teams that the NFL never sees. This playbook is the college layer on top of Levels 1 to 3. The NFL playbook covers the football basics it builds on, like pricing a hook, teasers and wind.

A huge, uneven market

Next week's openers usually go up on Sunday, and the market settles over the week as news and money correct them (see What happens to a line until kickoff). How far it gets depends on the game:

Big games are close to NFL-efficient. A national TV game between two major-conference teams draws heavy money and every serious modeler, and its closing line is very hard to beat.
Small games are softer, and smaller. A Saturday game between two small-conference teams, one of dozens that day, gets far less attention, so its opener can be further off and its line can move more before kickoff. But limits are low, and books move fast once respected money shows up.
The gap is early. Most of the value in smaller games is found early in the week, at low limits; by Saturday the price is better informed. Why lines move and Steam and market makers help tell information from noise, and closing line value is the honest scoreboard here too.

So many games is also a temptation. A full Saturday card is not a reason to bet more: pass on any game where you have no view.

Spreads and key numbers in college

College spreads run from pick'em to 40 points and beyond, because the gap between teams is so wide. Scoring still comes in field goals and touchdowns, so 3 and 7 still lead, but margins spread out further. In regular-season FBS games since 2015, about 1 in 10 ended on a 3-point margin, against about 1 in 7 in the NFL. 7 held up better, at about 1 in 11 in both. Next come 10, 14 and 17, each in fewer than 1 game in 20, and 6, a secondary NFL number, is no more common than 4 in college, at about 1 in 30. Key numbers explains the idea. In practice:

Value the hook where you are. In college games lined near 3, the favorite won by exactly 3 about 8% of the time, and in games lined near 7 it won by exactly 7 about 8% too, so a hook across 3 or 7 still matters. In games lined near 10, 14 or 17, the favorite won by exactly that number about 4% of the time, half as often; across 5, 8 or 9 a half-point is a sliver, often less than the extra price a book charges for it.
Don't import NFL rules of thumb. Advice on buying points and teasers assumes NFL margins, and college games land on 3 less often. Do the arithmetic on college numbers.

Big spreads and garbage time

A very big spread is often a bet on a game decided by halftime, and what happens after that is noise around the number:

Starters sit. A favorite that leads big pulls its starting quarterback, and backups may give up late scores or stop trying to add them.
Backdoor covers. An underdog trailing badly keeps throwing against those backups, and a late touchdown can beat the number without changing the result.
The market knows this. Big spreads already reflect it, so it is not a reason to take either side. It does mean results near big numbers often turn on decisions made after the game is settled, so judge these bets on the price, not on one cover or one miss.
Big favorites still lose. In games lined near 21, the favorite lost outright about 1 time in 12. A huge favorite's moneyline risks a lot to win a little.

Information that moves college lines

College news is messier than NFL news, and the market still prices most of it.

Injury reporting. College programs have historically disclosed far less than NFL teams, and the reporting rules have been changing. Treat any one report as one input, and expect late surprises.
The quarterback. As in the NFL, he moves lines more than anyone else, and the drop to a backup can be bigger in college. With less public reporting, his status often settles late (injury impact).
Roster turnover. Transfers and departures reshape college rosters far more each year than NFL ones, so last season's results say less about this season's team. A new head coach or coordinator can change a team's style and tempo too.
Motivation and bowl season. Bowl games bring opt-outs, as players skip the game to prepare for the draft or to transfer, and coaching changes before the game. Those are real information, and the market moves on them as they become known.

The common thread: well-known news in a big game is priced fast. Where an edge from news exists, it comes from being early or judging the replacement better, usually in smaller games and at low limits.

Over/unders

A college over/under can sit in the 40s for one game and the 70s for the next. What moves it:

Tempo. Some offenses snap the ball as fast as they can and others drain the clock, and the gap between them is far wider than in the NFL. Judge how the two styles meet, not each team's average alone.
The clock. College clock rules differ from the NFL's and have changed in recent seasons, which changes how many plays a game gets. Make sure play counts you compare come from the same rules.
Weather. Wind matters most, as the NFL playbook explains.
Landing points vary. In games lined near 55 or 59, about 1 in 20 landed exactly on that number, so a half-point across it is worth something; across 50 or 60 it is worth almost nothing.

Early season and late season

In September, books and bettors price teams on last year's results, recruiting, transfers and new staffs, with few games this year to check against. Mistakes are bigger in both directions, so early prices move more, and early results mislead too: a blowout of a much weaker opponent says little, and schedules vary widely (strength of schedule). As the season goes on, power ratings settle, the market gets sharper and edges get thinner. The early weeks offer more room to be right and more room to be wrong, so keep stakes small.

Home field and travel

Home field varies more in college than in the NFL: some stadiums are huge and loud, a few teams play at altitude, and some home crowds are small. The market already prices venues differently. Popular angles about long trips, early kickoffs after crossing time zones, or a look-ahead to a rivalry game rest on thin evidence, are widely known, and are not a bet by themselves: tiebreakers at most.

A pre-bet checklist

What kind of game is it? A big national game usually has a sharp price; a small-conference one has lower limits and bigger moves.
Who is at quarterback? If either starter is in doubt, know how far the line has already moved for it.
What changed since last season? Transfers, a new coach or new coordinators matter most early in the year.
Where is the number? Near 3 or 7 a hook matters; near 10, 14 or 17, about half as much; elsewhere the price usually matters more.
Is it a big spread? Then you are also betting on backups and garbage time.
Have you shopped the price? Compare books first (Line shopping).
What has the line done since it opened, and why?
Is the stake small? One shared misread, like a wrong read of a new coach behind several bets, can sink a few at once, so size for a bad Saturday (Variance and drawdowns).

How Consensus Edge helps

We compare prices across many sportsbooks and remove the vig, so you can see what the market really thinks and where the best price is. We blend Square (the books), Sharp (our models) and Crowd (prediction markets) into one chance, and we re-price regularly until kickoff, so line moves after news reach our numbers, though right after news any board, ours included, can trail the books. Each bet shows BET, LEAN or NO EDGE; BET and LEAN carry a small suggested stake, and we publish how our picks did against the close. NO EDGE is not a grade: it means a fair price, or a market we don't grade. Start on the slate or read how it works, and bet only what you can afford to lose.

Example. One point, two very different values (illustrative prices, not live data). A game between two small-conference teams opens on Sunday with the home favorite at −6.5, priced −110 on both sides. You bet the favorite on Monday at −6.5, −110. By kickoff it closes at −7.5, still −110 on both sides.

Remove the vig from the close by splitting the extra evenly, as in the CLV lesson: each side implies 52.38%, they add to 104.76%, and taking half the extra, 2.38, off each leaves a fair 50.00%. Your −6.5 wins everything the closing −7.5 wins, plus every game the favorite wins by exactly 7. In college games lined near 7.5, that happened about 7.8% of the time. If the close is right, your chance is 50.00% + 7.80% = 57.80%, and your $100 ticket, which wins $90.91, is worth 0.578 × $90.91 − 0.422 × $100 = $52.55 − $42.20, about +$10.35 on average. Small-game lines can move this far, but most bets that beat the close beat it by far less, and books take only small bets on small-conference openers.

Now move everything one point lower. Suppose you had bet the favorite at −5.5 and it closed at −6.5. You got the same full point better than the close, but the number you gained is 6: in college games lined near 6.5, the favorite won by exactly 6 only about 1.6% of the time. Your chance is 50.00% + 1.60% = 51.60%, below the 52.38% you need at −110, so the ticket is worth 0.516 × $90.91 − 0.484 × $100 = $46.91 − $48.40, about −$1.49. Same size of move, opposite verdict. Measured as the CLV lesson does, at your own number, your CLV is 51.60% − 52.38% = −0.78 points: a better number is not always a better price. The point still helped, since −6.5 at the close was worth about −$4.55, but it did not cover the vig. A point is worth only as much as the games that land on it: a lot through 7, very little through 6.

Key takeaway. College football rewards knowing which kind of game you are betting, so price big and small games differently, value each hook by college margins, respect roster news, shop every bet, and judge yourself on the close.

Terms in this playbook

Opening Line Closing Line Line Movement Spread Hook Moneyline Injury Impact Over/Under Pace SOS Power Rating Home Advantage Square Sharp Crowd BET LEAN NO EDGE