Saturday NCAAF CLV Workflow
FairLine to books to close, window by window
8 min read
The Quick Answer
A college Saturday is won on Monday, and graded on Sunday.
Compare each book's number to a vig-free reference price when the lines post early in the week, shortlist the gaps, place or discard them across midweek as the market settles, write down the number and price you actually took, then capture each game's close at its own kickoff and grade yourself on Sunday.
The part that makes this a college workflow rather than an NFL one: a Saturday has four or five separate kickoff windows, so there is no single close. The routine has to be interruptible, and the slate has to be triaged before it can be priced.
What this page assumes you already have
- The definition and the arithmetic. Converting two prices to implied probability and subtracting them is covered in what closing line value is, along with what counts as a good average. This page does not restate those thresholds.
- The early-versus-late decision. Whether to bet an opener or wait is its own question — see when to bet early vs late. Betting early only helps if you are ahead of the market rather than merely first.
- One inherited rule that governs everything below: CLV only means something measured against a sharp closing number. Grade yourself against a soft book and you will flatter yourself, because that is the number most likely to have been wrong.
The reference price this workflow compares against is the FairLine — a vig-free consensus number, not a prediction of the result.
The Week, Window by Window
| When | What you are doing | Why then |
|---|---|---|
| Sunday – Monday | Next Saturday’s numbers post. Compare each book to a vig-free reference and write down the gaps. | This is when a book is most likely to be off, because nobody has corrected it yet. |
| Tuesday – Thursday | Work the shortlist. Most bets get placed or discarded here. | Money and information arrive through midweek. A flag that survives to Thursday is usually a real one. |
| Friday | Availability and weather pass. Re-check anything wind- or temperature-sensitive. | College has no mandated, dated injury report. Friday is when what news exists tends to surface. |
| Saturday, pre-noon | Final pass, then stop. Record the number and price you actually took. | If you did not write down your number, you cannot compute CLV later. This is the step people skip. |
| Noon ET kickoffs | Capture closes for that window only. | Each game closes at its own kickoff. There is no one Saturday close. |
| 3:30, 7:00 and 10:15 ET windows | Repeat the capture as each window goes off. | Four or five staggered windows is the defining feature of a college Saturday, and why this workflow has to be interruptible. |
| Sunday | Grade. Compute CLV per bet, append to the log, look at the running average. | Results settle overnight. Grading while the slate is fresh is the only version of this that gets done. |
Triaging Forty Games Down to Ten
The board is the problem. Forty-plus games go off on a full Saturday and no one prices all of them, which means the useful skill is deciding what not to look at. Triage on where a price is structurally likely to be stale rather than on which games you would enjoy watching — those two lists barely overlap.
Attention, not quality
The marquee night game is the one the market has spent the most time on. The mispricings concentrate in the games fewest people are watching, which is a claim about oddsmaker bandwidth rather than about football.
Market conditions, not hunches
Give yourself a rule that removes games before you think about them. Ours is an example: college totals only interest us when the game is a genuine mismatch, because a close game and a blowout are different distributions and averaging them hides both.
A list you can finish
Six to ten games you can actually price beats forty you cannot. A shortlist that does not fit in the time you have is the same as no shortlist, and it fails on the busiest Saturdays rather than the quiet ones.
What to Write Down, and When
CLV cannot be reconstructed after the fact. If you did not record the number you took at the moment you took it, Sunday's grading is guesswork. Six fields are enough:
- Game and market — enough to find the close again.
- Side and line — the number, with its sign.
- Price — the juice, not just the line. On large spreads this is where most of the movement lives.
- Book — where you took it.
- Timestamp — so a negative average can be traced to when you bet.
- Stake — for results, not for CLV, but you will want them in one place.
Then, at each kickoff window: record the closing line and closing price from your reference book for the games in that window only. Trying to collect all of Saturday's closes on Sunday morning does not work — the numbers are gone, and the ones you can reconstruct are the ones that moved least. For a general system for this, see tracking your bets.
Half a Point Is Not Half a Point
The standard CLV example moves a line from +3 to +2.5, and that half-point is enormous because 3 is the most common margin in football. College football is where that intuition breaks, because the same half-point on a 24-point spread crosses almost nothing — margins around 24 are thinly and evenly spread, so the number barely encodes probability there.
The practical consequence: on large college numbers, the price carries more signal than the line. A spread that sat still at −24.5 while the juice went from −110 to −125 moved against you, and a log that records only the number will score that as flat. On small spreads near 3, 7, 10 and 14, the reverse holds and the number is nearly everything. Record both and you do not have to decide in advance which mattered.
Our own CLV, since we are asking you to measure yours
It would be poor form to teach this diagnostic and not publish our own result. As of 27 September 2026, our average closing line value was negative in all three markets. On moneylines it sat at roughly −1.0 points of implied probability across 439 graded bets; on spreads and totals it was slightly negative measured in line ticks, across 710 and 388 bets respectively. Between 29% and 35% of our graded bets beat the close, depending on the market. Those are site-wide figures across every sport we cover, not college football alone.
By the standard set out on our own CLV page, a negative average on a sample that size is a process problem rather than bad luck — and the honest reading is that it points at where and when we are betting, not only at what we select. We would rather you saw that number from us than found it yourself.
Two things worth separating, though. CLV asks whether you got a better number than the market settled on; results ask whether the bets won. Those can disagree for a long time, in both directions, and ours currently do. The graded record — every selection, win or lose — is on the public results page, which publishes the win-loss record, ROI and streak. It does not currently publish per-bet closing lines, so the CLV figures above come from our own grading rather than from anything you can recompute there. Treat the dated numbers here as a snapshot, not a live feed.
Start the Week With the Gaps Already Found
EdgeBoard compares every book to the FairLine continuously, so Monday's shortlist is waiting for you.
Frequently Asked Questions
What does a Saturday NCAAF CLV workflow actually look like?
It starts on Sunday or Monday when the following Saturday's numbers post, not on Saturday morning. You compare each book's number to a vig-free reference price, flag the gaps, place or discard those flags across Tuesday to Thursday as the market settles, re-check availability and weather on Friday, and then record the number you took. Saturday itself is mostly capture: college kickoffs run in four or five separate windows, so there is no single close to grade against — each game closes at its own kickoff.
How do I narrow a 40-game Saturday slate?
Not by trying to handicap all of it. Triage on where mispricing is structurally likely rather than on which games you want to watch: lower-profile conferences, games the market has spent the least attention on, and spots where a specific market condition holds. Our own filters are an example of that logic — college totals only interest us when the game is a genuine mismatch, because a close game and a blowout are different distributions. A shortlist of six to ten games you can actually price beats a spreadsheet of forty you cannot.
Which book's closing line should I grade my CLV against?
A sharp one, and then the same one every week. CLV only means something against a closing number that was probably right, so grading against a soft book flatters you — that is the book most likely to have been wrong, which is why you were betting there in the first place. Pick the sharpest close you can see, use it consistently, and note that a vig-free consensus price and a single book's close are two different references. Do not mix them mid-season.
Does half a point of CLV mean the same thing on a 24-point spread as on a 3?
No, and this is where college football differs most from the NFL. Moving from +3 to +2.5 crosses the single most common margin in football, so half a point there is enormous. Half a point on a 24-point spread crosses almost nothing, because margins around 24 are thinly spread. On large college numbers the price movement carries more of the signal than the number does, so record both the line and the juice — a spread that stayed at -24.5 but went from -110 to -125 moved, even though the number did not.
How many bets before my CLV average means anything?
Think in hundreds, not dozens, and never treat one Saturday as a sample. A single slate can run positive or negative on noise alone. The point of logging every week is that the average becomes informative somewhere in the low hundreds of graded bets, and until then the number tells you much less than the discipline of recording it does.
Is +EV Bets currently beating the closing line?
No, and we publish the figure rather than leaving you to assume. Across our graded sample site-wide - not college football alone - average CLV was negative in all three markets as of 27 September 2026, with between 29% and 35% of bets beating the close depending on the market. That is a process problem we are working on, not a rounding error, and it is exactly the diagnostic this page teaches you to run on yourself. One honest limitation: the public results page publishes the graded win-loss record, ROI and streak, not per-bet closing lines, so the CLV figures above come from our own grading rather than from anything you can recompute there today.