Learn

Closing line value is the leading indicator, and profit is the lagging one

Closing line value compares the price you took against the price the market settled on. It is the earliest honest signal that a bet was good, it is measurable in dozens of bets rather than thousands, and it is not a guarantee of anything.

Updated October 24, 2026 · 10 min read

Educational only. quant+ does not take wagers.

You take a side at -110. It closes at -130, with the other side at +110. All three numbers below are computed from those two prices.

Your price -110 converts to an implied 52.381%. 100 / 210

The -130 / +110 close carries a 4.141% overround. 56.522 + 47.619 - 100

Removing it leaves a vig-free close of 54.274%. 56.522 / 104.141

Relative CLV is 3.615%. (54.274 - 52.381) / 52.381

That is also 3.615% expected value per unit. 54.274% x 1.9091 - 1

Two prices, one subtraction, and a choice about which answer you want

CLV is not a single number. It is a family of related numbers, and the disagreements you see about it are usually because two people are reporting different members of that family.

The relative form

Divide the change in implied probability by the implied probability of the price you took. This is the version that carries information about money, because it is a percentage of the stake rather than a percentage of a probability scale.

CLV relative = (p_close - p_bet) / p_bet

The absolute form

Report the raw gap in implied probability, in percentage points. Easier to describe out loud, and misleading when you compare bets of different prices, for a reason the cents trap section below makes concrete.

CLV points = p_close - p_bet

Both forms need the same input, and getting that input right is the whole of the difficulty: a vig-free closing probability. Use the raw closing price and your answer is inflated by the margin, which in the worked example above is the difference between 7.905% and 3.615%.

If the closing price is fair, relative CLV is the bet's expected value

This is not an approximation. It is an algebraic identity, and it is the reason CLV is treated as evidence rather than as a scoreboard.

Write the vig-free closing probability as p. Take a bet at decimal odds d, where the implied probability of your price is a, so that d = 1 / a. Relative CLV is (p - a) / a. Expected value per unit staked is p x d - 1, which is p / a - 1, which is (p - a) / a. The two expressions are the same expression.

CLV relative = (p - a) / a and EV per unit = p x d - 1 = p / a - 1 = (p - a) / a

The consequence is worth stating plainly. Measuring your CLV is not a proxy for measuring your edge. On the assumption that the vig-free close is the true probability, it is the measurement. The assumption is the whole argument, and the section on illiquid markets is about where it fails.

Your price Vig-free close Relative CLV EV per unit
-110 55.556% (a -125 / +110 close) +6.061% +6.061%
-110 54.545% (a -120 / +105 close) +4.132% +4.132%
-110 53.488% (a -115 / +100 close) +2.114% +2.114%
-110 52.381% (no movement) 0.000% 0.000%
-110 50.980% (a -104 / -110 close) -2.674% -2.674%

Each row is one column computed two ways. The two right-hand columns agreeing is the identity, not a coincidence of the numbers chosen.

Why you can read CLV after 16 bets and profit only after thousands

Profit and loss is the outcome you care about and the worst instrument for measuring whether your process works. The variance swamps the signal, and it does so for years.

Suppose your bets carry a real edge of 2 percentage points and your CLV readings vary with a standard deviation of 4 percentage points. Detecting a 2-point average CLV at 95% confidence takes roughly 16 bets. Detecting a 2% return on investment from profit and loss takes roughly 8,668 bets, because a spread bet's per-bet outcome swings about 0.95 units around a mean of 0.02.

Signal Effect to detect Bets for 95% confidence Bets for 80% confidence
Closing line value 2.0 percentage points 16 7
Closing line value 1.0 percentage points 62 25
Profit and loss 5% return on investment 1,387 593
Profit and loss 2% return on investment 8,668 3,706

At a 2% edge the gap is a factor of about 542. That is the entire practical case for CLV: not that it predicts the future, but that it produces a usable reading of the present before your bankroll has had time to tell you anything at all.

A verdict on a month of bets

Sixteen bets is a slow weekend. CLV can tell you something about your process after one, when profit and loss is pure noise.

A verdict before the sample grows

Most bettors never place 8,668 bets. Judging yourself by profit and loss means judging yourself on evidence you will not live long enough to collect.

Not the same as predicting

Beating the close means your price was better than the market's final opinion. It does not mean you forecast the game better than the market did.

Ten cents is not a unit of anything

Bettors often report CLV in cents of line movement. The unit is intuitive and it is not comparable across prices, which makes it a poor basis for deciding whether your process works.

A cent is one point of American odds and, for a favourite, the number of cents in practice is the movement in the lay price. Two wagers can record the same 20 cents and carry very different values, because implied probability is not linear in the price. The table below uses the same vig-free-close machinery as above, with each pair's own overround removed.

You bet It closed Cents Implied change Relative CLV
-110 -130 20 52.381% to 56.522% +7.905%
-200 -220 20 66.667% to 68.750% +3.125%
-110 -115 5 52.381% to 53.488% +2.114%
-250 -300 50 71.429% to 75.000% +5.000%
+110 +100 10 47.619% to 50.000% +5.000%

Twenty cents at -110 is worth more than twice twenty cents at -200. Fifty cents at -250 is worth less than twenty cents at -110. A log that records only the movement cannot tell these apart, and a positive average of cents across a mixed book of favourites and underdogs can hide a negative average of value.

CLV is only as good as the market it is measured against

The identity above assumes the closing price is an efficient estimate. In thin markets that assumption is doing more work than the data supports, and a positive CLV reading can mean nothing.

Markets where the close is thin

  • Player propositions, where stake limits are small and prices can move on a few hundred of exposure rather than on a disagreement about the player.
  • Low-limit leagues and niche competitions, where the closing price is better described as a current guess than as a settled consensus.
  • Early-season college markets, where rosters are unsettled and the market is still finding the teams. A price moving here is not a market converging on truth.

The edge the market cannot absorb

There is a second and less discussed failure, which is the opposite of the first. A bettor can hold a genuine, repeatable edge in a liquid market and still record unremarkable CLV, because part of that edge is structural and never gets absorbed into the price. Systems built on information the market does not act on, or on a transaction the market cannot replicate, will understate their own quality if CLV is treated as the only scoreboard.

Both failures point the same way: CLV is a measurement of agreement with the market, and the market is not always right, and is not always the thing you are trying to beat.

Four ways to compute CLV wrong

If you are going to use this as your primary evidence, the log has to be right. These are the errors that make a CLV record flatter you.

Comparing against a raw close

The most common error. The closing price contains the book's margin, and the margin on a two-way market of average price is worth more than most edges. In the worked example, skipping the de-vig turns a 3.615% reading into a 7.905% one, overstating your edge by a factor of more than two.

Not recording the timestamp

Closing is a moment, not a day. A price captured hours before the start is not the close, and a price captured after the start is not a forecast of anything. Timestamps also matter because the closer to the start you recorded, the more of your figure is market movement rather than your own timing.

Ignoring pushes and settlement rules

Spread and total bets have three outcomes, and a push changes the arithmetic. A line that moves across a whole number does not have the same value as one that moves half a point, and a comparison that treats them as the same movement will misreport both.

Averaging a moving line

Where a line moves in several steps, averaging the prices you recorded is not the same as using the price at the close. Record the last price before the start, and record the others separately rather than blending them.

What a positive CLV record does not establish

It does not establish that you will profit. The identity above holds on average, on the assumption that the close is fair, and it says nothing about any individual bet or any individual month. A bettor can beat the close consistently and still lose money over a season, and the reverse happens constantly: someone with no edge at all can run a positive return on a few hundred bets.

It does not establish that your model is good. CLV measures the price you obtained against the price the market settled on. It is a joint verdict on your pricing and your timing, and a bettor with a weak model who posts early into slow markets can produce respectable CLV.

It does not transfer across markets. A CLV record built on major-league moneylines says very little about your ability in a thin prop market, because the thing being measured, the efficiency of the close, is not the same in the two places.

And it is not the goal in itself. Anyone who has spent time with this arrives at the same phrasing eventually: it is a comfort, not a crutch. The question that matters is still whether the bets were right.

Measure the process, not just the bankroll

A journal that records the price you took and the price it closed at turns every bet into evidence about your process, months before profit and loss can.

Open the bet journal ยท CLV calculator

Questions people ask

What is closing line value in simple terms?

It is the difference between the price you got and the price the market closed at, expressed as a percentage of your own implied probability. If you bet -110 and the vig-free close was -125, that is a positive 6.061%. If you bet -110 and the vig-free close was -104, that is a negative 2.674%.

What is the formula for closing line value?

Relative CLV is (p_close - p_bet) / p_bet, where both are implied probabilities and p_close is taken from a vig-free closing price. Without removing the margin, the number is inflated: in the -110 to -130 example, the raw close gives 7.905% and the vig-free close gives 3.615%. You can also report the raw gap in percentage points, but it is not comparable across bets at different prices.

Does positive closing line value guarantee profit?

No. It is a strong signal and it is the earliest usable one, but it is not a guarantee. Relative CLV equals expected value per unit only under the assumption that the vig-free closing price is the true probability. In liquid markets that assumption is a good one. In thin ones it is not, and a positive CLV figure there can be an artefact of low limits rather than evidence of skill.

How many bets do I need before closing line value is meaningful?

Far fewer than profit and loss. If your CLV readings have a standard deviation of 4 percentage points and your true edge is 2 points, about 16 bets give a 95% confidence reading of the average. Detecting the same edge from profit and loss takes roughly 8,668 bets. That difference, a factor of about 542, is the practical reason CLV is used as a process check.

Is closing line value the same as beating the market?

It is the measurement of it, and the two are not identical in thin markets. Beating the close means your price was better than the market's final opinion. Where the market's final opinion is well formed, that is meaningful. Where limits are small and the price is set by who happened to bet, it is closer to an accident than a skill, and the number should not be read as a verdict.

Sources

  1. SERP: closing line value, marginalia-search.com
  2. Getting Precise About Closing Line Value, unabated.com
  3. What's A CLV in Sports Betting? Closing Line Value Explained, thelines.com
  4. Closing Line Value: How to Use CLV to Track & Beat Betting Odds, sportsbettingdime.com
  5. What is closing line value profit and closing line value yield?, godsofodds.com
  6. Closing Line Value in Sports Betting and Why CLV Matters, valuepunter.com
  7. Closing Line Value (CLV) & Line Movement Explained, thespread.com
  8. Closing Line Value (CLV) Explained, mybookie.ag
  9. The Web EP. 3: Closing Line Value, jackweberstheweb.substack.com
  10. How Important is Closing Line Value in Sports Betting?, ff-winners.com
  11. What is Closing Line Value in sports betting? Basics of CLV explained, esportsinsider.com
  12. What is Closing Line Value in Sports Betting? How to Track Closing Line Value, oddsjam.com
  13. Attempted Pinnacle CLV page, pinnacle.com