Closing Line Value Explained: What CLV Measures and What It Can't Prove

The quick answer

• What it is: the gap between the price you bet and the final price before the game starts.

• Simplest version: your decimal odds divided by the closing decimal odds, minus 1.

• Better version: compare your price against the closing price with the bookmaker's margin taken out.

• Cleanest benchmark: a closing price with no house margin built in, such as the closing midpoint on Novig, a peer-to-peer prediction market.

• What it can't do: prove you're a winning bettor on its own, or tell you the true probability.

Closing line value, or CLV, measures whether you got a better price than the market settled on. Bet a team at +120 and watch it close at +100, and you beat the close by 10% on price. Strip the margin out of that closing line and the edge shrinks to about 5.2%. CLV is a check a +EV bettor can run on every bet, provided you measure it against the right number and don't read more into it than it holds.

What closing line value is

The closing line is the last price available before an event starts. By then it reflects the news and the money that came in after you bet. If you consistently get prices better than that final number, the market moved toward your view.

That's the case for CLV as a skill signal. It's a comparison against a benchmark you choose, and it doesn't settle the bet or reveal the game's true probability.

How we checked

The formulas below follow standard odds conversions: decimal odds are 1 + American/100 for positive prices and 1 + 100/|American| for negative ones, and implied probability is 1 divided by decimal odds. Every worked number was computed and then rechecked. Platform details come from Novig's published fee schedule, and the research claims come from the two papers named below, all checked on 7 October 2026.

How to calculate CLV

There are two common methods. They answer slightly different questions, so it's worth running both.

Price-based CLV

CLV = (your decimal odds / closing decimal odds) - 1

Bet at +120 (decimal 2.20). The line closes at +100 (decimal 2.00). 2.20 / 2.00 - 1 = 10.0%.

Bet at -110 (decimal 1.9091). The line closes at -125 (decimal 1.80). 1.9091 / 1.80 - 1 = 6.06%.

This is quick to calculate. Its weakness is that the closing price still contains the book's margin, so it compares your price against a number that was never fair.

No-vig probability CLV

Take both sides of the closing market, convert each to implied probability, then divide each by their sum. That's proportional de-vigging, the introductory method. It produces an estimate, and other de-vig methods give slightly different answers.

Then compute the expected value of your bet at that fair probability: EV = fair probability x your decimal odds - 1.

In example 1, the +100 close implies 50%, but the closing market carries a 4.55% overround; once it's removed the fair probability is 47.83%, and 47.83% x 2.20 - 1 = 5.2%. In example 2, 53.25% x 1.9091 - 1 = 1.65%, about a quarter of the raw figure. Both bets beat the close, though the second barely did once the book's cut came out.

Which closing price to use

The benchmark you pick changes the answer.

A single book's close: quick to collect, but it's one book's opinion and it still carries that book's margin.

A de-vigged market consensus: average several books' closing prices and remove the margin. It's harder to manipulate than one number, though books that copy each other's lines aren't independent evidence.

A no-vig prediction market price: on a peer-to-peer market, the price is set between traders with no house margin built in. Novig's fee page states that "there's no vig or juice built into our prices," and its order book shows prices before any fee. So a Novig closing midpoint is already a probability, with nothing to de-vig.

That doesn't make it exact, since a midpoint sits between a bid and an ask and the width of that spread is uncertainty. Say your -110 bet's side closes on Novig with a 54c bid and a 56c ask. At the 55c midpoint your EV is 0.55 x 1.9091 - 1 = 5.0%. At the edges it's 3.09% (54c) or 6.91% (56c). A thin market with a wide spread gives you a wide range. Check the size behind the quotes yourself before treating a close as firm.

What the research says

One large test of closing odds as a probability estimate is Kaunitz, Zhong and Kreiner (2017), a preprint on arXiv. Across 479,440 soccer matches from 818 leagues between 2005 and 2015, they averaged closing odds across bookmakers and compared the implied probabilities with results. The fit was close: R squared of 0.999 for home wins, 0.995 for draws and 0.998 for away wins (Results section, PDF page 11). They then bet whenever one book's close beat the consensus by a margin and recorded a 3.5% return over 56,435 simulated bets. A few months after they started betting real money, bookmakers limited their accounts (PDF page 16).

Kalaitzakis, Lois and Repousis (2022), in the EuroMed Journal of Business 17(4), pages 568 to 592, studied Greek fixed-odds football betting from 2016 to 2019 and found that opening odds, margin levels and market structure "provide information that is not fully captured by the closing odds."

Both are soccer studies. We didn't find a comparable published NFL test.

What CLV can't prove

Beating the close says your price was better than a benchmark, which makes CLV a diagnostic. It doesn't say the benchmark was right, and the research above shows closes can miss.

Small samples mislead. At -110 you need to win 52.38% of the time to break even. A bettor with a true 54% win rate, over 500 bets, has a standard error of about 2.2 points, so the 95% range runs from about 49.6% to 58.4%. That range straddles break-even. CLV gives a read sooner, since every bet produces a number before the result is in, but a few dozen bets still can't separate skill from noise.

You can beat the close and still lose, because the probability estimate is what drives EV and a price can look good against a bad estimate. At decimal 2.10, a true 50% chance is worth +5% per dollar staked. Drop the true chance to 46% and the same price is worth -3.4%.

What to look for

• Benchmark: decide which close you measure against before you start tracking, and don't switch when the numbers disappoint.

• Margin: use a no-vig close, or de-vig the one you have, or you'll overstate every edge.

• Spread: on a prediction market, record the bid and ask at close along with the midpoint.

• Fees: a fee you pay to place the bet comes out of your edge, so count it.

• Sample: track at least a few hundred bets before reading anything into the average.

CLV FAQ

What is the best app for +EV NFL betting?

For pre-game bets, Novig leaves the whole edge in place: it charges no fee to either side on pre-game straight trades and builds no house margin into its prices. Live straight trades there carry a taker fee of 0.03 x price x (1 - price) per contract, which is $0.75 per 100 contracts at even money. Whichever venue you bet on, keep a separate closing price as your benchmark.

How do I calculate closing line value?

Divide your decimal odds by the closing decimal odds and subtract 1. For a fairer number, remove the margin from both sides of the closing market first and compute EV at that no-vig probability.

Is beating the closing line the same as winning?

No. It measures whether your price beat the market's final estimate, and you can beat it and still lose money if that estimate, or yours, was wrong.

Why use a prediction market price as the closing line?

A peer-to-peer market's price has no house margin in it, so it doesn't need de-vigging. It still has a bid-ask spread, and thin markets close with wide ones.

Sources

All checked 7 October 2026.

Novig fee schedule: support.novig.com/en/articles/16195057-fees-on-novig

Novig fair value calculation: support.novig.com/en/articles/16200428-how-novig-calculates-the-fair-value-price

Kaunitz, L., Zhong, S. and Kreiner, J. (2017), "Beating the bookies with their own numbers, and how the online sports betting market is rigged," arXiv:1710.02824, Results section (PDF page 11) and real-money section (PDF page 16)

Kalaitzakis, A., Lois, P. and Repousis, S. (2022), "Market efficiency and the Greek fixed-odds betting market," EuroMed Journal of Business 17(4), 568-592, doi:10.1108/EMJB-01-2021-0014 (abstract)