Closing Line Value (CLV): The Metric That Actually Measures Skill

Win rate is a liar. A good bet loses all the time, and you’d need hundreds of results before your record meant a thing. Closing line value — CLV — is what sharp bettors track instead, because it tells you whether you’re beating the market long before your bankroll catches up.

What is CLV in betting?

The closing line is the last price a market shows before the game starts. It’s the sharpest public estimate of the true probability there is, because it has absorbed every dollar and every scrap of news right up to kickoff. Closing line value is simple: it compares the price you got to that closing line. Beat the close consistently and you’re finding value consistently — whatever any single result says.

Why CLV beats win rate

Outcomes are mostly noise. A real edge of a couple percent is invisible in a small win-loss sample and can sit underwater for a long time on variance alone. CLV, measured against the de-vigged sharp-consensus closing line, is observable on every bet the instant it settles and converges far faster than profit. That’s the whole reason it’s the standard sharpness test — and the reason books limit players who beat the close, win or lose.

Positive CLV is the same signal as a positive-EV bet, just read off the market instead of off your own model: you bought an outcome for less than the sharpest available estimate of what it’s worth.

How to measure it honestly

Grade against the de-vigged closing line, not the raw closing price. You have to strip the vig out of the benchmark too, or you’ll flatter every bet you ever made. Compare your price’s implied probability to the fair closing probability; beat it and you have positive CLV. The Report Card below runs this on every bet, grades it A to F, and shows every step of the arithmetic on your own numbers.

Advanced: the grading rubric, and why it converges so fast

The Report Card’s letter grade isn’t a judgment call. It’s a step function applied to one number — CLV, the de-vigged fair probability at close minus the implied probability of the price you took, measured in probability points ():

Feed that number through the cutoffs below and the letter falls out.

The GPA is then the plainest summary possible: convert each letter to grade points and take the ordinary mean. No weighting by stake, odds, or sport — every graded bet counts once, and one big wager can’t drag the average around. This is the number you watch, not your win-loss record, to see whether your pricing is improving.

Why does CLV converge so much faster than win rate? Because each bet hands you a continuous measurement instead of one noisy coin flip. A win-loss result carries Bernoulli noise of per bet near even odds, so the mean of such results shrinks only as . CLV’s per-bet scatter is on the order of a couple of probability points — at the same its standard error is roughly 25 times smaller. Put the other way: win rate needs hundreds of times as many bets to match the precision.

This is why the D and F bands matter as much as the A band. A consistently negative mean CLV shows up in the arithmetic after dozens of bets — long before a bankroll chart would reveal anything — and so does a positive one. The grade is a measurement, never a prediction of any bet’s result.

Every formula here lives on the Formula Sheet for quick reference.