Closing Line Value: The Best Measure of Betting Skill
Win/loss record tells you what happened. Closing line value tells you whether what happened was earned. CLV is the single best leading indicator that your process is profitable — and the single best way to know whether to keep trusting it during a losing stretch.
1. What CLV is
Closing line value is the difference between the price you got on a bet and the price that same bet closed at right before kickoff. If you bet the Bills –3 at –110 on Wednesday and the line closes at Bills –4 at –110 on Sunday morning, you got a much better number than the market settled on. That's positive CLV. If you bet Bills –3 at –110 and it closes at Bills –2.5 at –110, you got a worse number than the closing market. That's negative CLV.
CLV measures one specific thing: did you, at the time you bet, have a better number than the eventual consensus of all market participants? It's a measurement of whether you out-predicted the market — not whether you out-predicted reality. The game outcome is irrelevant to CLV. You can lose a bet that had massive positive CLV, and you can win a bet that had terrible CLV. The line at kickoff is the benchmark; everything else is variance.
The closing line incorporates all available information at game time — injury reports, weather, sharp money, public action. By the time the line locks, it's the sharpest number the market will produce. Beating it consistently means you're consistently ahead of where the market ends up.
2. Why the no-vig closing line is the benchmark
The closing price the book displays still has vig baked into it. To measure your skill cleanly, you compare your bet to the no-vig closing probability — the price the market would offer if it weren't taking commission. This is the same vig-stripping process covered in the Vig & Juice guide.
Why? Because the vig is the book's, not the market's. The no-vig probability reflects what the collective bettor and bookmaker information actually implies; the posted price reflects what the book wants to charge you. Skill is measured against the underlying reality, not the retail price tag.
noVigProb = impliedYou / (impliedYou + impliedOther)
# Example: line closes at -120 / +100
impliedYou = 120 / 220 = 54.5%
impliedOther = 100 / 200 = 50.0%
noVigProb = 54.5 / (54.5 + 50.0) = 52.2%
So if you bet that side at –110 (52.4% implied), your bet implied a 52.4% break-even probability — and the no-vig closing market is now saying 52.2%. You're slightly behind the close. If you'd bet that side at +100 (50% implied), you'd be well ahead of the close — sharp.
3. How to measure CLV on a single bet
There are two common ways to express CLV. Both have their place; serious bettors usually track both.
Price-based CLV (in cents)
The crudest measure: how many cents better was your price than the closing price? You bet a side at –107. It closed at –118. You got 11 cents of CLV. This is intuitive and easy to track, but it doesn't normalize for whether you got that 11 cents on a –200 favorite or a +400 dog (the same cent difference is worth different amounts on different prices).
Probability-based CLV (in %)
More rigorous: what was the no-vig closing probability minus your implied probability? You bet at –107 (51.7% implied). Close was –118 vs +100, no-vig 53.7%. Your CLV is +2.0 percentage points of no-vig probability. This normalizes across price ranges and gives you a clean number that's directly comparable bet-to-bet.
Quick read on what's good
An average CLV of +1.5% across a full season is excellent. +3% is professional-tier territory. If your average CLV is consistently negative, you are paying the book to bet — your record might be net positive in the short run thanks to variance, but the math will catch up.
4. Why CLV predicts profitability better than ROI
The argument for CLV-as-skill-metric has been validated repeatedly: bettors with positive long-term CLV are the same bettors with positive long-term ROI. Bettors with negative CLV who happen to be ahead are running good — their ROI will revert. CLV converges on its true mean far faster than win-loss record does, because every bet contributes a signal regardless of whether it wins.
| Metric | What it measures | Signal vs Noise |
|---|---|---|
| Win/Loss record | What happened | Mostly noise short-term |
| ROI | What happened, weighted by stake | Still mostly noise short-term |
| CLV | Whether your process beat the market | Signal-rich quickly |
| EV from your model | What your process predicted | Strong, but requires honest model |
Consider two bettors after 50 NFL bets:
- Bettor A went 28-22 (56%, +$580 at $100/bet). Their average CLV was −1.2%.
- Bettor B went 22-28 (44%, −$1,420 at $100/bet). Their average CLV was +1.8%.
Almost every recreational bettor would assume A is sharper. The CLV math says the opposite: A ran extremely hot on bets that were, on average, slightly below market value. B ran very cold on bets that were, on average, well above market value. Project both forward 500 bets and B will almost certainly be the profitable one.
CLV is the single best reason to keep betting through a losing month. If your weekly ROI is bad but your CLV is consistently positive, the process is sharp and the results are noise. Eventually the noise washes out and the process wins.
5. Sample size: when to trust your CLV
CLV is signal-rich, but it's still subject to sample size. A handful of CLV-positive bets after the line moved on a single injury report doesn't mean you're sharp — it means you happened to bet before some news broke. To distinguish skill from luck on CLV, you generally want:
This is dramatically faster than the 1,000+ bets ROI needs to stabilize. That's the whole appeal: you find out whether your process works within roughly one season, instead of three.
One caveat: CLV from in-play markets, exotic props, or extremely soft books is less reliable. The closing line at those markets isn't as sharp, so beating it doesn't carry the same predictive weight as beating the close on a heavily-bet NFL side.
6. Why winning bettors get limited
Books track CLV as obsessively as you should. They have models that score every customer's bet quality in real time. A bettor whose average CLV exceeds some threshold — typically around +1% over a meaningful sample — gets flagged.
Flagging escalates in stages. First it might mean reduced max-bet limits on certain markets. Then maybe restrictions on promos and bonuses. Then maybe slow grading or delayed payouts. Eventually the book may make it functionally impossible for you to place meaningful action there at all.
The single best metric proving you're a winning bettor is also the single most reliable trigger for getting your action limited. Books don't need to see you win — they just need to see your bets consistently beating the close. Sharp before the result is enough.
This is the cost of doing business as a winning bettor in the modern US market. Plan for it: have multiple books, vary your patterns where possible, expect that your "main" book may not stay your main book forever.
7. How to actually track CLV
Tracking CLV requires logging two prices per bet: the price you got and the closing price. The closing price is the last price displayed before the market closes — typically 30–60 seconds before kickoff for most sportsbooks.
A minimal tracking workflow:
- Log the bet at placement time. Note: sport, matchup, side, line, price (your side), implied probability, stake.
- Log the closing two-sided line. Right before kickoff, capture both sides of the closing market. A free odds-aggregator or a screenshot of your book works fine.
- Compute no-vig closing probability for your side. Use the vig calculator or do it manually: yourSide% / (yourSide% + otherSide%).
- Compute CLV. Subtract your implied probability from the no-vig closing probability. Positive = beat the close, negative = trailed it.
- Average across all bets. Your aggregate CLV is your skill grade. Look at it monthly, not weekly — small samples are still noisy.
If you do nothing else to upgrade your process, track CLV. It is the only metric that will tell you, within a season, whether your betting is skill or theater. Everything else can be explained by variance.