Win rate will gaslight you. CLV won’t.
If you’ve ever gone 7-3 over a weekend and felt like a genius… then gave it all back the next week, welcome to sports betting. Win rate is a noisy, short-run scoreboard. It tells you what happened, not whether you had an edge.
Closing Line Value (CLV) is different. CLV asks one simple question: did you beat the market price that mattered most? If you consistently bet a number that closes shorter than what you took, you’re doing something right. Not always. Not automatically. But in the long run, that’s the closest thing you’ll get to a skill metric that isn’t just variance wearing a disguise.
Here’s why this matters: sportsbooks don’t pay you for being “right.” They pay you for buying mispriced probabilities. That’s it. If you bought 55% when the book was selling 50%, you can lose today and still be printing long-term. If you bought 48% when the book was selling 55%, you can win today and still be lighting money on fire.
And the market moves a lot. Right now, across 6,585 tracked movements, the average movement is 21.49%. That’s not a rounding error. That’s the difference between “you found value” and “you donated.”
This post makes the case—cleanly, with math—why CLV beats win rate as a performance metric. Then I’ll give you a dead-simple daily workflow for tracking it, interpreting it, and catching the moments where “good CLV” is actually misleading (limits, stale books, low-liquidity markets).
CLV in plain English: you’re grading your bet, not the result
CLV is the gap between the price you bet and the price that closes.
Example in decimal odds (because it’s cleaner math):
- You bet Team A at 2.10.
- The market closes Team A at 1.91.
You beat the close. That’s positive CLV. You bought a bigger payout than the market eventually offered.
To make CLV comparable across bets, convert odds to implied probability:
- Implied probability = 1 / decimal odds
- Your bet at 2.10 implies 1/2.10 = 0.4762 (47.62%)
- Closing price at 1.91 implies 1/1.91 = 0.5236 (52.36%)
Your CLV (probability form) is roughly:
CLV% ≈ (Close Prob − Bet Prob) = 52.36% − 47.62% = +4.74 percentage points.
That’s a huge edge if it’s real. If you could repeatedly buy 47.6% outcomes that the market later prices like 52.4% outcomes, you’d be profitable even with a mediocre-looking win rate over small samples.
If you’re more comfortable in American odds, fine—just don’t mix formats and make mistakes. If you still fumble conversions, read Decimal vs American Odds: Convert Fast (and Stop Mispricing Bets). Most bettors leak EV just by misreading prices.
Important: CLV is not “line moved in my direction so I’m awesome.” It’s “my entry price was better than the best available consensus at close.” That’s the whole game.
The math: why win rate is a terrible skill metric (short-term)
Win rate ignores price. That’s the sin.
If you bet ten coin flips at -110 (about 52.38% break-even), you can easily go 7-3. That doesn’t mean you have an edge. It means variance smiled at you for ten trials.
Let’s put numbers on it. Suppose you bet two different styles:
- Bettor A takes mostly favorites at 1.40 (implied 71.43%). They go 7-3 (70% win rate). Sounds great.
- Bettor B takes dogs at 2.50 (implied 40%). They go 5-5 (50% win rate). Sounds worse.
But profitability depends on expected value (EV), not win rate.
Bettor A: if those 1.40 lines were actually fair (no edge), their expected win rate is 71.43%. Going 70% is actually below expectation. They’re likely losing to vig long-term even if they “win a lot.”
Bettor B: if those 2.50 lines should have been 2.30 (implied 43.48%) and they consistently got 2.50, they can be +EV while “only” winning 40–45% long-term. A 50% week is just variance in their favor.
This is why recreational bettors get crushed. They chase win rate. They brag about win rate. They tilt when win rate dips. Meanwhile, the only thing that matters is whether you bought good numbers.
CLV isn’t perfect, but it’s aligned with the math. Markets (especially liquid ones) incorporate information over time. If your bets consistently beat the close, you’re probably finding misprices before the rest of the money does.
And again, movement isn’t rare. Across 6,585 movements, the average move sits at 21.49%. If you’re routinely taking the worse side of moves like that, your win rate has to be unreal just to break even. Good luck with that.
How to calculate CLV the right way (with real-number examples)
You can track CLV in a few ways, but two are practical:
- Price CLV: compare your odds to closing odds.
- Probability CLV: compare implied probabilities (my preference).
Probability CLV is better because odds aren’t linear. The difference between 2.00 and 1.90 isn’t the same “edge” as 10.00 and 9.90.
Let’s use a real movement example you’ve probably seen in the wild: a massive drift (odds get bigger).
One recent move: Leeds United vs Brentford (EPL h2h). Brentford went from 5.0 to 10.0 at TAB. That’s a 100% move.
If you bet Brentford at 5.0 and it closed 10.0, that’s negative CLV. You beat nothing—you got the early, expensive price.
Convert to implied probability:
- 5.0 implies 1/5.0 = 20.00%
- 10.0 implies 1/10.0 = 10.00%
Probability CLV = 10.00% − 20.00% = -10.00 percentage points. That’s brutal.
Another example: Phoenix Suns vs Milwaukee Bucks (NBA h2h). Milwaukee moved from 2.0 to 4.0 at ESPN BET (another 100% move).
- 2.0 implies 50.00%
- 4.0 implies 25.00%
CLV = 25% − 50% = -25 percentage points. If you repeatedly buy 50% prices that close like 25% prices, you’re not unlucky—you’re mispriced.
Flip it around for positive CLV: if you bet 4.0 and it closes 2.0, you just bought a 25% implied price that the market later treated like 50%. That doesn’t guarantee a win. It guarantees you got a hell of a number.
One more important detail: use the same market type and rules. h2h vs draw-no-bet vs Asian handicap aren’t interchangeable. CLV tracking gets sloppy fast if you mix apples and chainsaws.
Your daily CLV tracking workflow (15 minutes, no nonsense)
You don’t need a PhD or a custom database. You need consistency.
Step 1: Log every bet immediately
- Date/time
- Sport/league/market
- Book
- Selection
- Odds you took
- Stake
If you want to be serious, log the timestamp. CLV depends on when you bet. A number at 9:00am and a number at 6:00pm aren’t the same opportunity.
Step 2: Pick a “closing reference” that isn’t your own book
This is where most people screw it up. If you grade your CLV against the same soft book you bet, you can “win CLV” while still betting into garbage prices.
You want a sharper reference: a market consensus or sharper books. If you’re using ThunderBet tools, Edge Finder fits naturally here because it lets you benchmark your bet price against stronger reference prices. That benchmark becomes your CLV yardstick.
Step 3: Record the closing price (or close probability)
Do this once per day for everything that closes that day. If the market closes at different numbers across books, use a consistent method:
- Take a sharpest-available close
- Or take a consensus average of sharp books
- Or use an exchange reference
When books disagree late (limits distortions happen), an exchange can be a cleaner “true market” snapshot. That’s exactly when something like the Exchange Terminal becomes useful: it gives you another reference point when sportsbook closes get weird.
Step 4: Calculate CLV for each bet
Use implied probability:
- Bet Prob = 1 / Bet Odds
- Close Prob = 1 / Close Odds
- CLV (pp) = Close Prob − Bet Prob
Step 5: Track a weighted average CLV
Weight by stake. A $10 flyer shouldn’t count the same as your $500 position.
Weighted CLV = Σ(stake × CLV) / Σ(stake)
Step 6: Review weekly, not emotionally
Daily CLV tracking keeps you honest. Weekly review keeps you sane. Win rate will mess with your head day-to-day. CLV gives you a calmer signal about whether your process is working.
When “good CLV” lies to you (limits, stale books, low liquidity)
CLV is a strong metric in liquid markets. It gets sketchy when the market is thin or when your “closing” number doesn’t mean what you think it means.
Three common traps:
1) Limits make the close fake
If a book drops limits late, the “close” might reflect tiny money, not true opinion. You can beat that number and still not have a real edge. You’re basically grading yourself against a price nobody could bet into at size.
2) Stale books hand you CLV on a platter
You can pick off slow-moving books and rack up beautiful CLV… until they limit you or the opportunity disappears. That CLV is still real in a sense (you got a better number), but it’s not always scalable or repeatable.
This is exactly why you should separate your CLV report by bookmaker. If all your positive CLV comes from one sleepy shop, you didn’t “solve betting.” You found a temporary leak.
3) Low-liquidity markets create clown moves
Some markets move like crazy because it doesn’t take much money to shove them around. You’ve seen it: 100% moves where odds double.
Just look at a few recent examples:
- CA Osasuna vs Girona (La Liga h2h) moved from 7.5 to 15.0 at FanDuel.
- Columbus Blue Jackets vs Seattle Kraken (NHL h2h) moved from 23.0 to 46.0 at LeoVegas (SE).
- Philadelphia Union vs Chicago Fire (MLS h2h) moved from 2.05 to 4.1 at Bet Victor.
Those are massive. And they’re a warning label: in thin markets, closing isn’t always “sharp.” Sometimes it’s just “last.”
How you protect yourself: tag bets by market liquidity (main lines vs alt lines vs niche props) and judge CLV expectations accordingly. Also, don’t fall in love with CLV that only shows up in weird, low-limit corners.
How to interpret your CLV report like a pro (and fix leaks fast)
Once you track CLV daily, you’ll start seeing patterns. Good. That’s where the money is.
1) Break it down by sport
Different sports behave differently because information and liquidity differ. Right now, movement volume is heavy in NHL (1,768) and NBA (1,185), with MLS (919) and NCAAB (824) also active. Soccer leagues like EPL (479) and La Liga (444) sit right there too. Translation: you’re not betting in a static environment. If you’re consistently late in fast-moving sports, CLV will expose you.
2) Split by market type
h2h, totals, spreads, props—don’t lump them. You might be great at totals and awful at moneylines. If you’ve read our work on where totals get hit early, connect the dots: 5,055 Moves: Where NBA/NHL Totals Get Hit First. CLV improves when you stop showing up after the market already ate.
3) Watch CLV vs ROI divergence
Over small samples, you’ll see:
- Positive CLV, negative ROI (you’re running bad)
- Negative CLV, positive ROI (you’re running hot)
The second one is the dangerous one. That’s when bettors “feel it” and press. Hell, that’s when they start parlaying to celebrate. If you need a reminder why parlays usually bleed EV, read Build Parlays That Don’t Leak EV: 3 Checks in ThunderBet.
4) Set realistic CLV expectations
You don’t need monster CLV on every bet. In efficient markets, beating the close by even a small amount repeatedly is meaningful. What you want is consistency and clean sourcing (real market closes, not your own book’s last-second number).
5) Use “CLV flags” to spot process issues
- If your CLV is negative mostly on favorites, you’re probably paying tax (betting popular sides too late).
- If your CLV is positive only on one book, you’re relying on stale lines.
- If your CLV swings wildly, your “close” source is inconsistent or you’re betting thin markets that whip around.
CLV isn’t a trophy. It’s a diagnostic tool. Treat it like one, and you’ll fix leaks before they show up in your bankroll graph.
If betting stops being fun or you feel out of control, take a break and set hard limits. Only wager what you can afford to lose.
