The question
"I'm down this month, but I keep getting bigger prices than the market closes at. Am I beating the market, whatever the results say?"
With closing line value explained, you have a way to answer that without waiting years. The closing price on Betfair is the market's final, best-informed view. Beating it consistently is the clearest early sign that your edge is real.
The idea in one sentence
Closing line value (CLV) is how much bigger the price you took was than the final price just before kick-off, and on average it tells you whether you're getting value.
The picture
A Betfair football price keeps moving until the off. Team news, late money and the sharpest bettors in the world all push it about. By kick-off it has soaked up almost everything anyone knows. That final price is the closing line.
If you backed at 2.20 and it closed at 2.00, you got a price the market later decided was too big. Do that on average across hundreds of bets, and you're very likely ahead of the market.
The beauty of CLV is that every bet counts, win or lose. A result is a coin flip weighted by the odds, which swings wildly. The gap between your price and the closing price is a small, steady number. Its average settles in a few hundred bets, while profit takes thousands (Lesson 1.5).
On Betfair the closing prices are already close to fair: the back and lay books sit within a fraction of a percent of 100%. Use the last traded price before the off, or the price where the best back and lay meet. For bookmaker closing prices, remove the margin first.
| # | Taken | Close | CLV | EV |
|---|---|---|---|---|
| 1 | 2.20 | 2.00 | +10.00% | +8.80% |
| 2 | 3.50 | 3.60 | −2.78% | −4.17% |
| 3 | 1.95 | 1.85 | +5.41% | +4.38% |
| 4 | 4.00 | 3.50 | +14.29% | +12.57% |
| 5 | 2.50 | 2.44 | +2.46% | +1.23% |
| 6 | 1.80 | 1.83 | −1.64% | −2.51% |
Worked Betfair example
Six back bets in football markets, each compared with the Betfair closing price at kick-off. (Illustrative figures.)
| Bet | Market | Taken | Closing | CLV | Close implied | EV if close is fair (after 2%) |
|---|---|---|---|---|---|---|
| 1 | Match Odds, home | 2.20 | 2.00 | +10.00% | 50.00% | +8.80% |
| 2 | Match Odds, draw | 3.50 | 3.60 | −2.78% | 27.78% | −4.17% |
| 3 | Over 2.5 goals | 1.95 | 1.85 | +5.41% | 54.05% | +4.38% |
| 4 | Match Odds, away | 4.00 | 3.50 | +14.29% | 28.57% | +12.57% |
| 5 | Under 2.5 goals | 2.50 | 2.44 | +2.46% | 40.98% | +1.23% |
| 6 | First Half Over 0.5 | 1.80 | 1.83 | −1.64% | 54.64% | −2.51% |
- CLV of one bet. Bet 1: 2.20 ÷ 2.00 − 1 = +10.00%. You got a price 10% bigger than the market's final view.
- What that's worth. If the close is fair, the home side wins 50% of the time. Your net odds at 2.20 are 1 + 1.20 × 0.98 = 2.176. Expected value = 0.50 × 2.176 − 1 = +8.80% per £1.
- A bet that lost value. Bet 2: 3.50 ÷ 3.60 − 1 = −2.78%. The draw drifted from 3.50 to 3.60 after you backed it, so by the off the market thought it was less likely than your price said. You took a smaller price than the close, and if the close is fair the bet was worth −4.17%.
- Average CLV. Add the six CLVs and divide by six: +4.62%. Four of the six beat the close.
- Average expected value. Averaging the last column gives +3.38% per £1 after commission, if the closing prices are fair.
- How sure can you be? The six CLVs have a standard deviation of 6.66%. The standard error is 6.66% ÷ √6 = 2.72%, and t = 4.62 ÷ 2.72 ≈ 1.70. Promising, but six bets is nowhere near enough. Keep going.
Why CLV gets there so much faster
Take 300 bets with an average CLV of +2.5% and a CLV standard deviation of 7% per bet (illustrative).
- CLV evidence. SE = 7% ÷ √300 ≈ 0.40%, so t = 2.5 ÷ 0.40 ≈ 6.2. Overwhelming.
- P&L evidence for the same 300 bets at average odds of 2.30 and +2.5% ROI after commission: one bet swings by about 1.13 per £1, so t ≈ 0.38, a p-value of about 0.35. Meaningless.
Same bets, same edge. CLV sees it clearly; the P&L can't see it at all yet.
The formula
CLV of a back bet
- O taken is the decimal price you backed at.
- O close is the Betfair price at the off.
In plain English: how much bigger your price was than the final price. Positive is good.
CLV of a lay bet
In plain English: a layer wants the price to drift after laying, so the fraction flips. Laying at 3.40 and seeing it close at 3.60 is +5.9%.
Expected value implied by the close
- c is commission, 0.02 on Betfair.
- 1 ÷ O close is the closing implied probability, treated as the true chance.
In plain English: if the closing price is right, this is what each bet was worth after commission. It's a little below the raw CLV because commission trims the winnings.
Is your average CLV real?
- CLV bar is your average CLV.
- s is the standard deviation of your CLVs.
- n is the number of bets.
In plain English: the same t-statistic as Lesson 1.1, applied to CLV instead of profit. Because s is small, t climbs quickly.
Try it
Paste in the six bets from the table and check the tracker shows an average CLV of +4.62%, four bets beating the close, and t ≈ 1.70. Then add a few more rows with small positive CLV and watch the running average settle.
Common mistakes
- Using the wrong closing price. Take the Betfair price at the off, not a price from the morning or a bookmaker's raw price with its margin still on.
- Judging CLV on a handful of bets. Six bets gave t = 1.70. You still need a few hundred to be confident.
- Mixing up backs and lays. For a lay, you want the price to drift after you've laid it. Flip the formula.
- Ignoring CLV because profit looks fine. Profit with negative CLV is usually luck, and it tends to run out. CLV over P&L.
- Assuming the close is perfect. It's the best estimate available, not the truth. In less efficient markets the close can be wrong too, but beating it is still the best early sign you have.
Why price beats results as the measure: why good models still lose money.