Closing line value (CLV) measures how the price you took compares with the final price before the event started: CLV = your odds / closing odds - 1. With illustrative prices, take $1.90 on a team that starts at $1.73 and you beat the close by 1.90 / 1.73 - 1 = 9.8%. Take $1.90 on a team that starts at $2.05 and your CLV is 1.90 / 2.05 - 1 = -7.3%.
CLV is worth tracking because it answers, in far fewer bets, the question profit takes thousands of bets to settle: are my prices better than the market's? For sport, take the bookmaker's margin out of the close first, and that 9.8% shrinks to about 4.4%. A good CLV figure grades the price you took; it pays nothing by itself.
How to calculate closing line value
Three steps for each bet:
- Record the price your bet was accepted at.
- Record the closing price for the same outcome, in the same market, on the same line: the last price before the game starts or the race jumps.
- Divide the first by the second and subtract 1.
With illustrative bets and the bookmakers' raw closing prices:
| Bet | Price taken | Closing price | CLV |
|---|---|---|---|
| Team A, head to head | $1.90 | $1.73 | +9.8% |
| Team C, line at +7.5 points | $1.92 | $1.85 | +3.8% |
| Over in a total points market | $1.87 | $1.95 | -4.1% |
| Player to score the first try | $9.00 | $9.00 | 0.0% |
The working for the third row: 1.87 / 1.95 - 1 = -0.041. The average of the four, (9.8 + 3.8 - 4.1 + 0.0) / 4 = +2.4%, matters more than any one row.
Some trackers quote CLV as a gap in implied probability instead. For the first bet, 1 / 1.73 - 1 / 1.90 = 57.8% - 52.6% = 5.2 percentage points. Both forms agree on whether you beat the close, but they are different measures, not one figure in two units; the ratio form lines up directly with edge and expected value, so it is the one used here.
Why the closing price is the market's best estimate
By the start, a market has taken in the most news it will get: team lists and late changes, weather, the ground or track, and in racing the scratchings. It has also had the most money bet into it, and bookmakers have had the longest to move their prices towards that money and news. No earlier price folds in as much.
Australian law gives sport a clean close. Under the Interactive Gambling Act 2001, online betting on a sporting event is lawful only through a provider licensed in an Australian state or territory (s 15AA). The bet must also be placed, made, received and accepted before the event begins (s 8A(3), read with s 10B), so the last price before the start is the close.
The close is still an estimate, not the truth. A close formed on little money is a weak one, a close can be wrong about a single game, and a bookmaker's close still has its margin inside it, which the next step removes.
Taking the margin out of the close for sport
A bookmaker's closing prices add up to more than 100%, so measuring against them flatters every bet. The simplest fix, the proportional method, scales each closing price by the market total:
margin-free closing price = closing price x market total
With illustrative closing markets:
| Closing market | Closing prices | Market total | Your bet | CLV against the raw close | CLV against the margin-free close |
|---|---|---|---|---|---|
| Head to head, two outcomes | $1.73 and $2.11 | 57.8% + 47.4% = 105.2% | $1.90 on Team A | 1.90 / 1.73 - 1 = +9.8% | 1.90 / (1.73 x 1.052) - 1 = 1.90 / 1.82 - 1 = +4.4% |
| Home, draw, away | $2.00, $4.00 and $3.20 | 50.0% + 25.0% + 31.25% = 106.25% | $3.50 on the away team | 3.50 / 3.20 - 1 = +9.4% | 3.50 / (3.20 x 1.0625) - 1 = 3.50 / 3.40 - 1 = +2.9% |
The last two columns differ by the market total, as a ratio: 1.098 / 1.044 = 1.052 for the first market. A bet at exactly the raw close of the first market scores 0% against it, and 1 / 1.052 - 1 = -4.9% against the margin-free close. Under the proportional method, beating a bookmaker's raw close by less than its margin is not beating the market at all. Methods that load more of the margin onto long shots take less off a favourite, so check the sum with the method you use.
Two choices shape the result:
- Which close. Use the deepest market you can: the exchange where plenty of money has traded, or several bookmakers' margin-free closes combined by averaging their chances, then taking 1 / that average. One bookmaker's last price is a single opinion.
- Which method. Other de-vig methods shift a bigger share of the margin onto long shots, so the choice changes CLV most at long prices. How to remove the bookmaker margin compares them; pick one and use it on every bet.
What beating the closing line tells you
Take the margin-free close as the market's final estimate of the chance. On that estimate, the amount a bet beat the close by is its edge, and a record's average CLV is the yield it should expect. That is the link between CLV and value betting: a value bet is a price above fair, and the close is the market's last word on fair.
Beating the closing line on most bets is not the test; the size of the average is. With illustrative figures, six bets that beat the close by 2% and four that miss it by 6% average (6 x 2 - 4 x 6) / 10 = -1.2%. Most of those bets beat the close, and the record still trails it.
The reading works in both directions. A losing run with positive CLV points at luck; a winning run with negative CLV points at luck too, the kind that runs out. Within a wider betting strategy, CLV is the check that keeps either story honest.
Risk: Betting involves risk. Beating the close is a verdict on prices already taken, silent on whether the next bet wins, and a close is itself an estimate. Past results are no guarantee of future results; see responsible gambling for limits and support.
CLV shows an edge sooner than profit
Each bet's profit lands at one of two far-apart points, minus the stake or plus (odds - 1) times it, while its CLV is the small gap between two nearby prices. The smaller the swing per bet, the sooner an average settles.
Example: Illustrative: bets at an average price of $2.50, a true edge of 3%, and a bet-to-bet standard deviation in CLV of 7 percentage points. Work out your own from your record; 7 is only for illustration.
| Bets | One standard error of yield | One standard error of average CLV |
|---|---|---|
| 50 | 17.3 points | 0.99 points |
| 200 | 8.7 points | 0.49 points |
| 1,000 | 3.9 points | 0.22 points |
The working for 200 bets: standard error of yield = about square root of (2.50 - 1) / square root of 200 = 1.2247 / 14.142 = 0.0866, or 8.66 points. Standard error of CLV = 7 / square root of 200 = 0.49 points. After 200 bets, a 3% average CLV sits 3 / 0.49 = 6.1 standard errors above zero, while a 3% yield sits 3 / 8.66 = 0.35 of one.
How many bets a record needs before either figure means something is worked through in sample size in betting. CLV answers a narrower question, though: whether your prices beat the market's final estimate, not whether that estimate was right.
Risk: Betting involves risk. A clear CLV signal after a few dozen bets says nothing about your balance, which can keep falling for hundreds of bets at a real edge. See responsible gambling for limits and support.
Why racing uses a different close
Racing has a ready-made close. The Betfair Starting Price (BSP) is struck on the exchange at the jump from the money waiting on each side, with no bookmaker margin built in. That leaves CLV = price taken / BSP - 1 needing little de-vigging. Exchange commission, place bets, late scratching deductions and a bookmaker's own starting price each change the sum, and closing line value in horse racing works through them.
Edge cases that change the sum
| Situation | How to handle it |
|---|---|
| The line or total moved after you bet | Compare like with like: the closing price for your exact line, where bookmakers still offered it. Say, with illustrative lines, you took +7.5 points and the line closed at +4.5: you beat the close by 3 points, a gain measured in points rather than price |
| The bet was voided or the game abandoned | Leave it out of your yield, and keep its CLV in a separate tally |
| The price came with an odds boost or other promotion | Work out CLV on the price before the boost as well: a boost lifts the price above the bookmaker's own by design, which inflates CLV and says nothing about your judgement |
| A multi across separate games | Multiply the legs: (1 + CLV of leg 1) x (1 + CLV of leg 2) - 1, so illustrative legs at +5% and +3% give 1.05 x 1.03 - 1 = +8.15% |
| A same game multi | The legs are linked and the bookmaker prices that link, so CLV cannot be built from the legs; you need the bookmaker's own price for the same combination at the start |
| A late team change after you bet | Count it. The market moved on news you did not have, and over many bets that is part of what your timing costs or earns |
Mistakes that make CLV misleading
| Mistake | What it does to the figure |
|---|---|
| Measuring against the raw close | Overstates every bet by about the margin: +9.8% instead of +4.4% for the first bet above |
| Choosing the close after the event | Picking the bookmaker whose close flatters your bet turns a check into a story |
| Trusting a thin close | A close formed on little money is a weak benchmark, so give those bets less weight |
| Mixing in bonus bets and boosted prices | Inflates the average with value that came from the promotion, not your prices |
| Treating CLV as profit | It is evidence about prices; results still swing, and the money comes only from results |
| Recording only some bets | A CLV record with the misses left out is as worthless as a profit record with the losers left out |
Checking CLV with B337
From Terminal View up, the Terminal's racing board shows closing lines, the last prices before the jump, with each bookmaker's flucs and Betfair back and lay, so you can check a racing price you took against where the market closed. For a sports bet, note the last price before the start yourself. The Terminal reads prices in cycles rather than every tick, so what is on screen can trail the bookmaker's own price.
The +EV screener lists sports prices above a margin-free fair price, ranked by edge, and that fair price is an estimate. Its rows are part of Terminal Pro and Full Automation, or come on their own as a Screeners plan arranged with the team, while a free account shows the live counts with the rows locked. Bets placed through B337 use credits.
On copy betting, each tipster feed shows a tracked closing-line value beside its strike rate, ROI and turnover, worked out by B337 from the tips published through the platform and how they settled. A tracked record is not an audit, a rating or a recommendation. Betfair is a trade mark of its owner, and B337 is not affiliated with it.
For free and confidential support call 1800 858 858 or visit gamblinghelponline.org.au.
Risk: Betting involves risk. A positive CLV record grades your prices and can sit beside a losing balance, closing lines on any screen can trail the bookmaker, and bookmakers can restrict accounts and void bets. Past results are no guarantee of future results; see responsible gambling for limits and support.