StrategyUpdated 21 Jul 20265 min read
Closing Line Value
The closing line is the market's best estimate of true probability. If you keep taking prices better than the close, you are consistently disagreeing with the market and being right — and that shows up long before profit could.
Every betting method needs a scoreboard. The obvious one, profit, is almost useless over any period a human is willing to wait. CLV is the alternative, and it is the reason Tofiko grades itself the way it does.
What the closing line is#
The closing line is the last price a market offers before kick-off. Prices move between opening and closing as money and news arrive: an opener of 2.40 might drift to 2.60 or shorten to 2.10 by the time the whistle goes.
CLV compares your price to that final one. Back a side at 2.20 that closes at 2.00 and you have raw CLV of +10%, because 2.20 ÷ 2.00 - 1 = 0.10. You took a price 10% better than the market's final word. Whether the bet won is a separate question, and a noisier one.
Why the close is sharp#
Two forces make the closing price hard to beat.
Volume. The overwhelming majority of money on a football match arrives in the final hours. Prices that are wrong at that point get hit until they aren't. An opening line is one trader's opinion; a closing line is everybody's, weighted by how much they were prepared to risk on it.
Late information. Confirmed lineups, late injury news, the pitch, the weather — all of it lands before the close and is priced immediately. Anything you knew at 9am that mattered is in the number by 3pm.
That's why an opening-line comparison flatters you and a closing-line comparison doesn't. Beating an opener means you were faster. Beating the close means you were right.
Raw CLV versus margin-adjusted CLV#
Here is the part almost every tipster's CLV claim quietly skips. The closing price is not the market's fair estimate — it's the fair estimate with the bookmaker's margin baked in. Beating a margin-loaded price by less than the margin means you are still betting below fair value.
Work it through. Take a book whose closing prices sum to 105% of probability — a 5% margin. Your side closes at 2.00, which implies 50%. Strip the margin proportionally and the fair probability is 0.50 ÷ 1.05 = 47.62%, so the fair price is 1 ÷ 0.4762 = 2.10.
Beating a 5% book's close by 5% is exactly break-even. Not a small edge — none. Everything below that line is a loss dressed as a win, and every claim of "we beat the close on 60% of bets" that doesn't state the margin is uninterpretable.
Put your own bets through the CLV calculator. Enter the price you took, the closing price, and the margin the closing book carries — then read the margin-adjusted number rather than the raw one.
Stripping 5.0% of margin out, the closing market rated your pick at 47.6% — fair odds of 2.10. You took a bigger price than that, so the bet carried real expected value at the time you placed it.
One bet proves nothing. CLV is worth measuring because it converges far faster than profit does — but it still needs hundreds of bets before the average means anything.
Dividing by the overround, as above, spreads the margin evenly across every outcome. Real books load more of it onto longshots, so the even split slightly overstates fair prices on favourites and understates them on outsiders. Tofiko removes margin with the power method instead, which fits the distortion rather than assuming it away — the fair odds guide covers the difference.
Why ROI is a worse scoreboard#
ROI has one fatal property: the signal is tiny next to the noise. A 2% edge over 500 flat even-money bets is worth 10 units of profit, while the standard deviation of that same 500-bet total is about 22 units. The thing you're measuring is half the size of the measurement error. You'd need something like 10,000 bets before the profit figure separates from zero at two standard errors, and rather more before you'd reliably detect it at all — the arithmetic is in how many bets you actually need.
CLV has the opposite property. It records a number on every bet, including the ones that haven't been settled and the ones that lost. It doesn't wait for the outcome, so it isn't diluted by it. The same handful of hundreds of bets that tells you nothing about ROI can already tell you whether you're systematically on the right side of the market.
How to record closing prices honestly#
CLV is easy to fake without meaning to. Four rules:
- Same book, same market, same line. Comparing your Asian handicap at one book to a different line at another isn't CLV, it's shopping.
- Log the price at the moment you bet, not the best price you saw that day. Retrospective bests are fiction.
- Capture the close automatically, right before kick-off. Doing it by hand guarantees you'll capture the ones you remember, which are the good ones.
- Keep the losers and the voids. A record with survivors only measures your memory.
What CLV cannot tell you#
It is not a shortcut past sample size. CLV converges faster than ROI, not instantly — you still need hundreds of bets before an average CLV means anything, and the average is dragged around by a few big movers, so look at the distribution too.
It measures the decision, not the outcome. You can beat the close on every bet in a season and finish down; that's variance doing its job, and it doesn't invalidate the CLV. It also says nothing about whether your probabilities are well-shaped — for that you want calibration and the Brier score, which grade the numbers themselves rather than the prices you got.
And it only works where a meaningful close exists. Thin markets that barely trade have no crowd to be wise, which is awkward, because thin markets are exactly where any value would live.
We grade ourselves on CLV, calibration and Brier score — not ROI — because profit converges far too slowly to prove anything. On that scoreboard, Tofiko has no demonstrated edge over closing prices. Our published results are indistinguishable from the closing line to date. We say so here and on the performance page rather than quoting a flattering profit figure from a sample too small to mean anything.
Related
- Sample Size: How Many Bets Before You Know You Have an Edge
- Fair Odds: What the Bookmaker's Margin Hides
- What Is Value Betting? A Plain-English Guide
- The Brier Score: How We Grade Our Own Predictions
- Closing Line Value Calculator — Grade a Bet Without Waiting for the Result
- Fair Odds Calculator — Remove the Bookmaker Margin from Any 1X2 Book
Frequently asked questions
What is closing line value in betting?
Closing line value is the difference between the odds you took and the odds the same bet was available at just before kick-off. If you backed a team at 2.20 and it closed at 2.00, you beat the close. It measures your decision at the moment you made it, independently of whether the bet won.
Why is the closing line considered accurate?
By kick-off the market has absorbed almost all the money and almost all the information — team news, lineups, weather, and every sharp opinion that was willing to back itself. No single participant knows more than that aggregate, which makes the closing price the best cheap estimate of true probability available.
Does beating the closing line mean I will make money?
Not necessarily, and the gap matters. If the closing book carries a 5% margin, you must beat the closing price by about 5% just to reach break-even against fair value. Consistently beating it by more than the margin is evidence of skill, but it still has to survive the commission and limits you actually face.
Is CLV better than ROI for judging a betting method?
For anything short of tens of thousands of bets, yes. ROI is buried in variance at realistic sample sizes, so a season of profit or loss is close to uninformative. CLV records something on every single bet and needs no result at all, so it accumulates evidence far faster.