The first time I properly tracked my closing line value I discovered I'd been profitable for six months by accident. My win rate looked healthy. My bank balance looked healthy. My CLV was negative 1.2%. I'd been betting into unfavourable market moves for half a year and hadn't noticed because nothing had caught fire yet. Six weeks later it did, and I gave back most of the profit in one very ugly month.
That experience changed how I think about measuring a betting process. Profit is a noisy, late signal. CLV is an earlier, quieter one, and it happens to be the metric that actually predicts whether the next 500 bets will go well.
What CLV is, exactly
Closing line value is the gap between the price you got on a bet and the price the same market closed at. If you back Collingwood at $2.00 and the market closes at $1.85 right before kickoff, your CLV on that bet is +8.1%. If you bet at $2.00 and the market drifts out to $2.20, your CLV is -9.1%.
The reason the number matters is that closing prices are the most accurate outcome predictor anyone has ever built. Better than any pundit. Better than any model. Better than any AI. By the time a bookmaker market closes, it has absorbed injury news, weather, lineups, and every dollar of sharp money that flowed in during the week. The close is as close to the real probability as you are going to get from any public source.
So if you consistently bet at prices better than the close, you are consistently betting above market. You are +EV by definition, regardless of whether this particular month's results look pretty.
Why win rate lies to you
Sports betting variance is brutal at small samples. Say your real edge is 3% at decimal odds of $2.00. Long-run, you win 51.5% of the time. Over 100 bets, your observed win rate can land anywhere from about 42% to 61% by pure chance. Over 500 bets, it's still 47% to 56%. You need something like 2,000 placements before raw win rate starts to tell you anything you can actually trust.
Most punters don't have 2,000 bets of clean data. They have 150 and a feeling. That's fine for someone betting weekends for fun. It is nearly useless for anyone trying to work out whether their process has an actual edge.
CLV converges much faster. After 50 to 100 bets you can usually see whether your price-taking is beating the close. The signal isn't perfect. A run of markets that barely move will dampen it. But it gets you to a reliable read in a fraction of the sample size you'd need from results alone.
Calculating CLV
The formula is unglamorous: CLV% = (your_odds / closing_odds − 1) × 100.
Bet at $2.10, market closes at $1.95, CLV is (2.10 / 1.95 − 1) × 100 = +7.7%. You beat the close by 7.7%.
One thing to watch: record the closing price at the same bookmaker where you placed the bet. Using the best close across every AU book over-rewards the bets you took at slow bookmakers and punishes the ones you took at sharp ones. That defeats the point. The market you want to beat is the one you actually bet into.
The no-vig version
A slightly sharper calculation strips the bookmaker margin out of the closing prices before comparing. You convert each side of the closing market to an implied probability, normalise them so they sum to 100%, and compare your odds to the resulting fair odds.
In practice the no-vig version gives tidier numbers but the relative ranking of your bets almost never changes. Raw CLV is fine for almost everyone. The only time it matters is if you are trying to compare performance across sports with very different vig levels, or across AU bookmakers with structurally different margin policies.
What good CLV actually looks like
Rough benchmarks, from my own tracking and from other AU punters I compare notes with:
- +5% or better average CLV: something in your process is genuinely beating the market. Keep doing what you're doing and start worrying about account longevity more than edge.
- +2% to +5%: a legitimate edge. Probably from slow bookmakers, promo boosts, or a narrow prop niche. Real, but not elite.
- 0% to +2%: could be skill, could be noise. You need more sample to know, and in the meantime you shouldn't bet like you've got a confirmed edge.
- Negative CLV: you are betting into moves. Stop, audit, figure out which categories of bets are dragging you down before you lose more money.
Numbers get strange at the extremes. Someone running +15% CLV is almost always exploiting a narrow edge, like a specific player prop market on a slow bookmaker, and it tends to scale badly. The second the edge closes, the CLV collapses.
Where CLV misleads you
CLV is the best single metric most punters have, but it is not magic. A few honest limitations.
Dead markets and flat closes
If you bet into a market that barely moves — a late-night A-League fixture, an obscure NBL game, a midweek NBA game between two tanking teams — the close might sit within cents of where you bet. Your CLV reads as roughly 0% whether you were right or wrong. That's noise, not signal.
The way around this is to look at CLV grouped by market type. A bettor with +4% CLV overall but 0% on late-night A-League isn't making money on those bets. They're making money somewhere else and diluting the average by betting into dead markets. That's worth knowing.
Promo-boost illusions
Sportsbet boosts an AFL match from $1.85 to $2.20 as a money-back special, you take it, your CLV on that bet is around +19%. But the whole thing is capped at a $50 stake and you can do it once. Tracking CLV on a boost tells you the boost was good. It doesn't tell you you're sharp.
The fix is to bucket promos separately and look at your CLV on unboosted bets as the cleaner signal.
How to actually track it
Two realistic options.
Option one: a spreadsheet. Columns for date, event, pick, your odds, closing odds, stake, outcome, P&L, and a CLV% column that computes automatically. It works. It costs nothing. The problem in practice is that you have to remember to come back and log the close, which means checking markets right before every kickoff, which means you forget. My spreadsheet had about a 30% fill rate on the closing-line column, which made the average CLV value functionally worthless.
Option two: use the Krok Odds Bet Tracker, which captures the closing line automatically at kickoff and tags every bet with its CLV. I built it after I got tired of not trusting my own data.
Either way works. The one that doesn't work is not tracking and assuming your profit column tells you what you need to know.
CLV vs profit: which one matters
Profit pays the bills. CLV tells you whether the process that produced the profit is repeatable. Those are two different questions.
A bettor up $1,200 over three months with -2% average CLV is headed for a cliff. They've been lucky. The market has been quietly telling them they're on the wrong side of most of their bets, and the variance will catch up.
A bettor down $400 with +4% average CLV is almost certainly fine. The process is sharp. The results will come.
If you're only going to track one thing about your betting, track CLV. It is the earliest honest signal you have. Use the positive-EV guide to understand why the signal works, and the bankroll guide to keep your stake sizing from undoing the edge you're proving.