Expected value and closing line value are the two most important metrics in sports betting. They're often confused or treated as interchangeable. They're not — they measure different things, at different times, using different inputs. Understanding the distinction is essential for any advantage bettor serious about improving their process.
This guide covers what each metric is, what the relationship between them should be, and how to use both in practice.
Expected value (EV): the prediction
Definition: EV is the average long-term return a bet would produce if it could be placed an infinite number of times, based on your estimate of the true probability of the outcome.
Formula: EV% = (true probability × decimal odds − 1) × 100
When you calculate it: before placing the bet.
What it tells you: whether to place the bet. A +EV bet is profitable on average over the long run. A −EV bet loses money on average.
Key limitation: depends entirely on the accuracy of your true-probability estimate. If your estimate is wrong, your EV calculation is wrong.
For the full walkthrough, the EV calculation guide covers the worked examples. The short version: you need a true probability input (usually market consensus from AU bookmakers) and decimal odds. The formula does the rest.
Closing line value (CLV): the verdict
Definition: CLV is the percentage by which the price you got on a bet differs from the closing price of the same market at the same bookmaker.
Formula: CLV% = (your odds / closing odds − 1) × 100
When you calculate it: after the market closes, typically immediately before kickoff.
What it tells you: whether you beat the market's final price. The closing line is the most accurate available estimate of true outcome probability. Beating it consistently means you're betting above market, which is mathematically positive EV.
Key advantage: doesn't require a separate true-probability estimate. The closing line is the probability estimate. This makes CLV much more reliable than EV for verifying process quality.
See the CLV guide for the full walkthrough and the one-year CLV tracking piece for what real CLV data actually looks like over time.
The key difference in one sentence
EV is your prediction of the bet's value at placement time. CLV is the market's verdict on your prediction.
If you're a sharp bettor with accurate probability estimates, the two numbers should approximately match across a large sample of bets. Your average claimed EV at placement should equal your average CLV at close.
What the relationship reveals
The relationship between your tracked EV and CLV is diagnostic. Four patterns matter:
EV and CLV both positive, roughly equal
The healthy pattern. Your average claimed EV is, say, +3% and your average CLV is +2.8%. Your probability estimates are calibrated and the market is roughly agreeing that your picks were above fair value at placement. This is what an advantage bettor's process looks like in practice.
EV positive, CLV negative
The most common problem pattern. You're claiming +4% EV on your bets but your average CLV is −1%. This means your true-probability estimates are systematically over-optimistic — you believe your picks are better than they actually are. The market disagrees, moves away from your position, and the close settles at a price worse than yours.
Action: recalibrate your EV process. Either your true-probability estimates are wrong (most common), or you're betting into moves (your bets are informative to the market, which adjusts before close). Either way, your sizing is too aggressive for your actual edge.
EV near zero, CLV positive
Unusual but real. You're placing bets that seem close to fair at placement, but the market moves in your favour before close. This can indicate early-market timing — you're getting your bets down before the market tightens. The edge exists but your EV calculation is understating it because it's comparing to already-shifted market consensus.
EV negative, CLV negative
You're losing. Stop. Audit your process, identify whether specific categories are dragging the overall numbers, and address them before placing more bets.
Why CLV is often the more reliable metric
EV requires a true-probability estimate, which is the single biggest source of error in advantage betting. CLV requires no such estimate — it directly measures your price against the market's most accurate available estimate (the closing price).
Practical implications:
CLV stabilises faster than EV-implied returns. After 50-100 bets you can get a reasonable read on your average CLV. Getting the same confidence on realised returns or pure EV accuracy takes thousands of bets.
CLV doesn't lie to you about your skill. A bettor who consistently over-estimates EV will show positive realised returns in lucky stretches and convince themselves they have edge. CLV catches the mis-calibration immediately because it compares to the actual market, not to your own potentially flawed numbers.
CLV protects against model-decay. If your EV model was working but has stopped working because the market adjusted, CLV will show the shift before your P&L does. Early warning signals matter.
When to use EV and when to use CLV
Practical workflow:
Before placing a bet: calculate EV. Compare your true-probability estimate to the offered odds. Place the bet only if EV is above your threshold (typically +2% or higher to account for estimation error).
After the market closes: record CLV. Log the closing price on every bet and calculate CLV. Track the rolling average, segmented by bookmaker, sport, market type.
Monthly review: compare average EV to average CLV. These should track closely. Persistent gaps indicate process problems to investigate.
The Krok Odds Bet Tracker automates both — calculating EV at placement and CLV after market close, and aggregating the metrics for review.
Frequently asked questions
Can CLV be more reliable than actual profit?
Yes, at small sample sizes. Variance dominates realised profit at 50-500 bet samples. CLV is much less affected by variance because it compares the price you got to the price that existed at close — independent of the actual match outcome. Pro bettors often trust CLV over short-term P&L for this reason.
What's a good CLV for an advantage bettor?
Sustained +2% to +3% CLV is real edge. +5% or better is elite. Negative CLV is a red flag regardless of short-term results. See the one-year CLV piece for benchmarks from a tracked dataset.
Do CLV and EV always match for sharp bettors?
They should approximately match over large samples. Small deviations (1-2 percentage points) are normal due to the inherent imprecision in both metrics. Large persistent gaps indicate a process issue.
Is CLV only relevant for pre-match bets?
Primarily yes, because CLV is defined against the closing price, which is a pre-match concept. In-play betting doesn't have a clean “closing” reference in the same way. For AU punters, who are mostly constrained to pre-match online betting under the IGA, this isn't a practical limitation.
Can you have positive CLV without positive EV?
Not in any meaningful sense. Positive CLV means you're getting prices above the closing line, which is the best available proxy for true probability. Beating the closing line is mathematically equivalent to positive expected value against the market. You might have positive CLV with negative realised short-term profit (variance), but the expected value is still positive.
Daniel writes about the maths underneath advantage betting — expected value, Kelly sizing, closing line value, bankroll theory. Translates the theoretical side into practical decisions AU punters can actually apply.