Sports Betting ROI vs CLV: Which Number First

CLV tells you whether the prices you took beat a sharp close; ROI tells you whether those bets, after results, paid more than they cost. They measure different things. Over a short sample they disagree often. Over a very large sample they tend to line up if the close you use is honest. Change staking and selection rules on CLV. Read ROI as the lagging bill, not the steering wheel.

This page is the comparison. For the CLV formula, start with closing line value. For how many bets before CLV is evidence, CLV sample size. For how to log, the CLV tracker guide.

Definition. ROI (return on investment) is (total returns − total stakes) ÷ total stakes, after results. CLV (closing line value) is the gap between the odds you took and a later closing benchmark for the same selection, usually Pinnacle with the margin removed: odds taken ÷ fair close − 1. CLV does not wait for the match to finish.

ROI vs CLV at a glance

Neither number is a promise of future returns. Positive CLV is not a locked profit. Negative ROI on a small sample is not proof the process is broken.

Two formulas, two clocks

Take a bet at 2.10. Pinnacle's fair close on the same line is 2.00.

• CLV = 2.10 ÷ 2.00 − 1 = +5%. You have this as soon as the market closes. The match has not started.
• ROI on this bet is +110% if it wins (you receive 2.10 per 1 staked, net +1.10) and −100% if it loses.

Average those results over many bets and you get portfolio ROI. Average the CLV figures and you get average CLV. They are not interchangeable. A week of +8% ROI and −1% CLV means you got worse prices than the close and the results came in anyway. A week of −8% ROI and +2% CLV means the opposite. The second week is the one that should keep the rules. The first should not scale the stake.

Why ROI stays noisy for so long

Each result is a coin with a small bias. In betting discipline we published the binomial arithmetic for a true +3% edge at odds of 2.00 (a 51.5% chance of winning each flat-stake bet). Chance of still being behind on results:

After 500 bets, almost one process in four with a real +3% edge is still losing money. That is ROI. It is the wrong number to abandon a rule on.

CLV does not remove variance. It measures a different thing, on every trial. The share of bets with CLV > 0 needs about 380 bets before a 95% interval on a 55% share sits entirely above 50%, and about 2,400 if the share is only 52%. Those figures are original arithmetic, first published in CLV sample size on 6 November 2026. Average CLV% still needs the standard deviation from your log; do not invent one.

When they do line up

Joseph Buchdahl used 35,570 matches (106,710 1X2 prices, 22 European divisions, 28 July 2012–20 February 2017) to test whether prices longer than Pinnacle's de-vigged odds at four UK books predicted actual returns (football-data.co.uk, 24 February 2017, re-read 6 November 2026). On the 24,150 prices where expected value was above 1.00, average expected return was 1.016 and actual return 1.018. By book: bet365 6,250 bets, 101.67% expected vs 103.30% actual; Ladbrokes 3,935, 101.81 vs 102.64; William Hill 7,669, 101.54 vs 101.20; Betvictor 6,296, 101.45 vs 100.53.

That is EV from a sharp fair price versus realised ROI, on tens of thousands of bets. It is evidence that beating Pinnacle's fair price, as a process, showed up in results at that scale. It is not a forecast of your next 50 bets. Buchdahl's own warning in the same article: as you raise the value threshold, the sample shrinks and luck dominates; run a range of outcomes, do not treat the lucky bins as a target.

The reverse test in the same paper matters as much: using the four UK books as the "fair" guide to bet Pinnacle produced no useful correlation with actual returns. A soft close is a bad CLV benchmark. Beating bet365's close is not the same as beating Pinnacle's.

Four ways the two numbers disagree

Variance. The binomial table. Do not "catch up" on stake after a losing month with positive CLV. Do not raise the stake after a winning month with negative CLV.

The wrong close. Raw CLV against a soft book includes that book's margin and its delay. No-vig CLV against Pinnacle is the usual honest figure. Mixing them in one average is how a dashboard looks healthy while the process is not. Trackers are compared in CLV tracking tools.

Thin markets. A close on an ITF games handicap or a lower-league total is a weaker benchmark than a Premier League 1X2. Positive CLV there can fail to show up as ROI because the close itself was noisy. Split the log by league and market.

What you actually got paid. CLV uses the odds you logged. ROI uses what settled. Voids, rule mismatches, a price that moved on the slip, and a limit that forced a smaller stake all open a gap. Log the accepted price, not the alert price.

Which number to change the rules on

• After a few dozen bets: neither. Direction only. A 7-day tool trial is a screen; see how to test a value betting tool.
• After a few hundred bets with honest logging: CLV (share above the fair close, then average once you have a standard deviation). If it is still a coin flip, keep logging. If it is negative on main markets against Pinnacle's no-vig close, the process is not beating the market; change selection, not the stake.
• After thousands of bets: ROI becomes readable. Use it to check that the bankroll did what the process said, not to pick next week's markets.

CLV is the leading indicator of price quality. ROI is the lagging confirmation, and only on a sample large enough that the binomial table has cooled down. Neither one is a reason to promise a yield.

How Pinnacle Terminal handles this

Pinnacle Terminal's tracker records Pinnacle's closing price on every bet you log, from the app or from Telegram, and reports no-vig CLV with four devig methods plus a worst case (Power by default). You can export taken price, close and result and compute both ROI and CLV yourself in the CLV calculator. It does not print a "projected ROI" from last week's CLV. Live in-play football drops are shown and filtered, without a grade. The terminal does not take bets and is not affiliated with Pinnacle. Setup: tracker and CLV guide. Product: Pinnacle dropping odds. Plans: pricing.

FAQ

Is ROI or CLV more important in sports betting?

For deciding whether your process works, CLV first. For seeing what the bankroll did, ROI. On a few hundred bets, CLV is the readable one. On a few thousand, both should be in the log.

Does positive CLV mean I will make money?

No. It means you took better prices than the close you chose. Results still swing. Limits, voids and a weak close can all keep ROI from matching CLV. There is no guaranteed return.

Why is my ROI negative if my CLV is positive?

The usual reason on a small sample is variance: see the 23.7% figure after 500 bets at a +3% edge. Next checks: fair close, complete log, thin markets pulled out of the average.

Can I use ROI to judge a value betting tool?

Not on a week's trial. A week's results are mostly luck. Judge the trial on CLV of bets you could actually place at your books.

Should I stop betting if ROI is down and CLV is up?

Not on that pair alone. Recheck the close and the log. Keep the written rules. If CLV itself has gone negative on main markets against Pinnacle's no-vig close, then change selection.

Sources

• Binomial table (+3% edge at 2.00; 100 / 500 / 1,000 / 2,000 bets): original arithmetic, first published in betting discipline (31 October 2026).
• Share-of-bets sample sizes (52% / 55%): original arithmetic in CLV sample size (6 November 2026).
• Joseph Buchdahl, "The Wisdom & Market Efficiency of Pinnacle.com compared to other Bookmakers", football-data.co.uk, 24 February 2017, re-read 6 November 2026 (35,570 matches; 24,150 +EV prices, expected 1.016 vs actual 1.018).

Try it

To log every bet against Pinnacle's fair close automatically, start the Pinnacle Terminal 7-day free trial. Nothing is charged before day 8. Plans: pricing.

18+ · Bet responsibly. Only bet money you can afford to lose.