The idea
By kickoff, a market like Pinnacle's has absorbed every bet, injury report and line-up. Its closing odds are the most informed price available. If you consistently got odds that were better than that closing price, you were ahead of the information, and that tends to show up as profit over time.
This has been measured. Betting analyst Joseph Buchdahl compared 87,960 prices with Pinnacle's closing odds and found that how much a price beat the close predicted its actual return almost one to one. In other words, how far you beat the close is a good measure of your edge.
How to calculate it
CLV = (odds you took ÷ closing odds) − 1. Take 2.10 on a team that closes at 1.95 and CLV = 2.10 ÷ 1.95 − 1 = +7.7%. Take 1.90 on a side that closes at 2.00 and CLV = −5%.
The closing odds still contain the bookmaker's margin, so a fairer comparison removes it first. Sharp Bets does that: it fetches Pinnacle's closing price, strips out the margin, and measures your price against the fair closing odds.
An example: a match closes at 1.95 on both sides. The implied chances are 51.3% each, 102.6% together, so the fair closing odds are 2.00 for both. If you took 2.10, your CLV against the fair close is 2.10 ÷ 2.00 − 1 = +5.0%, not the +7.7% you'd get against the raw 1.95. Skipping the margin flatters every bet.
Why it beats looking at profit
Profit is noisy. A bettor with a real 5% edge can be down after 200 bets and a bettor with none can be up. CLV strips out most of that luck, because it's measured on every bet at the moment of the bet rather than on who happened to win.
That's why it's the fastest honest test of whether a method works, and why our leaderboard ranks CLV rather than profit.
Does beating the close pay? Our numbers
Sharp Bets has 4,630 settled picks with a closing line. The 3,115 that beat the close returned +10.39% (95% range +4.75% to +16.02%). The 1,515 that didn't returned −16.20% (−22.46% to −9.94%). The two ranges don't overlap, so a gap that size is very unlikely to be luck.
Part of that gap is built in, and that's the point. When news moves a line before the start, it tends to move it towards the side that goes on to win, so a pick that closes worse also tends to lose more often. The closing line carries the information that arrived after you bet, which is exactly what makes it a fair yardstick.
How many bets until you know
Single CLV readings vary a lot: across Sharp Bets' picks the standard deviation is about 11 percentage points. At an average of +2.66%, that means around 60 picks are enough for the 95% range to sit clearly above zero.
Profit is far noisier. At the same edge, the spread of our picks' results means profit alone would need around 11,000 bets to rule out luck. Our own record shows it: after 4,887 settled picks, ROI is +2.95% with a 95% range of −1.03% to +6.93%, a range that still includes zero, while CLV is +2.66% with a range of +2.38% to +2.94%. CLV gives a clear answer after about 60 bets; profit needs thousands.
What it can't tell you
CLV needs a decent sample, a few dozen bets at least, before the average means anything; My Bets shows a 95% range so you can see how sure to be. Thin markets, where the closing line itself is weak, give less reliable readings. And positive CLV isn't a guarantee of profit, only good evidence that the process is sound.
Common CLV mistakes
- Measuring against a slow bookmaker's close. A soft bookmaker's last price often just copies the market late; use a sharp one, such as Pinnacle.
- Leaving the margin in. As in the example above, the raw close makes every bet look a few percent better than it was.
- Taking the close too early. Lines still move in the last hour, often the most. Sharp Bets records it 5 minutes before the start.
- Reading single bets. One bet's CLV says little; the average over many bets, with its range, is what counts.
- Counting only some bets. If you log the bets you remember and skip the rest, the average means nothing. Log every one.
Measuring it with Sharp Bets
Every pick is logged with a timestamp the moment it's found, and the system captures the closing line minutes before kickoff. The Track Record shows the result across all picks. On My Bets, bets logged with one click get their closing odds filled in automatically, so your own CLV builds up without any typing.
What real CLV looks like
Sharp Bets has measured the closing line on 4,701 of its picks so far. Their average CLV is +2.66%, and 67.3% of them beat the closing price. Given how many picks that is, the true average most likely lies between +2.38% and +2.94% (a 95% range), so it is clearly above zero rather than a lucky streak.
That is what a working method tends to look like: a few percent per bet, not tens of percent, with plenty of single picks that close worse. The numbers update on their own as new closing lines come in, and the full breakdown is on the Track Record page.
Common questions
What is a good CLV?
Anything consistently above zero over a few hundred bets. An average of +1 to +3% is a strong result; Sharp Bets' picks average +2.66%.
Can you have positive CLV and still lose money?
Yes, for a while. CLV measures the price, not the result, so losing runs are normal even with positive CLV. Over hundreds of bets the two tend to line up, but nothing guarantees it.
Where do the closing odds come from?
They are the last price before the start, ideally from a sharp bookmaker such as Pinnacle. Sharp Bets records them automatically for every pick, and for every bet you log from the board.
How many bets do I need before my CLV means something?
With a typical spread, around 60 bets give a clear answer if your average is about +2.66%. A smaller edge needs more. My Bets shows the 95% range as you go.