Closing Line Value (CLV): How to Calculate It and What’s Good

10 min read
Closing Line Value (CLV): How to Calculate It and What’s Good

Closing line value (CLV) is the difference between the odds you bet at and the final odds before the game starts. If you take a team at +150 and it closes at +120, you beat the closing line. If it closes at +170, you didn’t. Because the closing line holds the most information the market will ever have, beating it consistently is the best early sign that your bets have positive expected value, long before your win-loss record can prove it.

This guide shows how to calculate CLV the right way (with the vig removed), what counts as a good number, why it works for spreads, totals and moneylines, and what Mysports.AI Research found when it checked whether AI-favored teams moved toward the AI pick before the close.

📈 Picks filtered by EV and CLV

Mysports.AI removes the vig, compares its win probability with the market, and keeps only positive EV picks, with every result on the public record.

Get the AI predictions →

What is closing line value in sports betting?

Closing line value measures whether you got a better price than the market’s final price. The closing line is the last spread, total or moneyline posted before kickoff or tip-off. It has absorbed every injury report, lineup change, weather update and sharp bet, so it is the market’s best estimate of each side’s chance.

A simple example. On Monday you bet the Over 47.5 at -110. By Sunday at 1:00 p.m. ET the total has risen to 49.5. You hold a number two points better than the close, so you have positive CLV. The bet can still lose; CLV says you bought at a good price, not that you won.

Our guide on why sports betting odds change covers what moves a line. CLV simply records which side of those moves you ended up on.

How to calculate closing line value

To calculate closing line value, remove the vig from the closing line, then compare that fair probability with the price you bet. Your CLV in percent is your decimal odds times the no-vig closing probability, minus 1. That figure is the expected value of your bet if the closing line were the true price.

Step by step, using a moneyline you bet at +150:

  1. Get both sides of the closing line. Your team closes at +120, the opponent at -140.
  2. Convert each to implied probability. +120 is 45.45%; -140 is 58.33%. Together they add up to 103.79%. The extra 3.79% is the bookmaker’s margin.
  3. Remove the vig. Divide each side by the total: 45.45 ÷ 103.79 = 43.80% fair probability for your team.
  4. Compare with your price. +150 is 2.50 in decimal odds. 2.50 × 0.4380 − 1 = +9.5%. In probability terms, your price implied 40.00% and the fair close says 43.80%, a gain of 3.8 percentage points.

The no-vig fair odds calculator handles step 3 for two-way and three-way markets. If you need to switch American odds to decimal first, use the odds converter.

Why remove the vig? Raw prices flatter you. A bet at +105 that closes at -105 looks like a 5% better price. But if the other side closed at -115, the fair close on your side is only 48.92%, and your real edge is about +0.3%. More on the margin itself in understanding the vig.

Closing line value examples

These five bets show how CLV changes with your price and the close. The first column is each bet’s expected value against the no-vig closing line.

CLVYour bet → close
+9.5%+150 → +120 / -140
+1.7%-110 → -125 / +105
+0.3%+105 → -105 / -115
-4.5%-110 → -110 / -110
-12.7%-110 → +110 / -130
CLV = your decimal odds × no-vig closing probability − 1. The close shows your side first, then the other side. Vig removed proportionally (each side’s implied probability divided by the total), the same method used by the Mysports.AI no-vig fair odds calculator. Worked examples, not real bets.

The fourth row is the one most bettors miss. Taking -110 and watching it close at -110 is not neutral. The fair price on a -110 / -110 market is +100, a 50% chance, so a -110 bet that never moves loses about 4.5% over time. That is the cost of the vig, and matching the close does not cover it.

Spreads and totals often move in points instead of price. If you bet a team -3 and it closes -4.5, you have 1.5 points of CLV. How much each half-point is worth depends on the sport and on which numbers the line crossed: crossing a common NFL margin like 3 or 7 matters more than moving from 44 to 44.5 on a total.

What is a good closing line value?

A good closing line value is any average that stays positive after the vig is removed, measured over hundreds of bets. A single bet with +10% CLV means little. A small positive average across 500 bets is a far stronger signal than a hot week, because it shows you keep buying below the market’s final price.

There is no official benchmark, so treat these as rules of thumb for reading your own log:

  • Below 0% on average: you are buying after the market has moved. Even good handicapping will struggle to overcome that.
  • Around 0%: you are betting close to fair prices but still paying the vig. Expect a slow loss.
  • Clearly above 0%: your timing or your numbers are ahead of the market. Keep tracking to confirm it is not noise.
  • Beating the close on more than half your bets is a useful second check. An average driven by two or three huge moves is weaker evidence.

Why closing line value predicts long-term profit

Closing line value predicts profit because the closing line is usually a sharper estimate than any single bettor’s, and because CLV can be measured far sooner than wins and losses. Results take thousands of bets to separate skill from luck. CLV gives a usable signal on every bet the moment the game starts.

The math on win rates shows the problem. At -110 you need to win 52.38% of bets to break even. Over 100 bets, normal luck moves your win rate by about ±5 percentage points (one standard deviation). Over 1,000 bets it is still about ±1.6 points. A bettor with a real 54% edge can easily post a losing season.

CLV cuts through that. Sportsbook Pinnacle calls consistent positive CLV one of the strongest indicators of a sharp strategy, and notes that at recreational sportsbooks it is a key factor in whether an account gets limited.

Pair CLV with your own probability estimates. The guide to positive EV betting shows how to judge a bet before it is placed; CLV checks afterward whether the market agreed with you.

Do AI picks beat the closing line?

Not the general-purpose chatbots we tested. In a Mysports.AI Research look-back at 821 games of the 2026 MLB regular season, the team favored by Claude, ChatGPT and Gemini saw its implied win chance change by just +0.01 percentage points on average, from about 10 minutes after the AI made its pick until 5 minutes before first pitch. In other words, the market did not move toward the AI picks.

What else the study found, using Polymarket prices, where a $0.60 share roughly means a 60% chance:

  • The AIs were not more accurate than the market. In the full sample the AI error was larger than the market’s at the same moment (Brier score 0.249 vs 0.238, lower is better). Among the 416 games with third-party proof that the picks were made before the game, the result was not conclusive.
  • Adding AI to the odds did not help. When the AI numbers were blended into the market price, the weight chosen for the AI was 0.
  • Prices often barely moved. In a check of about 520 games, the price changed in only about 4% of them within 15 minutes after the AIs answered. A re-run with real trade prices on 541 games put the move at about +0.05 points, still essentially zero.
  • ChatGPT and Gemini were overconfident; Claude was closest to the market.

The lesson for CLV: a confident pick is not evidence that a line will move your way. The forecasts came from our sister site, aisportsinfo.com, and the AIs could see ESPN’s win probabilities. An NBA follow-up will be analyzed after the regular season ends in April 2027. Full method: Mysports.AI Research, AI predictions vs pre-game odds.

How Mysports.AI uses closing line value

Mysports.AI uses closing line value as a backtesting check, not as a pick by itself. The model first removes the vig from market odds, keeps only bets with positive expected value, then compares the price it would have bet at with the closing odds. Action is concentrated where both EV and CLV are positive.

This is a dedicated sports model, a different system from the general-purpose chatbots in the study above. Per the Mysports.AI model methodology, it works in six steps:

  1. Remove the vig so implied probabilities add up to 100%.
  2. Calculate EV: AI win probability × decimal odds − 1. At a 58% win probability and 1.90 odds (about -111), EV is +10.2% per bet.
  3. Backtest with CLV to check whether the model’s prices beat the close.
  4. Size stakes with fractional Kelly.
  5. Refresh daily as line moves, injuries and lineups update, then lock the day’s picks close to game time.
  6. Publish every result, misses included, on the AI picks track record.

Locking near game time has a trade-off you should know: picks made close to the close have less room to gain CLV than bets placed days earlier. The upside is that they use the latest injury and lineup news.

How to track closing line value

To track closing line value, log every bet with the price you got and both sides of the closing price, then calculate no-vig CLV once the game starts. A spreadsheet is enough. Review average CLV every 100 bets.

  • Record: date, league, market, your odds, sportsbook, time placed (ET), and closing odds for both sides.
  • Use one reference close. Take the closing line from the same low-margin market every time, or an average across books, so your numbers are comparable.
  • Capture the close right at start time. A line 30 minutes before kickoff is not the close; late injury news often lands in that window.
  • Split by market. You may beat the close on NBA totals but not on NFL sides. Average CLV by league and bet type.

Limits of closing line value

Closing line value is a strong signal, not a guarantee. The close can be wrong, especially in thin markets such as small-conference college games, low-tier soccer and many player props, where a few bets move the line. Line shopping also inflates CLV: grabbing a stale price at one book can look like skill even when you had no view on the game.

The Mysports.AI figures above come from a look-back at the 2026 MLB regular season with forecasts from our own sister site, and several follow-up checks were added after seeing the data, so the research page labels them post-hoc and for reference only. They are past averages and guarantee nothing for any single game.

21+. Betting involves risk. Only bet what you can afford to lose. If gambling stops being fun, call or text 1-800-GAMBLER.

Grace Walker

AI Sports Data Analyst

Grace Walker is the AI sports data analyst at Mysports.AI. She focuses on football, covering moneyline, both teams to score (BTTS) and point spread markets, with close attention to expected goals (xG) and odds movement.

Closing line value FAQ

What does closing line value mean?

Closing line value is how much better or worse your odds were than the final odds before the game started. If you bet a team at +150 and it closes at +120, you beat the close. Because the closing line holds the most information, beating it consistently suggests your bets have positive expected value.

How do you calculate CLV?

Remove the vig from the closing line to get a fair probability, then multiply your decimal odds by that probability and subtract 1. A +150 bet (2.50) on a team that closes +120 against -140 has a fair closing probability of 43.80%, so CLV is 2.50 x 0.438 – 1, or about +9.5%.

What is a good CLV percentage?

There is no official benchmark. Any average that stays above 0% after removing the vig, across several hundred bets, is a good sign. Matching the close at -110 on both sides is not break-even: it works out to about -4.5% because you still pay the vig.

Can you lose money while beating the closing line?

Yes, in the short run. A single bet with positive CLV can lose, and at -110 your win rate swings about 5 percentage points either way over 100 bets from luck alone. Positive CLV across a large sample is what points to long-term profit.

Do AI predictions move the line before the game?

In a Mysports.AI Research look-back at 821 MLB games in 2026, the team favored by Claude, ChatGPT and Gemini saw its implied win chance change by only +0.01 percentage points before first pitch. The market did not move toward the AI picks, and the AIs were not shown to be more accurate than the market.

Found this useful? Share it