Closing line value (CLV) demystified by expert Joseph Buchdahl

Do not index
Do not index

What is the Closing Line?

The term ‘closing line’ comes from the world of American sports betting, specifically, line betting or point spread markets, a kind of handicap market designed to equalise the odds on each side by adjusting the points total. In US betting parlance, this is known as the line. Furthermore, what UK and European bettors might call fixed odds, for example, the prices for a home, draw and away in football, are known in the US as moneyline. The closing line (or reasonably closing odds or closing price) is simply the line (or price or odds) at the moment the sportsbook closes the betting market just before a match begins. Having opened their betting market, sportsbooks are known to frequently move their lines (or odds) in response to two main things: news and money. The two are not mutually exclusive. News about a particular game, and the teams and players involved, may shift the line (or odds). Similarly, money coming from the sportsbooks’ customers may help shift the sportsbook’s opinion about how accurate their betting odds are. More money on one side may cause the sportsbook to shorten their price on that side. In theory, then, the more opportunity there has been for more news and more money to flow into the market, the more accurate the line or odds reflecting that news and money should be. Since the closing line or odds is the last one, theory tells us that it should, on average, be the most accurate of all. Just how accurate the average will be to a significant extent depends on the betting market, but for very popular like the English Premier League or the NFL, there is ample evidence to show that the closing line/odds provide a reasonable measure of the true outcome probabilities, at least on average. Not all sportsbooks are equal, however; some are more accurate than others. Scoring rules like Brier and rank probability scores can be used to show which ones are better than others. The most accurate sportsbook is Pinnacle Sportsbook. This is so for a reason, largely dependent on its business model of accommodating sharp bettors and their money, rather than restricting or banning them like more recreational sportsbook. Pinnacle’s closing line (or odds), then, are widely accepted as providing the most accurate numbers and is hence why they form the basis of both my Wisdom of the (Pinnacle) Crowd betting methodology and the one used by the Pinnacle Odds Dropper (POD) service.

What is the Closing Line Value?

Now we know what the closing line is, what about closing line value? Any serious bettor knows that to be successful in betting is not a matter of just finding winners, but finding more winners than the sportsbook believes you should be finding. The way we measure this metric is by means of expected value. Tossing heads is a 50% position, but suppose I have secret information that tells me my coin is biased, such that 51 times out of 100 it will show heads. My expected value is 51/50 or 1.02. If I bet $1 on heads every time, I can expect to win 51 times and lose 49 times out of every 100 tosses, for an expected profit of $2 per $100 wagered (2%), or a total return of $102.Since decimal odds are just the reciprocal of the implied outcome probabilities, it’s really easy to calculate the expected value. If the true odds of a proposition were 2.00, but the sportsbook was offering 2.10, then the expected value is just 2.10/2.00= 1.05 (or 5%). Closing line value, when expressed using decimal odds, is then just the sportsbook's odds you have bet divided by the closing odds, assuming that the closing odds do indeed represent a fair reflection of the true outcome probability.

Making the Closing Odds Fair

The observant amongst you may have spotted an error. If the Pinnacle’s closing odds contain their betting margin, sometimes known as the vig, how can they be a fair reflection of the true outcome probability? This is true, and to calculate the fair odds, we need to remove it. In the absence of knowing precisely how Pinnacle applies itsvig, there are various documented ways for removing it. These include the rather crude and frankly inaccurate equal margin method, which assumes the margin or vig is equally distributed on all sides of a betting market, to more sophisticated methods that assume a favourite-longshot bias in asymmetric betting markets that show over-betting of underdogs, including the odds ratio, logarithm function and Shin methods. Pinnacle Odds Dropper provides an easy summary for a number of these methods and subscribers have access to their own de-vigging calculator. For those wanting to dive deeper into the mathematics behind the methodology, my Wisdom of Crowd PDF and a paper in the American Journal of Sports Science provide further reading.

The Significance of Closing Line Value

What’s so special about holding closing line value? If the theory of more accurate odds at market closing holds, this tells you that if you are beating the closing no-vig-price (abbreviated to NVP), your bets should hold expected value. Holding expected value, not picking winners, is the holy grail for all serious sports bettors. CLV, then, offers a useful proxy measuring stick of the size of your expected value. If the closingNVP is a reasonable reflection of the true price, then the amount you beat it by will be a reasonable reflection of your expected value. If you beat it by 5%, then the theory dictates that your expected value should be 5%, implying over the long term your profit over the total amount wagered should be 5%.Over the course of nearly 20,000 bets in my Wisdom of the Crowd betting system, the actual profit over turnover of 3.4% compares closely to the expected value of 4.0% and is well within the bounds of statistical significance. Closing line value is also far more useful than your results in determining whether a forecasting method is doing what it is supposed to be doing, or whether it is all just happening by chance. Why? Because it’s so much faster at providing statistically significant information in this respect. Profits and losses, for example for unit-stake even-money betting markets, come in+1s and -1s. How much total profit you are showing after 100 or 1,000 bets can vary quite considerably just because of random good and bad luck. Hence, we need a large number of results before any statistical testing can tease out a skill signal against a background of random noise. For such an even-money bettor with a CLVand expected value of 5%, it might take several thousand bets before I would confidently rule out chance as the only explanation. By contrast, CLV and expected value for bets vary in much smaller increments. Whilst the typical standard deviation in even-money profits and losses will be about 1.00, the equivalent for CLV will be about 0.1. Hence, if we are consistently beating the closing line, it will take far fewer bets, perhaps as few as just 50, to demonstrate the statistical significance that will tell us something causal is at work, and not just good fortune. [You can find out more about the mathematics involved here in my last book Monte Carlo or Bust: Simple Simulations for Aspiring Sports Bettors.]Furthermore, because CLV can so quickly reveal a signal, it will also quickly tell you when CLV may have disappeared, and whether it might be time to adjust or change forecasting strategy.

CLV and Pinnacle Odds Dropper

Both the Pinnacle Odds Dropper service and my own Wisdom of Crowds methodology use Pinnacle’s odds as a measure of the true odds. Indeed, I maintain that Pinnacle’s pre-closing odds are good enough to be used to find betting value at other sportsbooks and the betting record I referenced above has demonstrated that. The Pinnacle Odds Dropper service detects falls in Pinnacle sportsbook betting odds, allowing subscribers to place value bets at other sportsbooks where their prices have not yet adjusted to reflect the new information and money that Pinnacle is basing their judgment on. Do these bets show any closing line value? The answer is a resounding yes. In Monte Carlo or Bust, I analysed a sample of football betting odds where Pinnacle’s odds revealed 952 positive expected value bets at other sportsbooks. Of these 952 bets, 756 of them saw either price shortening or no price movement at the sportsbook where the bet was placed, with an average price shortening of 3.94%. A statistical test revealed it was effectively impossible these price movements happened by chance. [Indeed, I calculated it would have needed only 65 bets to prove this.] If it wasn’t chance, the obvious explanation for such closing line value is that as the expected value made available by the recreational sportsbooks was exploited, they were forced to shorten their prices, and by closing were nearly 4%shorter than at the time the bets were placed.

Summing Up

Not all sharp bettors will show closing line value. Odds originators, for example, those harp bettors with novel forecasting models who for a time remain unknown to sharp sportsbooks like Pinnacle, will not, initially at least, have their information reflected in the betting markets. Over time, as such bettors wager more money, and as Pinnacle starts to take notice of the information their modelling brings, this might start to change. The kind of ‘line-shopping’ methodology utilised by Pinnacle Odds Dropper and Wisdom of the Crowd, however, will most certainly reveal itself in the closing line value. And so long as you see it, you can be confident that these services will remain as robust and reliable as they were when first introduced.

Sign Up to Pinnacle Odds Dropper with our 21-Day Money Back Guarantee

Sign Up

Written by