Closing line value: the only fast feedback loop in betting

Why the outcome is the least informative thing about a bet

A bettor with a genuine 55% edge at standard juice needs roughly 2,250 settled bets before the record separates from break-even at conventional confidence. At 53% it is close to 40,000. Almost nobody has that, and the ones who do accumulated it over years during which their actual edge changed several times.

Closing line value sidesteps the problem by measuring something with far less noise in it. You are not asking whether the team won. You are asking whether the price you took was better than the price the market settled on, and prices carry information that single outcomes do not. Tens of bets start to say something. Hundreds say a lot.

Measure it against a de-vigged close, or do not bother

Comparing your minus 105 against a closing minus 110 at the same book, before removing margin from either, measures the book margin as much as it measures you. Take the de-vigged closing probability, take the raw price you actually got, and express the gap as expected return per unit staked. That converts a vague sense of beating the close into a number you can average.

Choose the closing book deliberately. The close at a market-making book that accepts large stakes from winning accounts is an estimate. The close at a book that limits anyone who wins is a marketing number, and beating it proves you can find a slow price rather than that you can forecast.

• Record the exact price and time you took, on every bet, without exception.
• Record the closing two-way price at a chosen reference book.
• De-vig the close with one method and stay with that method.
• Convert the gap to expected return per unit staked, then average across bets.

Three ways closing line value lies

The first is when you moved the line yourself. If your stake is large enough relative to the market to shift the price, the close is partly a record of your own action, and you are measuring your influence rather than your accuracy. The second is stale-line hunting: consistently beating a slow book on news that has already moved the sharp books is real money, but it is a speed edge with a short shelf life, and it will not survive the account restrictions it provokes.

The third is the subtle one. Closing line value in illiquid or exotic markets, props, minor leagues, novelty lines, mostly measures the variance of a thin market rather than any edge. The closing price there is not a consensus, it is whatever the last person to trade left behind.

Our position: track closing line value on every bet, treat it as your primary feedback signal for the first several hundred bets, and switch to profit only when the sample can carry the question. A bettor showing positive closing line value and negative profit over 200 bets is almost certainly running badly. A bettor showing profit and negative closing line value over the same sample is almost certainly running well.