01
CLOSING LINE VALUEA better price is worth tracking.
Judge the decision before judging the result.
Closing line value compares the price you took with the final pregame market. For the same selection and line, a higher decimal price pays more if the bet wins. Track that comparison consistently, using a named reference book and a timestamp.
CLV is a useful process signal alongside units, return on stakes, and the accuracy of your probability estimates. It is not the only scoreboard. A thin market, a different settlement rule, or an unreliable closing source can make the comparison misleading.
WORK IT THROUGHExample: you take +110 (2.10 decimal), and the same selection closes -110 (1.9091). Price CLV is 2.10 ÷ 1.9091 − 1 = +10%. That includes the book’s margin; it is not automatically a 10% expected profit.
Your next stepRecord the bet price when you enter. Add the close later. Never compare -3 with -3.5 as though they were the same wager.
Try the CLV grader →Further reading: closing-market efficiency ↗02
DEVIGGINGThe price includes a margin.
A market estimate starts with every outcome.
Implied probability is 1 divided by decimal odds. When the probabilities for all outcomes sum to more than 100%, that excess is the overround. Proportional devigging divides each implied probability by their total.
The result is a no-vig market estimate, not a discovered truth. The power method allocates the margin differently. Use all mutually exclusive outcomes from one book and timestamp, and keep your method consistent.
WORK IT THROUGHExample: -110 on both sides implies 52.38% + 52.38% = 104.76%. Divide each side by 104.76% and the proportional no-vig estimate is 50% each. Fair odds are +100.
Your next stepFor a two-way market with refunded pushes, interpret the probabilities conditional on no push. For three-way markets, include the draw.
Remove the vig →Further reading: betting margins ↗03
BANKROLL 101Survive the ordinary losing streak.
Your stake is part of the decision.
Keep betting money separate from essential spending. Define a unit before the slate starts, and record the amount risked as well as the net result. A $100 stake returning $190.91 creates $90.91 profit, not $190.91.
Kelly uses your estimated probability and the offered price to calculate a growth-oriented stake fraction. That estimate can be wrong. Fractional Kelly reduces the calculated exposure, but does not make the bet safe. Account for all open bets, especially bets that depend on the same game.
WORK IT THROUGHExample: a $1,000 bankroll, 55% win probability, and -110 odds give 5.50% full Kelly. Quarter Kelly is 1.375%, or $13.75. If your estimate falls to 52%, the positive-edge stake becomes $0.
Your next stepSet a maximum exposure before betting. Do not raise it to recover a loss. When your probability estimate is weak, the calculator cannot fix it.
Explore stake sizing →Original research: Kelly (1956) ↗04
STEAM & SHARP MONEYA moving line is a question.
Ask what changed, who moved, and when.
“Steam” describes a fast move across multiple sportsbooks. “Sharp money” describes wagers believed to come from informed bettors. A moving screen alone does not identify who placed the wagers or whether the new number still offers value.
Compare books and timestamps. Check injuries, confirmed lineups, market limits, and whether a single book is catching up. A screenshot without timing is not evidence of coordinated action. Public betting percentages are also incomplete without knowing their source and sample.
WORK IT THROUGHExample: a total moves from 47.5 to 45.5 after lineup news. Following at 45.5 is a different decision from buying 47.5. Recalculate your edge at the price you can actually get.
Your next stepUse our market watch to notice a change, then investigate. Opening-to-latest quotes do not tell you that a move happened overnight.
Compare the live board →05
WEATHER & THE PRICEConditions need context.
A forecast becomes useful when it changes your estimate.
Check the venue, start time, roof status, and the latest local forecast. Look at wind direction as well as speed, and distinguish sustained wind from gusts. A city-wide forecast does not automatically describe conditions on the playing surface.
Then ask whether the market has already adjusted. Rain is not an automatic under, and a weather headline is not a quantified edge. Forecast uncertainty matters: revisit your assumptions as kickoff or first pitch approaches.
WORK IT THROUGHExample workflow: note the forecast and its timestamp, describe the effect you expect, compare the current total with your estimate, and write down what new information would make you pass.
Your next stepUse the weather desk as a starting point. Verify the game-time details with a local forecast before acting.
Open the daily pulse →Source: National Weather Service forecast process ↗06
THE POSTGAME REVIEWKeep the losses in the picture.
A bad beat is a story. The full ledger is evidence.
Grade every play using the stated settlement rules. Keep the original selection, entry price, stake, date, tier, and final outcome. Wins, losses, and pushes all belong in the same ledger.
Separate outcome review from process review. Ask whether your assumptions were reasonable at the time, whether you secured the intended price, and whether your stake matched your plan. Do not rewrite your reasoning after the score is known.
WORK IT THROUGHExample: an over misses by half a point. Record the loss first. The narrow margin alone says nothing about whether your original probability estimate or bet size was sound.
Your next stepReview a group of bets, not only the most painful result. Filter our Record by sport and tier, then compare the full units trajectory.
Inspect the full record →