Glossary
The words this site uses, defined.
Football pricing has a vocabulary, and most of it is used on this site without ceremony. Each definition below says what the term means here and what it does not claim. Longer explanations live on how it works and methodology.
Implied probability
The probability a decimal price corresponds to before any margin is removed, calculated as 1 ÷ decimal odds.
A price of 2.50 implies 40%. Across a complete market these numbers sum to more than 100%, and the excess is the bookmaker margin — so an implied probability is a price restated, not an estimate of what will happen.
Bookmaker margin
Also: Overround · Vig · Vigorish · Juice
The amount by which a complete market's implied probabilities exceed 100%, also called the overround.
On a three-way football market the implied probabilities might sum to 102.8%. Those 2.8 points are the margin. It is charged whichever side you take, which is why it has to be removed before a price and an estimate can be compared at all.
De-vig
Also: De-vigging · Margin removal · No-vig conversion
Removing the bookmaker margin from a complete market so that its probabilities sum to 100%; Clauseground currently removes it proportionally, dividing each raw implied probability by their total.
Proportional removal — also called the multiplicative method — is the simplest of several. It assumes the margin is spread evenly across the outcomes, which is not exactly true: books usually load more of it onto longshots. Clauseground uses it because it is the method the live board actually runs, and says so rather than describing a more sophisticated one it does not use.
Fair probability
Also: No-vig probability
A market-implied probability after the bookmaker margin has been removed across every outcome in the same market; interchangeable with no-vig probability.
This is what the market is pricing once the house charge is taken out. It is a cleaner reading of a price, not a promise about the result, and it is only as good as the market it came from.
Market probability
No-vig probability derived from the current designated reference market.
Clauseground checks designated reference books in configured order and uses the first with a complete market. It does not yet build a liquidity-weighted consensus across every venue, so the site calls this the reference market view rather than the market consensus.
Model probability
Independent statistical model output.
A Poisson scoring model with a Dixon-Coles correction, run without reference to the price. Keeping it independent is the point: a model fitted to the market can only ever agree with it.
Combined estimate
Market-specific weighted combination of market and model probabilities.
The blended number the board ranks on. Its provenance is labelled everywhere it is shown, because a blend is neither of its inputs and should not be read as either.
Difference
One probability minus another, with both bases named.
Reported in percentage points. A difference is a question — is the model missing something, or has the price not caught up? — and never on its own an answer.
Expected value
Also: EV
What one unit staked at a given price returns on average if the estimated probability is correct — an arithmetic consequence of a price and an estimate, never a forecast of profit.
The conditional is the whole thing. Expected value inherits every weakness of the probability it is computed from, so a large positive number is most often evidence that the estimate is wrong rather than that the price is. Clauseground does not currently claim a proven historical edge.
Potential value
A candidate price comparison that still depends on freshness, fees, uncertainty, and user judgment.
The deliberately hedged phrase the product uses instead of "value bet". A comparison made against a price that has since moved, at a venue you cannot use, is not an opportunity.
Best price
The highest decimal odds Clauseground has collected for an outcome across the venues it surveys, which is not necessarily a price any one account could have taken.
Every match page names the venue behind each price and the time it was collected. The caveat in the definition is load-bearing: the surveyed set includes books that limit winning accounts, books unavailable in a given jurisdiction, and quotes that were never simultaneously obtainable.
Line shopping
Also: Price comparison
Comparing the same selection across the venues you can actually use and taking the highest available price.
The largest effect this site can demonstrate, and it is a property of the market rather than a result of ours. It has been measured — see the research article — and the measurement is an upper bound on what a real account captures, not a return.
Closing line value
Also: CLV
The difference between the price recorded when a selection was published and the market price near closing, read with its sample size and never as proof of future profit.
Closing line value is outcome-independent, which is what makes it useful on small samples: it asks whether a process found prices the market later agreed were good. Clauseground uses it internally as a steering metric and does not publish it as a performance claim.
Pick
An explicit published selection, not every disagreement.
Most differences between the model and the market are never published. A pick is one that cleared the publication rules, and it is decision support rather than advice — the decision, including the decision not to bet, stays with the reader.
AI analyst
The interface that retrieves and explains Clauseground data.
It reads structured outputs from the pricing and modelling tools and explains them in plain language. It does not generate probabilities, prices, or results of its own, and it is instructed not to invent a data source.
Decision support, not betting advice. No guaranteed outcomes. 18+/21+ where applicable. Gamble responsibly.