How to Read Decimal Odds and Implied Probability

A decimal price of 1.36 on the Oklahoma City Thunder for their October 23 game against Denver, as listed on Oddsportal's NBA page, means the market rates the Thunder's chances at about 73.5%. The Nuggets' price of 3.40 puts Denver at about 29.4%. Those two figures add up to 102.9%, which is impossible for real outcomes and entirely deliberate for a sportsbook.

The conversion takes one division. Its most useful output is not either team's chance of winning. It is the size of the cut the house takes from every bet, whichever side wins. This page takes a reader who has never converted a price to the point of doing it unaided on any two-sided line, with the house's share visible in the result.

Decimal odds are the total amount a bet pays back for each peso staked, including the stake itself. The stake is the money put down. The return is stake multiplied by the decimal price. Profit is return minus stake. A price of 2.50 returns 2.50 for every 1 wagered, which is 1.50 of profit, as the NBA Game Betting explainer sets out. A lower decimal price marks the favorite and a higher one marks the underdog.

Implied probability is the chance of an outcome that a price assumes. The formula, given by OddsIndex and others, is:

Implied probability = 1 ÷ decimal odds × 100

So 2.00 is 50%, 1.50 is 66.67%, 4.00 is 25% and 1.91 is 52.36%. A ₱200 bet on the Thunder at 1.36 would return ₱272, which is ₱72 of profit, and 1 ÷ 1.36 gives 73.5%. In plain terms, the price asks the bettor to risk ₱200 to make ₱72 on an outcome the book treats as roughly three-in-four.

The second idea follows from the first. Add the implied probabilities of every possible result in a game, and an honest coin would total exactly 100%. A sportsbook's prices total more. The excess is called the overround, also known as the margin or vig (short for vigorish, also called juice), and it is the house edge. Oddsportal's odds calculator page puts it bluntly: "Implied probability always factors in the vig, or the sportsbook's profit margin. Don't confuse it with 'true' or 'fair' odds or probability, which factors out the juice."

Very little, and nothing that costs money. The first conversion takes about ten minutes. Later ones take under one.

The 2026-27 NBA regular season opens on October 20 and runs to April 11, 2027, with the NBA Cup group stage from October 30 to December 11, according to Oddsportal's fixtures page. That gives a live set of decimal prices to work on. The example below uses the Thunder and Nuggets game dated October 23, priced 1.36 and 3.40. Any two-sided line works the same way.

Done once, the method applies to any game. The three examples below are all from the Oddsportal NBA page as it stood before opening week. The margin varies from one game to the next, and on none of these lines does it reach zero.

The last row of that table is hypothetical. No source available to this blog carries a decimal PBA Governors' Cup line, so the figures below illustrate the shape of a close PBA game priced the way spreads and totals usually are. A price of 1.91 is the decimal form of the standard -110 American line.

Say a bettor puts ₱100 on each side of a Governors' Cup game priced 1.91 and 1.91. That is ₱200 staked. Whichever team wins, the winning ticket returns ₱191. The bettor has backed both possible results and is still ₱9 short, a loss of 4.5% on the money put down. That ₱9 is the overround made visible. A bettor who picks only one side pays it too. The cost is simply hidden inside the price rather than shown as a separate charge.

The obvious objection is that 2.9 points on a Thunder game, or ₱9 on ₱200, is too small to bother about. On a single bet, it is. A ₱200 stake at a 2.9-point overround costs roughly ₱6 in expected terms, less than a jeepney fare. The objection fails on repetition rather than on size. The margin is charged on every bet, on winners and losers alike, across a regular season that runs from October to April. A cost that is trivial once and certain every time is exactly the kind a household budget is built to notice.

Treating implied probability as a forecast. A figure of 73.5% is what the price assumes, margin included. It is not what will happen, and it is not even the book's clean estimate. The sign this has happened is a bettor thinking "the Thunder are 73% to win" rather than "the Thunder are priced at 73%". The fix is to run step six every time and to remember that the stripped figure is still only an opinion.

Reading one side of the line. Converting only the team being backed hides the margin completely, because the overround appears only when both sides are added together. The sign is a conversion sheet with one number per game. The fix is never to convert a price without converting its opposite.

Using a stale or second-hand price. Lines move, and outlets disagree. The Fox Sports opening-night preview quoted DraftKings odds as of August 11, with the Knicks at -162 against Philadelphia. A later Covers.com preview listed the same game at -175, which implies about 63.6%. OddsTrader showed spread juice of -118 and -102 on that game, where another book had -110 on both sides. The sign is a conversion done on a number from an article rather than from the operator's own screen at the moment of betting. The fix is to convert the live price, and to treat every published figure, including the ones on this page, as dated.

Running the formula on something that is not a sportsbook line. Prediction markets quote prices that look like decimal odds but do not obey 1 ÷ price. Kalshi's PBA page, for example, has listed Ginebra and Meralco both at 1.36x while showing 49% for each. Under the formula, 1.36 would mean 73.5%. The sign is a pair of percentages that cannot be reproduced by division. The fix is to use the formula only on bookmaker decimal prices from a PAGCOR-licensed operator.

The natural second step is the other common odds format. Most American previews quote moneylines rather than decimals. OddsIndex gives the two conversions. For a negative number, divide its size by its size plus 100: -175 becomes 175 ÷ 275, or 63.6%. For a positive number, divide 100 by the number plus 100: +145 becomes 100 ÷ 245, or 40.8%. Running both sides of the Covers.com Knicks and 76ers line gives 104.4%, a wider margin than the Oddsportal decimal line on the same game. That gap is a lesson in itself.

After that come spreads and totals, which are usually priced near 1.91 on each side and so carry roughly the 4.5% cut shown in the hypothetical PBA example. The PBA calendar offers a natural place to practise. The Governors' Cup resumed on October 7, with Meralco against Blackwater and Macau against TNT on October 13, according to the Inquirer's schedule. The Manila Times reports that another pause is planned after the quarterfinal phase so that Gilas can play Syria and Iran in late November. A conference with a break in the middle is a good reason to set a budget for the whole conference rather than week by week.

The formula is sport-agnostic, so it applies unchanged to a two-sided price on a fight or a football match. Converting a decimal line from a licensed operator on any of them, and adding both sides, is the habit worth building.