Betting odds explained: what −110 actually costs you

American odds look arbitrary and are not. A negative number is what you must risk to win $100. A positive number is what you win from a $100 risk. −110 means risk $110 to win $100, and +150 means risk $100 to win $150.

The useful move is converting a price into the probability it implies, because that is the number you have to beat.

Price to probability

For a negative price, implied probability is odds / (odds + 100) using the absolute value. For a positive price it is 100 / (odds + 100). A quicker way to think about it: implied probability is always 1 / (1 + profit per unit risked).

Two things fall out of that table immediately. First, a coin flip at −110 loses money: you win half the time and are paid less than even. Second, the difference between −105 and −120 is over three percentage points of required hit rate — larger than almost any edge that genuinely exists. Shopping price is not a refinement; on most bets it is bigger than the handicapping.

The vig, and how to strip it out

When both sides of a market are −110, each implies 52.38% and the two add to 104.76%. That excess is the hold. To recover what the market actually thinks, divide each side by the total: 52.38 / 104.76 = 50.0%. That is a de-vigged probability, and it is the market’s real opinion.

De-vigging matters because the market’s opinion is the thing you are trying to beat. If your model says 55% and the de-vigged market says 54%, your edge is one point — not the five points you get by comparing against 50%.

The conversion that governs everything

Here is the number that puts every handicapping claim in perspective. The difference between the final margin and the closing spread has a standard deviation of about 12.72 points. Push one point of line through that distribution and you move the cover rate by roughly 3.13 percentage points. For totals the spread of outcomes is slightly wider, so a point is worth about 3.03 points of probability.

Run it backwards and the bar becomes concrete:

• beating −110 requires 52.38%, which is 0.76 points of true edge over the closing line;
• a 55% handicapper would be beating the closing line by about 0.8 points every week;
• a 60% handicapper would be beating it by roughly 2.4 points, which is the width of a whole key number.

That is the right lens for any claimed record. The closing line is set by people whose full-time job is setting it and then moved by everyone willing to bet into it. Beating it by two and a half points, every week, is an extraordinary claim.

A useful baseline

One number worth carrying around: since 2006, home teams have covered the spread 48.81% of the time. Not 50% — the market slightly over-prices home field, and has for two decades.

Before getting excited about any situational angle — a team off a bye, a Thursday game, a revenge spot — check it against that baseline rather than against 50%. Almost every slice of the board reports something near 48.81% on the home side, and its mirror near 51.2%, which is still under the 52.38% you need. Angles that look significant against 50% routinely turn out to be that one fact wearing a costume.

Where the edge actually tends to be

If a point of line is worth 3.13 points of probability and the market is very good at setting that line, the practical implication is that the number and the price are usually a bigger lever than the opinion. Taking +3 instead of +2.5 buys the densest outcome in football. Taking −105 instead of −120 saves three points of required hit rate. Neither requires knowing anything about the game.

The markets built on top of that arithmetic — teasers and half points — are covered in the teaser guide, and the distribution underneath it in NFL key numbers.

Common questions