The Vig You Can Actually See: How Sportsbook Odds Are Really Priced

Introduction

Every sports bettor knows what -110 means in practice — risk $110 to win $100. Far fewer can tell you what it actually costs them over a season. That is strange, because the number is printed on the screen every time a line is posted. Most people just never learned to read it.

Part of the reason is that the number does not sit still. A betting line is a manufactured product, rebuilt continuously by a trading operation running dedicated sports betting software solutions — pulling data feeds, modelling probabilities, and repricing thousands of simultaneous markets. The margin loaded into that price is not a constant. It moves by sport, by market, and sometimes by the minute.

Which means the cost is knowable, but you have to calculate it yourself. This is not a handicapping piece and it will not tell you how to beat anything. It is an argument that a betting market should be read the way any other priced product is read — starting with what the seller is charging. The sports coverage here already argues in those terms whenever it debates contracts, trades, and what a team overpaid for. Odds deserve the same scrutiny.

Start with the arithmetic

Take the standard American football spread: -110 on both sides.

Convert -110 to an implied probability and you get 110 ÷ 210, or 52.38%. Do the same for the other side and you get 52.38% again. Both sides together add up to 104.76%.

Probabilities cannot sum to more than 100%. That extra 4.76% is the overround — the amount by which the book's stated probabilities exceed certainty. Expressed as a share of total handle, it works out to roughly 4.55%.

That is the theoretical hold on a balanced -110 market. A coin-flip spread at standard pricing costs about 4.5% of everything you put through it.

Now do it on a line you are actually looking at. If both sides sum to 110%, you are paying closer to 9%. The arithmetic takes fifteen seconds and almost nobody does it.

Why the hold is only theoretical

One important complication, and it cuts in the bettor's favour.

That 4.55% assumes proportionate action at the quoted prices. A sportsbook only realises something close to that theoretical margin if the distribution of stakes and outcomes works in its favour. It frequently does not.

A book carrying lopsided exposure on a popular team is not collecting a guaranteed margin — it is taking on additional risk, and its actual result will depend on the outcome. This is one reason lines can move even without new information about the teams. A book shading a number to attract money onto the quiet side may be managing its own exposure rather than simply reassessing the game.

The margin is what the book builds into the price. What it actually keeps depends on how wagers are distributed and how the events are settled.

The margin is not the same everywhere. This is the part many bettors never internalise. The 4.55% figure is a useful reference point, not a universal standard.

Heavily traded markets are generally cheaper. Major NFL sides and totals at competitive books can carry relatively low margins. Volume and competition tend to compress the price — the book cannot charge as much when the same market is widely quoted.

Obscure markets can be expensive. Lower-division football, niche props, and minor tournaments can carry 8% or more. Fewer people are watching, fewer people are comparing or arbitraging the price, and the book may be pricing with less confidence.

Three-way markets stack differently. A football match priced 1/X/2 has three implied probabilities to add up, and the overround is spread across all three. The total margin is often higher than on a comparable two-way market.

Parlays compound it. Three independent legs at -110 each have a true probability of 12.5% if every leg is genuinely a coin flip. Fair payout would be 7:1. A 6:1 offer produces an expected loss of roughly 12.5% under those assumptions — substantially higher than betting the same three games individually.

Live markets price wider. In-play odds often carry more margin than the same market pregame. Part of that reflects the difficulty of repricing constantly as the event unfolds, as well as the operator protecting itself against latency.

Uncertainty is the thread running through all of it, and totals markets show it clearly. Scoring outcomes can carry substantial variance. On 18 July 2023, twelve MLB teams scored double-digit runs on the same night — the first time that had happened since 1894, with three games ending 11-10. Books do not price for that specific night. They price for a range of possible outcomes across every total on the board, and that uncertainty is part of what the margin reflects.

The same bet at two prices

Here is the practical consequence of all of the above.

"The Chiefs -3" is not one product. It is a materially different product at -105 than it is at -115. Same team, same game, same outcome — different cost of entry.

A bettor placing a hundred wagers a season at -115 instead of -105 is paying more for the same selection. No improvement in handicapping is required to identify the difference; it is visible simply by comparing the prices available. The regulars in the sports betting forum will tell you the same thing, usually at some length.

Comparing prices is one of the simplest ways to see how much a sportsbook is charging for the same underlying wager, even though it can feel more like admin than strategy.

Why latency shows up in the price

The in-play point deserves unpacking, because it is where the machinery becomes visible to the customer.

A pregame line can be set at leisure. A live betting market has to be repriced continuously — every goal, every red card, every break of serve, across thousands of simultaneous events, fast enough that the number on screen still reflects the state of play.

An operator whose data feed lags by several seconds is exposed to anyone watching a faster stream. Common defences include widening the margin or suspending the market entirely, which is why in-play prices can be worse and why markets often freeze at exactly the moment customers want to bet.

An operator with faster, more reliable data may be able to price more tightly and keep markets open longer. Put plainly: some of the additional margin in a live market compensates for the extra pricing and latency risk involved. Better infrastructure can reduce some of that exposure, and that can affect the odds customers are quoted.

The practical version

None of this makes anyone a winning bettor. It just makes the cost easier to see.

Adding up the implied probabilities shows how much margin has been built into the market. If both sides of a two-way market sum to 104.76%, the built-in cost is around the level associated with standard -110 pricing. If they sum to 110%, the margin is substantially higher.

Parlays can carry a higher effective margin than equivalent straight bets, particularly when several priced legs are combined.

Comparing prices across books can also show how much the cost of the same wager varies from one operator to another.

And in-play betting should be understood as a product whose immediacy and rapidly changing pricing environment can come with a wider margin.

Conclusion

Sports betting has never advertised its mathematics particularly loudly, and that can make the pricing easy to overlook. But the cost is not hidden in a back office or buried in terms and conditions — it is sitting in the odds, in plain sight, waiting for someone to add the implied probabilities together.

It takes fifteen seconds. This is a readership that has never been shy about doing the arithmetic.