Sharp sportsbooks, and why their prices are the benchmark

A sharp sportsbook is one that sets its own prices, takes large bets from bettors who win, and moves its line in response to them. Its odds, with the margin removed, are the best readily available estimate of an outcome's true probability, because anyone who thinks the price is wrong can bet into it at size, and the price moves until they stop. Positive EV betting is mostly the practice of comparing recreational books against that benchmark.

Two ways to run a sportsbook

A market-making book treats its customers' bets as information. It posts a price, lets winning bettors hit it, and learns from where their money goes. To attract that money it needs low margins and high limits, and to profit on thin margins it needs volume. A retail book runs the opposite model: it often opens from or follows lines set elsewhere, charges a wider margin, spends on promotions to win recreational customers, and restricts accounts that win consistently (the limits guide explains why). Most US-licensed books are closer to the retail model, which is exactly why their prices are sometimes wrong.

The best-documented example

Pinnacle is the book most often cited as the sharp benchmark, and it describes its own model publicly: it advertises low margins, high limits, and a "winners welcome" policy of not restricting customers for winning, and it says its limits rise as an event approaches and more money comes into the market. Pinnacle does not accept customers in the United States, so US bettors use its prices as a reference rather than a place to bet. Other offshore books and some Las Vegas books are also described as sharp; how sharp any one of them is varies by sport and market.

Exchanges are a benchmark too

On an exchange or a prediction market there is no house setting the price. Buyers and sellers do, and the price sits where they meet. Where plenty of money is resting on both sides, that price carries little or no built-in margin and moves on information as fast as any sharp book. Where the order book is thin, a single small order can set the last price, and it tells you very little.

One sharp book, or a consensus

Using a single sharp book as the anchor is simple and transparent, and it works well where that book prices the market with high limits. It works badly where it does not: a sharp book's price on a minor prop may carry low limits and little information. A consensus devigs several reference-grade sources and blends them, which smooths out one book being slow or wrong. The cost is that you are now trusting a weighting, and a blend is only as good as the sources in it. Blending in recreational books pulls the estimate toward the prices you are trying to beat.

Where the benchmark is weakest

Early in the week, before limits rise, even sharp prices are softer. Player props, lower leagues, and alternate lines are priced with lower limits everywhere, so the sharp price there is a weaker signal. And a sharp book can be the stale one: if news breaks and a retail book moves first, the "edge" you see is against a number that is about to change. The EV mistakes guide covers that trap.

How TrueEdge builds its fair price

The Positive EV board devigs every reference-grade venue it collects that prices the whole market (sharp books and exchanges), blends them, and leaves the book being measured out of its own benchmark. Where no reference-grade source prices a market, it falls back to a weighted consensus of the books quoting it and labels the result lower confidence rather than hiding the difference. Tap the fair figure on any row to see which sources and method built it; the walkthrough shows where.