Market makers and sharp books

Most US bettors interact only with retail-facing sportsbooks: FanDuel, DraftKings, BetMGM, Caesars, Fanatics. Above that layer is a smaller, less visible category of market makers and sharp books that operate on different economics. Understanding the structural difference is what makes line movement legible.

The retail book model

Retail books are entertainment companies wearing financial-products clothing. Their customer base is heavily recreational. They optimize for customer acquisition (bonuses, brand spend) and customer retention (parlays, promotions, app design). Their hold runs 4 to 5% on majors, higher on parlays and props. They limit or restrict winning customers; this is well-documented and structural to the model.

The line at a retail book is a managed price. The book starts close to consensus, then moves to balance the action from its specific customer base. Because that customer base is mostly recreational, retail lines often diverge predictably from sharp consensus on popular sides.

The market-maker model

Market-maker books take action from professional bettors. The two most prominent globally are Pinnacle (legal in many countries, not most US states) and BetCRIS (offshore Costa Rica, not legal in the US). In the US, Circa Sports operates closest to this model.

Their hold is much lower (around 2% on majors). Their limits are higher. They take losing positions when the sharp action lands lopsided. They use the inflow of sharp money as price discovery, then publish lines that are close to the actual market consensus.

Why this matters even if you cannot bet at a sharp book

Most US bettors cannot legally bet at Pinnacle or BetCRIS. That does not mean their lines are irrelevant. The opposite. Their lines are the cleanest reference for what the actual market thinks the fair price is.

Practical use:

1. Watch the sharp consensus line. If FanDuel is at -3.5 -110 and Pinnacle is at -3.5 -103, the sharp price is closer to even money. The retail book is charging a markup.
2. Look for divergence. If your retail book sits 2 points away from the sharp consensus, the public is on the opposite side of the sharp money on that game.
3. Trust steam from sharp books. When Pinnacle moves first, the rest of the market follows. By the time the move shows at retail, the value at sharp books has compressed.

Circa Sports as the US sharp option

Circa, headquartered in Las Vegas, is the closest thing to a US sharp book. They take limits comparable to a sharp offshore book and run reduced hold. Their NFL early lines (often released hours before competitors) become a benchmark for the rest of the market.

Circa is legal only in Nevada and Colorado at the time of writing. For US bettors outside those states, Circa's lines are visible (via odds aggregators) and useful as a reference even if not bettable. The legal sportsbooks page covers the operating jurisdictions of the major US-legal books: see /sportsbooks.

How sharp books detect informed action

Sharp books welcome informed action because the action sharpens their line. When a known sharp bettor places a $10,000 bet on the underdog, the book reprices. The book is not afraid to lose that bet; the book is using the bet as a price-discovery signal and adjusting both its own line and (indirectly, via market follow-through) the rest of the global line.

Retail books do not operate this way. Retail books usually treat informed action as a problem to suppress, either by lowering limits, voiding bets after the fact, or restricting the account. The two business models lead to opposite postures toward the same customer.

Reading the no-vig consensus

When sharps cite 'fair price' on a line, they usually mean the no-vig price computed from a sharp book. Strip the hold (which is small at sharp books anyway) and you have an estimate of the market's fair probability.

PINNACLE NFL EXAMPLE
  Side A: -103   implied 50.7%
  Side B: -103   implied 50.7%
  total          101.5%   (1.5% hold)

  no-vig fair = ~50% / 50%

  Compare to FanDuel:
  Side A: -110   implied 52.4%
  Side B: -110   implied 52.4%
  total          104.8%   (4.8% hold)
  no-vig fair = ~50% / 50%

  The fair probability is similar.
  The retail book is just charging more juice.

When sharp and retail no-vig prices agree, the retail book is just collecting more hold. When they disagree (e.g., the retail book is 50% / 50% but the sharp book is 53% / 47%), the retail book is mispricing relative to sharp consensus and one side is the value.

What this means for the WagerBird approach

WagerBird's confidence model uses sharp consensus as a primary input. The model reads the line the way a trader reads a market maker. The proprietary specifics of how those signals are weighted are not disclosed; the framework (sharp consensus and bookmaker posture as model inputs) is. See the institutional approach and Terminal.

What to read next

Sharp vs square money covers the daily-action level of the same dynamic. The institutional approach covers the methodology of treating sports betting as a market.