How Sportsbooks Set NFL Player Prop Lines

Step One: The Projection

Everything begins with an expected production estimate, and it’s built from the same components a researcher would use: how many chances the player's role should generate, what those chances have historically been worth, and how the opponent and the expected game script change both. Team play volume, pace, and pass rate set the size of the pie. The player's snap share, route participation, and target or carry share determine his slice. Efficiency history converts that slice into yards, catches, or completions. It's the same opportunity-and-efficiency split described in the four-question framework, viewed from the other side of the counter.

Worked example

Say a book is pricing Jaxon Smith-Njigba's receiving yards prop. The chain from team volume down to one player's number looks like this. Every figure below is invented for the example.

Change one input and the whole chain moves. Drop the target share from 20 percent to 15 and the projection falls to about 40 yards, which is a different prop entirely. That's why a line that looks wrong often reflects disagreement about one assumption instead of about the player.

Step Two: Turning a Projection Into a Number

The calculation above gives an average of about 54 yards. It does not tell us the probability of going over 54.5. Keep 54.5 as a hypothetical posted line for this example; estimating either side's probability requires a distribution of possible outcomes. The half-point avoids a tie for this whole-yard market.

The width of that distribution matters as much as its center. Two players can both project for 54 yards while deserving very different confidence. A possession receiver catching six short passes gets there the same way most weeks. A deep threat gets there with one catch, or misses badly without it. Same line, very different spread of outcomes, which is part of why a receiver's role changes which prop holds up.

Step Three: The Price and the Margin

The book doesn’t charge a commission on top of a bet. It builds the vig into the price instead. Carrying the same example forward, here's what the two sides of that 54.5 line imply once a price is attached in American odds:

A real coin flip would add up to 100 percent. These two add up to about 105, and those extra few points are the book's theoretical hold. The practical consequence is that breaking even at -110 takes winning about 52.4 percent of your bets, not 50.

The price can also move on its own. If that same over drifts from -110 to -125 while the number stays at 54.5, the implied probability goes from 52.4 percent to 55.6 percent. The bar the player has to clear hasn't moved at all. The bet just got more expensive, which is the case people miss when they only compare opening and current numbers.

Strip the vig back out of both prices and what's left is the no-vig price, the implied probability without the book's cut. That's the number bettors use to compare the same prop across different books.

American Odds to Implied Probability

The math scales the same way at any price. Minus odds divide the number by itself plus 100. Plus odds divide 100 by itself plus 100. Here's the implied probability behind ten common prices:

Add both sides of a two-way market together and anything over 100 percent is the book’s hold. A market priced at -110 on both sides holds about 4.8 percent. A market priced at -120 on both sides holds more, since 54.5 percent plus 54.5 percent leaves an even wider gap above 100.

Why the Line Isn't the Book's Honest Guess

A posted number is a price, and prices carry the seller's position. A book that expects heavy public action on a popular player's over may post a slightly higher threshold than its own projection supports, because it would rather sell the popular side at a worse price than balance the book afterwards. Existing exposure on correlated markets can pull a number too. None of this makes the line dishonest. It makes it a different object from a projection. That distinction matters for research: disagreeing with a line isn’t the same as disagreeing with the book's estimate of the player.

Why Props Arrive Late and With Lower Limits

A game total can be posted as soon as the schedule’s known. A player prop can’t, because it depends on the player being active and holding a role, and neither is settled until practice participation and game-status designations are published late in the week. Props also concentrate risk: a single piece of information, such as one inactive teammate, can invalidate a number entirely. Books manage that with later postings and lower betting limits. That thinness is also why prop numbers can be moved by relatively little money. See reading line movement for what a move does and doesn’t tell you.

Sources

• DraftKings Sportsbook, How to Read Odds is an operator explaining its own price format, including what the number next to a bet represents and why both sides are priced below even money.
• FanDuel, Sports Betting Guide covers the same ground from a second operator, including odds formats, bet types, and the margin built into a posted price.
• Nevada Gaming Control Board, Gaming Revenue Information publishes what licensed sportsbooks actually wrote and won by month. It’s the primary-source version of the margin described above, reported by the regulator instead of the books.

The implied-probability figures come from arithmetic on the posted American odds. They aren’t a claim from any source: -110 implies 110 ÷ 210, or about 52.4 percent, and -125 implies 125 ÷ 225, or about 55.6 percent. The projection walkthrough uses invented numbers to show the shape of the calculation. Prop Lab isn’t a sportsbook and has no affiliate or commercial relationship with the operators linked above.

Week 7's biggest line moves, live from the board.

Get the AI Analyst's Read on Every Prop

Every prop on this week's board gets a side and a write-up from our AI Analyst. Pro gives you all of them.