What Are Prop Bets? How Proposition Markets Work and Why They Matter

Sportsbooks start with a probability estimate—usually a statistical model or an experienced trader’s view of an event’s chance. That raw probability is the foundation, but it rarely appears as a published line.

From probability to price

Conversion follows a few practical steps:

• Model output: a probability for each outcome (e.g., 47% chance of Over).
• Smoothing and rounding: probabilities are adjusted to avoid extreme spikes and converted into convenient price points (standard odds increments or whole-number lines).
• Built-in margin: books add a vig (vigorish) so the sum of implied probabilities exceeds 100% — that gap is the bookmaker’s edge.

Operational concerns then reshape lines in real time. Exposure management (limiting liability on one result), the mix of retail versus sharp money, and risk limits can push a line away from the pure predictive probability. Heavy action from professional bettors typically moves prices more than casual bets.

For readers wanting the arithmetic — how implied odds, margin and rounded lines are calculated step‑by‑step — consult the detailed calculations for player prop odds.

Settlement rules determine whether a prophecy becomes a win, loss, or push — and different books rely on different official sources. Typical providers include league box scores and data vendors such as Elias, Sportradar, Opta/Stats Perform, and sometimes the league’s official report. To see a sportsbook’s preferred feed, check the official stat source used to settle bets.

Common void or tie scenarios

• Overtime: some props include OT, others don’t. Clarify before wagering.
• Partial or abandoned games: early stoppages often void props unless a minimum is specified.
• Stat corrections and overturns: official corrections after the game can flip results.
• Off-field technicalities: bets on player appearances may void if listed but inactive.

Edge cases that flip outcomes often look trivial: a 99-yard receiving game later credited as 100 yards when a stat is reassigned; a first-goal scorer changed after review; or a punt return TD in a game that’s later ruled a dead ball. Even a late stat correction published hours later can turn a losing ticket into a winner — or vice versa.

Habit: always read the sportsbook’s settlement rules for that market. When in doubt, screenshot the rule page before placing a bet.

Line shifts carry two broad messages: public sentiment (volume from casual bettors) and sharp money (few large, informed wagers). Books display both but sometimes smooth or delay changes to manage exposure — small early bets may disappear from the public ledger even if serious information exists.

Watch for information-driven jumps tied to injuries, weather, or inactive lists. Sudden, sustained moves after official team announcements or verified reports are more actionable than slow, scattershot drift.

Key signals to track:

• Sharp money: big size + little ticket count; lines move quickly and stick.
• Public pressure: high ticket count with little size; books may move limits or adjust vig instead of price.
• Information pulses: line spikes right after injury reports or weather alerts often reflect real probability change.

For timing details and common pregame patterns, consult the pre-kickoff movement explainer.

Q&A

Is early value real? Early value can exist but is risky; only persistent, info-driven movement or clear sharp action validates it.

When does movement matter most? Movement matters most within tight windows after credible information or when sharp money forces a price that remains stable into close.

Start by turning the posted price into an implied probability. Convert American or fractional odds to decimal odds, then compute: implied probability = 1 / decimal odds.

Next, remove the bookmaker's vig so probabilities sum to 100%. For two-outcome markets, normalize each implied probability by dividing it by the sum of both implied probabilities. The result is the fair market probability.

Quick EV checklist

• Convert the line to decimal odds and compute implied probability.
• Normalize probabilities to remove vig: fair_prob = implied_prob / sum(implied_probs).
• Compare model probability (internal estimate) to the fair_prob.
• Compute expected value per $1: EV = model_prob * (decimal_odds – 1) – (1 – model_prob).

A simple decision rule: if model_prob > fair_prob and EV > 0, the wager is +EV in expectation. Prefer bets with larger EVs and reasonable edges, not tiny differences that vanish with variance.

For a step-by-step walkthrough with worked numbers, consult a worked example on calculating EV after juice.

Track outcomes and evaluate performance over many bets — statistical edges become meaningful only across a sample, not one-off plays.

Player props are most sensitive to a few clear, measurable inputs. Focus first on four items that regularly change outcomes:

• Snap share — how much of the offense a player is on the field for; a drop usually kills counting props.
• Target share — for receivers and pass-catching backs, targets map directly to yards and catches.
• Usage — running back carries + routes + pass-game role; touchdown chances rise with goal-line usage.
• Game script — expected score margin and pace change opportunities for volume and garbage-time stats.

When a heuristic is enough vs when to model

A simple rule (e.g., take a three-game snap/target average and adjust for injury news) is often adequate for high-volume, stable roles. Build a small model when samples are sparse, roles rotate, props depend on rare events (touchdowns), or cross-game factors matter. Start with opponent-adjusted rates and a game-script simulator; add more features only if backtesting shows consistent edge.

See related deep dives and data sources below to move from rules-of-thumb to reproducible models.