Clemson @ Duke - Brother Willies

Market	50%	—
House Model (v1)	94%	43.7%

What is this page? This is the deep-dive analysis for a single game. You'll see side-by-side comparisons of what the market thinks, what our house model thinks, and what your personal model thinks. For completed games, you'll also see the final score and outcome.

Game Status and Scores — Games display their current status:
- No badge = Scheduled (hasn't started yet)
- LIVE badge = Currently in progress, with live scores shown next to each team name
- FINAL badge = Game is complete, final scores displayed
Scores are pulled from ESPN (CBB) and CollegeFootballData.com (CFB) during each data refresh cycle. The probability analysis remains visible for completed games so you can compare what the model predicted to what actually happened.

Market Probability — The win percentage implied by the current betting odds. This is what sportsbooks and the collective betting market believe. For example, if Alabama is -300 on the moneyline, the market implies roughly a 75% chance of winning.

House Model Probability — What our model calculates as the actual win probability based on team ratings, home field advantage, and injuries. The model uses a logistic function: it takes the rating difference between the two teams, adds a home field adjustment (3.0 points for CFB, 3.5 for CBB), applies injury impact adjustments, and converts the combined score into a probability. This is an independent calculation — it doesn't look at the market odds.

Your Model Probability — The same calculation but using YOUR custom weight settings from the "My Model" page. If you've cranked up the injury weight, games with key injuries will shift more in your model than in the house model.

Edge — The difference between a model's probability and the market's probability. Example: If the house model says 72% and the market says 65%, the house edge is +7.0%. Positive edge means the model thinks the team is undervalued by the market. Delta shows the gap between your model and the house model.

Line movement — The arrows (▲ ▼) show how the odds have changed since they were first posted. If the spread moves from -3.5 to -7, that's significant movement, usually meaning sharp money or new information has come in. Movement toward a team means the market is becoming more confident in that team.

Injury impacts — Key player injuries are listed with their estimated impact on the game (HIGH / MED / LOW). A starting quarterback being out (HIGH impact = 6% shift) has a much larger effect than a reserve player (LOW impact = 1% shift). These injuries are factored directly into the model's probability calculation.

Confidence — Reflects data quality behind the calculation:
- HIGH = Odds updated within 2 hours AND injury data available
- MED = Odds updated within 12 hours
- LOW = Odds older than 12 hours or no odds at all

Where does this data come from?
- Scores come from ESPN (CBB) or CFBD (CFB), updated on each data refresh cycle.
- Odds and line movement come from The Odds API, which pulls real-time lines from major US sportsbooks (DraftKings, FanDuel, BetMGM, and others). We capture multiple snapshots over time, which is how we show line movement.
- Injury reports are sourced from ESPN, including player status and our estimate of the injury's impact on the game outcome.
- All probabilities and edges are calculated by our model — they are independent estimates, not copied from any sportsbook or external source.

Help

Game Detail