NBA predictions and matchup analysis
Each estimate removes the bookmaker margin from paired moneyline prices, then averages the resulting probabilities across available sources.
Informational estimate only—not a guarantee or betting recommendation. Prices can change, and this calculation does not independently evaluate injuries, lineups or team performance.
Prediction calculation explained
The fields used for each API-powered matchup card.
| Input | How it is used | Status |
|---|---|---|
| Paired moneyline prices | Converted to implied probabilities with the bookmaker margin removed | Connected |
| Multiple bookmakers | No-vig probabilities are averaged across available sources | Connected |
| Point spread | Shown as supporting market context for the predicted team | Connected when supplied |
| Game total | Shown as scoring-environment context | Connected when supplied |
| Injuries and team statistics | Not included in the current calculation | Separate feed required |
How the predicted winner is calculated
For each bookmaker with both teams listed, decimal moneyline prices are converted into implied probabilities. The two values are normalized to remove the listed margin, then averaged across sources. The team with the higher average becomes the market-based prediction, while the percentage shows the strength of that pricing estimate.
How to read the matchup analysis
The short analysis compares the teams' no-vig probability gap and adds the average spread and total when those markets are available. It does not claim to measure injuries, rest, pace or recent performance unless those datasets are connected and clearly labeled.
Why predictions can change
Prices can move when new information reaches the market. Check the displayed PHT update time and revisit the matchup closer to tip-off. A close probability split means the market sees a more balanced game and should not be presented as a strong prediction.
