NBA Machine Learning Projections

Matchups

Games Today: 0
No matchups found for today.

NBA Machine Learning Projections

Compare baseline projections, same-day snapshot context, player-history proof, and XGBoost probability signals before choosing cash or GPP lineup builds.
Last Updated:

How To Use This Page

Start with the tracked Cash and GPP lineups, then use the player boards to understand which players are model-supported upside, volatile, or downside risk.
1. Start With The Lineups
  • Tracked Cash Lineup favors stable minutes, safer projection range, and lower downside.
  • Tracked GPP Lineup allows more volatility when the ceiling and leverage signals justify it.
  • Use alternates when you want similar builds without locking into the first lineup.
2. Read The Boards
  • Top Upside is where the model sees the strongest over-base opportunity.
  • Middle / Volatile means useful upside may exist, but the signal is less clean.
  • Downside Risk highlights players the model is warning against.
3. Key Metrics
  • Base is the starting projection; Snapshot FP includes same-day context.
  • XGB Delta and Over % show the model's lean versus the base projection.
  • Cash/GPP Score, Ceiling, and Smash are the main action signals.

Member Access Required

Guest visitors can see the page overview and today's matchup widget. XGBoost predictions, upside boards, downside boards, and lineup actions are hidden until login.