2026 NBA All-Star Teams
The NBA went USA vs. World this year. How do the voted rosters compare to what RAPM would pick?
🏟️ New Format: USA vs. World
The 2026 All-Star Game featured a round-robin tournament between three teams: two American squads (Stars & Stripes) and one international team (World). Four 12-minute games, top two advance to a championship game. No more East vs. West.
🏀 Voted All-Stars
⭐ Team USA Stars
8 players🇺🇸 Team USA Stripes
9 players🌍 Team World
11 players📊 RAPM All-Stars
If we picked All-Stars purely by RAPM (min 5,500 possessions), split USA vs. World:
🇺🇸 RAPM Team USA
16 players🌍 RAPM Team World
12 players📉 Voted All-Stars RAPM Says No
LeBron James — RAPM: -2.4. The brand is still massive, but the on-court impact has cratered this season.
Anthony Edwards — RAPM: -1.1. Highlight-reel offense, but defensive impact dragging him underwater.
De'Aaron Fox — RAPM: -2.3. New team, rough adjustment. The counting stats don't tell the full story.
Scottie Barnes — RAPM: +0.3. Positive but barely. Not in the same tier as the RAPM top 16 USA players.
📈 RAPM Picks the Voters Missed
Derrick White (+2.9) — Elite two-way guard. Doesn't sell jerseys, but our model loves him.
Dyson Daniels (+3.5) — Third-highest RAPM in the entire pool. Defensive impact is absurd.
Franz Wagner (+3.3) — Two-way forward with elite defensive RAPM. Overlooked on a bad Orlando team.
Josh Giddey (+2.7) — Quietly producing real impact in Chicago. RAPM catches what box scores miss.
📐 Methodology
RAPM (Regularized Adjusted Plus/Minus) measures a player's impact on their team's point differential per 100 possessions, controlling for every teammate and opponent on the floor. Decomposed into offensive (O-RAPM) and defensive (D-RAPM) components.
Selection criteria: Minimum 5,500 possessions played in the 2025-26 season, ranked by blended RAPM (rolling + Bayesian prior). USA/World split mirrors the actual 2026 format. Rosters are frozen as picked at the February break; displayed RAPM values track our latest published snapshot.
Why the disagreements? RAPM favors two-way players and penalizes poor defenders. Fan voting rewards name recognition and offensive highlights. Neither method is “correct” — they measure different things. That's what makes the comparison interesting.