How the 73-9 game is calculated
No black boxes. Here is exactly how the game turns real box-score stats into player ratings, salaries, and a win percentage against the 73-9 Warriors.
Player rating (RTG)
Each player's rating is built from their real per-game averages for that exact season: points, assists, rebounds, and field-goal percentage, plus a nudge from how many minutes they played (a proxy for role and trust). Scoring and playmaking are weighted most heavily, efficiency is rewarded above roughly league-average shooting and penalized below it, and the whole thing is capped so no single player runs away with the scale.
The result lands prime superstars around a rating of 24β30, quality starters in the low-to-mid teens, solid role players around 8β12, and deep-bench players in the low single digits. It's intentionally simple and readable β you can look at a stat line and roughly predict the number.
Salary against the 2026 cap
Salary is a convex function of rating: it climbs gently through the role-player range and then explodes at the top. That mirrors how real NBA contracts work β the difference between a good starter and an MVP is worth far more than the difference between a bench player and a good starter. A rating in the low 20s prices a player near a max slot; the very best seasons ever are priced well beyond it.
This is the mechanism that makes $100M feel tight. Five max-level seasons would cost several times your bankroll, so every roster is a negotiation between talent and thrift.
The season simulation and the Warriors matchup
Your five ratings are added into a single team-strength number. That number feeds a logistic win curve for the 82-game record and a separate head-to-head model against the 2015-16 Warriors, calibrated so that a team-strength around the mid-80s plays them roughly even. Because the Warriors sit near the top 0.1% of team strength ever assembled, most rosters win only a small share of games against them β and getting that number into double digits is a genuine achievement.
The 'best squad you could've signed' uses the same math on the five team-seasons you landed on, solving a small optimization to find the highest-rated legal five under the cap. It's the same wheel results, played perfectly β the benchmark you're chasing.
Related guides
How to play the 73-9 game Β· About the 73-9 game Β· Start the wheel