PLAYBASKET.GAMES
NBA DRAFT · 5 PLAYER RUN
Roll a club. Pick any available player. Build a roster strong enough to make a run.
2025–26 snapshot · Ratings are PlayBasket estimates.
Your first pick sets your franchise.
NO CLUB DRAWN · ROSTER
NBA SEASON RUN
30 TEAMS · 82 GAMES · CONFERENCE PLAY-IN · BEST-OF-SEVEN PLAYOFFS
LEAGUE PHASE · FINAL TABLE
TOP 8 · FIRST ROUND
The league table is ready.
—
NBA ROSTER BUILDER
Choose five existing NBA players through random team draws, then see how your roster handles a full NBA season. Build around stars, balance open positions and decide when to spend your single re-roll. A strong lineup improves its game strength; it does not guarantee a title.
Every draw offers the eligible historical roster of one club. Inspect the entire drawn roster at once, with each player’s G, F or C roles and rating shown beside their name. Your first pick sets the franchise your five will replace.
Inspect the conference standings after the regular-season simulation. A top-six finish qualifies directly; seeds seven through ten must survive the Play-In.
Advance into the postseason and watch individual results and cumulative series scores. Reach the conference finals and NBA Finals, or start a new draft after your run ends.
Data and simulation: the pinned 2025–26 dataset contains 30 teams and 502 eligible players. It is a historical played-roster snapshot, not live registration or injury status. Ratings and the five-player replacement model are PlayBasket estimates. A club’s best feasible five is a strength proxy, not a claim about its actual historical starters. Team baselines are fitted to the same captured season, so their fit does not prove future predictive accuracy. Source policy · Report a data issue.
Replay and saving: Use New Draft after the season ends to start again with a fresh team draw and one re-roll. The current lineup stays only in the open page session; closing or reloading the page does not restore it. Identical player IDs and drawn memberships reproduce the same simulated run. This game has no public score submission or leaderboard.
Roll an NBA team and choose one eligible player for an open role. Fill two guard slots, two forward slots and one center slot. Your fifth pick starts the simulated season. This is a roster-building game using existing NBA players, rather than a mock rookie draft. The club on the draw used for your first selected player becomes your franchise. Your five replaces that club in its conference and pinned 82-game schedule. The franchise name appears above the court. Other clubs retain their model strength; selected players are not removed from opponent rosters.
You have one optional team re-roll for the entire draft. Selecting a player locks that pick; there is no undo. A player can be chosen only once, even if the historical dataset records appearances for multiple teams. Only teams with an available player for an open role can be drawn.
The game uses a pinned 2025–26 regular-season snapshot, not live transfers. It preserves 582 statistical identities and 661 played team memberships. The draft pool contains 502 players meeting at least five games and 100 total minutes, with usable source-backed roles. Eighty lower-sample players are retained in the source review but cannot be selected. Profiles for 57 identities use explicitly scoped current-provider supplements and do not certify their exact historical positions.
No. PlayBasket grades use a custom box-score production formula, per-minute normalization, a workload adjustment and a 600-minute prior to reduce small-sample effects. They are game estimates, not official NBA grades, complete defensive-impact measures or validated predictions. No player receives a manual popularity override.
Each conference ranks 15 clubs after 82 regular games. Seeds one through six qualify directly; seeds seven through ten enter the Play-In to decide the final two places. The first round, conference semifinals, conference finals and NBA Finals are best of seven. The simulation uses simplified wins, head-to-head, conference-record and point-difference tiebreaks rather than the complete official multi-team procedure.