Rules Coverage & Accuracy

Enforced by the rules engine and covered by unit/golden tests.

• ✓2026-27 official thresholds: cap, luxury tax, first apron, second apron, MLEs, BAE, minimums
• ✓All exception / tier amounts cross-checked against the CBA's own %-of-cap formulas (BAE 3.32%, NT-MLE 9.12%, Room MLE 5.678%, Taxpayer MLE, min-team 90%, max 25/30/35%)
• ✓Trade salary matching — expanded bands (200%+$250k / outgoing+$7.5M / 125%+$250k)
• ✓Apron teams limited to 100% matching (no 110%)
• ✓Below-cap absorption capped at cap+$250k (or standard matching if larger)
• ✓Hard-cap triggers: NT-MLE / BAE / sign-and-trade → first apron; Taxpayer MLE → second apron
• ✓Second apron: no aggregation (bin-packing combination test), no cash out; exact-boundary tiers use strict 'exceeds'
• ✓Exception gating by apron tier: cap room, NT-MLE, Taxpayer MLE, Room MLE, BAE, minimum, Bird
• ✓Maximum-salary tiers by years of service (25% / 30% / 35%)
• ✓Bird / Early-Bird / Non-Bird re-signing ceilings (175% / 120%-of-prior-or-min, etc.), with each FA's real Bird-rights status and UFA/RFA designation
• ✓Multi-year committed-salary cap sheet (4 seasons per team)
• ✓Veteran extension & extend-and-trade first-year ceilings (140% / 120% of prior-or-estimated-average)
• ✓Renegotiation ceiling (under-cap teams only, raise limited to cap room) and the stretch provision (2N+1 years, 15%-of-cap guardrail)
• ✓Renounce free-agent cap holds to drop below an apron / open cap space; a kept own-FA hold converts to salary on re-sign (no double-count), and renouncing forfeits Bird rights
• ✓Contract extensions add future years at the veteran-extension ceiling (140% rule, 8% raises)
• ✓Free-agent cap holds by Bird status (Non-Bird 120% / Early-Bird 130% / Bird 150–190%)
• ✓Trade eligibility: a free agent signed this offseason is trade-restricted; a just-acquired player can't be aggregated for ~2 months
• ✓MLE / exception consumption tracking within an offseason session
• ✓Sign-and-trade acquisition: second-apron block + first-apron hard cap on the acquiring team
• ✓Base-year compensation: re-sign your own FA to a >20% raise over the cap, then trade — outgoing value = max(50% of new salary, prior salary)
• ✓Trade kicker bonus boosts the acquiring team's incoming matching value (applied when kicker data is present)
• ✓Sign-and-trade outgoing leg: build a return package to the FA's old team — validates both the acquirer's first-apron hard cap and the old team's salary matching
• ✓Pick-ownership ledger: executed trades actually transfer draft picks; boards show real inventory
• ✓Ted Stepien rule against full pick inventory — sees picks traded in PRIOR moves, not just the current proposal
• ✓Restricted free agents: Gilbert Arenas first-year cap (1-2 YOS → NT-MLE) and a real match flow — the original team can match your offer sheet and keep the player at your terms
• ✓Roster limits: 21-player offseason hard cap on signings, 15-by-opening-night warning
• ✓Sign-and-trade contracts enforce the 3–4 season term (§8(e)(1)(ii)) AND the acquirer must have room or match salary with the return package (§8(e)(1)(vii))
• ✓Trade freezes per Art. VII §8(d)/(f): rookie signings 30 days; FA signings Dec 15; over-cap Bird re-signs at >120% until Jan 15; extensions beyond extend-and-trade limits 6 months; matched RFA offer sheets one year (§5(j))
• ✓Renegotiation Mar–Jun blackout window
• ✓Validated against reality: every real July 1, 2026 trade, sign-and-trade, and signing replays as legal through the engine (see lib/realmoves.test.ts)

Directionally correct; limited by available data. Flagged in-app.

• ~Bird sub-type is sourced from public free-agent listings for ~half of free agents; the rest default to full Bird
• ~Rookie-scale salaries are scaled estimates until the official 2026 scale posts
• ~Team salary totals include some non-guaranteed / dead-money rows from the source
• ~Years of service covers ~91% of players; the rest default to a mid-career value
• ~Early-Bird / extension average-salary alternative uses an estimated figure until the official one posts
• ~Trade-value numbers (5–99) estimate production vs. contract — wins produced, dollar-valued, against salary, term, and age — calibrated to real 2026 deals; directional, not gospel
• ~Draft-pick trade values project the origin team’s slot from roster strength (mean-reverting for far-out years) onto a rookie-contract surplus curve — same units as player values

Real CBA rules on the roadmap — the honest gap to front-office-complete.

• ✗Pick protections & swap rights on ledger picks (protections shown in deal text; the ledger tracks unprotected ownership)
• ✗The >10%-renegotiation-blocks-a-later-extension rule
• ✗Traded-player exceptions as separate expiring objects
• ✗Likely vs. unlikely incentives in matching / apron math
• ✗Second-round pick exception & two-way contracts as signing mechanisms (reconciled from real data, not user-simulable)
• ✗Poison-pill provision (engine function exists; not wired to the trade UI — needs structured extension data)
• ✗Designated-player criteria (All-NBA/MVP triggers, 6-year DV extension length)