NBA Advanced Player Metric

AUTH — Authority Rating

Created by Evan Lynch

A first-principles composite metric built from three independent pillars: offensive authority, defensive authority, and win contribution — each constructed from raw box-score inputs, z-scored within season, then blended and adjusted for team quality and minutes played.

—
Players Qualified
—
Seasons Covered
—
Top AUTH · Final Seasons
—
MVP-tier · Final Seasons

What is AUTH?

Core Design

AUTH does not re-average finished metrics like BPM or Win Shares. It is built from three independent pillars — offensive authority, defensive authority, and win contribution — each assembled from raw box-score inputs. Every input stat is z-scored within its season before combining, eliminating scale differences between stats and making AUTH peer-relative across eras. A score of 8.0 in 2023-24 represents the same peer percentile as 8.0 in 2025-26.

The three pillars are combined under adjustable weights and then multiplied by a team quality adjustment — players on weaker teams receive a mild amplification because their individual production isn't inflated by elite teammates. The final score is linearly scaled so that a league-average qualifier lands near 5.0. Historic outlier seasons exceed 10.

A minutes-played factor then scales the score by availability: players get essentially full credit at ~2,000+ minutes and about 72% credit near the 600-minute floor, so elite rate stats posted in a few cherry-picked minutes can't outrank full-time starters.

In-progress seasons (marked *) are handled on a pace basis: the minutes floor becomes max(150, 600 × games ÷ 82), minutes are scaled to an 82-game pace before the factor is applied, and VORP / Win Shares are measured per team game so teams that have played fewer games aren't penalized. Early-season samples are noisy, so these scores are provisional and are excluded from Top Performers and the headline KPIs.

Pillar 1 · Offensive Authority
How effectively does the player create and convert offense?
OBPM ×0.40 — offensive pts per 100 poss above avg. The anchor: captures scoring, playmaking, and turnovers in one rate stat.

TS% ×0.25 — true shooting efficiency across all shot types. Rewards efficient scorers regardless of shot diet.

PER ×0.15 — new. Player Efficiency Rating captures free throw generation, offensive rebounding, and full scoring volume — elements OBPM systematically underweights.

AST% ×0.10 — reduced from 0.20. OBPM already prices playmaking; this adds the signal without doubling the PG premium.

TOV% ×−0.10 — strengthened from −0.05. High-volume creators with sloppy handles are now meaningfully penalized.
Pillar 2 · Defensive Authority
How much does the player protect the rim and disrupt the offense?
DBPM ×0.60 — increased from 0.55. The most reliable single defensive stat: captures positioning, rotations, and help defense that counting stats miss entirely.

STL% ×0.20 — z-scored within position group. A PG posting 2.0 STL% is now correctly valued above a C with the same rate — guards and bigs face different defensive assignments.

BLK% ×0.20 — increased from 0.15, also position-neutralized. Rim deterrence and shot alteration, fairly rewarded across all five positions.

DRB% removed — a center's high defensive rebound rate is structural, not authority. It was adding a size bias that DBPM already accounts for internally.
Pillar 3 · Win Contribution
How many wins did the player actually generate for their team?
VORP ×0.60 — increased from 0.50. The single best "value" stat — availability × impact above replacement level. Now the true anchor of win contribution.

WS ×0.40 — increased from 0.35. Win Shares grounds AUTH in real team outcomes, independent of the box-score inputs used in the other pillars.

DWS removed — defensive win shares was double-counting what DEF_AUTH already captures. Removing it sharpens each pillar's independence and eliminates circularity.
Full Formula
OFF_AUTH = 0.40×z(OBPM) + 0.25×z(TS%) + 0.15×z(PER) + 0.10×z(AST%) − 0.10×z(TOV%)
DEF_AUTH = 0.60×z(DBPM) + 0.20×zpos(STL%) + 0.20×zpos(BLK%)
WIN_AUTH = 0.60×z(VORP) + 0.40×z(WS)
────────────────────────────────────────────────────────────
RAW = (w₁ × OFF) + (w₂ × DEF) + (w₃ × WIN) ← z() = within-season normalization
TQ = 1.0 + (0.500 − team_win_pct) × 0.15 ← that season's record; amplifies impact on weak teams
MP_F = min(1, 0.72 + 0.28 / (1 + e−0.006·(MP_eq − 1400))) ← minutes credit; MP_eq = MP × 82 ÷ team games played
AUTH = ((RAW × TQ) + 2.5) / 5.0 × 10.0 × MP_F ← avg full-time qualifier ≈ 5.0, historic peak ≈ 11–12
MVP-tier: 8.5+ ~top 2%
Elite: 7.0–8.5 ~top 8%
All-Star: 6.4–7.0 ~top 15%
Starter: below 6.4

Customize Pillar Weights

Offensive Authority40%
OBPM · TS% · AST% · USG% · TOV%
Defensive Authority30%
DBPM · STL% · BLK% · DRB%
Win Contribution30%
VORP · WS · DWS

Top Performers

Analytics

AUTH score distribution by season
Dashed lines mark tier thresholds (6.4 · 7.0 · 8.5)
Top 15 players — AUTH score
Filtered view · color encodes tier
AUTH vs Offensive BPM — scatter
Color = AUTH tier · hover for player details
MVP-tier (8.5+)
Elite (7.0–8.5)
All-Star (6.4–7.0)
Starter (<6.4)
Average AUTH by position
Filtered view · under current pillar weights

Player Database

Season
Team
Position
Min tier
Search player Rank shown is position in the full filtered pool
— results
# Player Season Pos G MIN VORP BPM WS DBPM USG% OFF▸ DEF▸ WIN▸ AUTH Tier
Player Analysis Tool

Compare Players

Search up to 4 players and compare their AUTH scores, pillar breakdowns, and raw stats side by side. Select a season per player to analyze any year in the dataset.

⚡ Quick Insights
Roster Construction Tool

Trade Machine

Build trades between 2–5 teams — players and draft picks — and check legality against the current CBA: salary matching tiers, tax aprons, and hard-cap triggers, calculated live.

📅 Salary data as of November 2026 (Basketball-Reference contracts).
🏀 Draft pick data via RealGM, current through 2032-33.
$165.0M
Salary Cap
$201.0M
Luxury Tax
$209.0M
First Apron
$222.0M
Second Apron
2 of 5 teams
⚡ AI Trade Analysis Beyond legality — what this trade actually means Player value uses each player's most recent completed season (rookies: current season on an 82-game pace)
🏆 Wins Impact
📋 Roster Fit (post-trade)
💰 Cap Flexibility
🎯 Draft Capital
⏳ Timeline Shift
⚖️ Trade Balance
CBA Rules Reference — 2026-27 Season
Below Salary CapCan absorb salary freely up to the cap + $100K. No matching required.
Cap to First ApronTiered matching: outgoing ≤$7.5M → up to 2× + $250K. $7.5M–$29M → outgoing +$7.5M. >$29M → 125% + $250K.
First Apron ($209M)Matching tightens to 110% of outgoing. Cannot send cash. Sign-and-trade only if it lands the team below the apron.
Second Apron ($222M)Matching capped at 100% — cannot take back more salary than sent out. Cannot aggregate 2+ outgoing player salaries. No trade exceptions from prior years.
Stepien RuleA team can never go two consecutive future draft years without controlling at least one 1st-round pick — own, swapped, or acquired unconditionally. Protected incoming picks don't count toward satisfying this rule, since they may not convey.
Draft PicksPicks carry no salary cap impact and don't count toward aggregation restrictions. Protection status is shown for picks acquired via trade — many carry lottery protections or complex conditional swap language not fully modeled here.
Multi-Team TradesEach team's matching legality is evaluated on its net incoming vs. outgoing player salary across the whole deal — exactly how real multi-team trades are structured and approved by the league office.
Visual Analysis Tool

Scatter Lab

Pick any two stats for the X and Y axes. Tap or click a dot (or a name in the list under the chart) to see who it is. Dashed lines mark the average of the players shown, splitting the chart into four quadrants. Save any chart as an image.

Quick start
Swipe the chart sideways to see all of it · tap any dot for details
Players on this chart