Glossary

Every term the board, war room, and player pages use, in plain English. Anywhere on the site, hover or tap an underlined word to see its definition without leaving the page.

Career tiers

Out of league: Never carved out a real NBA role.
Fringe: Stuck on a roster, but barely: spot minutes, not a rotation spot.
Rotation: A real rotation piece, logging regular minutes in the game plan.
Starter: A full-time starter's workload, night in and night out.
All-Star level: Produced at an All-Star level of impact, whether or not he actually made an All-Star team.
Elite: A tier above All-Star level: the rare, true franchise-cornerstone production peak.

Board calls

STEAL: Drafted well below where the model ranked him this class, a bargain on draft rank.
FAIR: Drafted about where the model ranked him.
REACH: Drafted well above where the model ranked him, a premium on draft rank.
SLEEPER: An undrafted player the model rated as draftable. Steal and reach grade where a player was picked, so they don't fit someone who went undrafted; a sleeper is one the whole league passed on but the model would have taken.

Value

Model value: The model's expected career value for a player, on a 0 to 40 scale: 1 is a fringe player, 8 a starter, 20 an All-Star, 40 an all-time great.
Value gap: Model value minus what his draft slot has historically returned. Positive means the model prices him above his slot.
Slot price: The average career value a draft slot has returned historically. What that pick costs in value terms.
Consensus-implied EV: The career value players ranked where the consensus put him have historically returned. Like slot price, but from the mock-draft consensus instead of his actual pick.
STAR %: The model's best guess at his chance of reaching All-Star level or Elite. The range shows how sure it is: a wide spread like 26 to 80 means low confidence, a narrow one means high.

Ranking

Consensus rank: Where the pre-draft consensus of mock boards ranked the player.
Model rank: Where the model ranks the player within this class by expected value.
Draft-day call: The model's rank and expected value frozen on draft night, before any Summer League evidence. Shown next to the current numbers as the receipt: what the model said then, versus what it says now.
Traded pick: The pick that selected this player changed hands, on draft night or after, so the team shown is where he actually is, not who was on the clock. The popover names the team that traded the pick.

Coverage

Outside coverage: No Division-1 college season on record, so the model cannot grade him. The market's prior is shown instead, as a dashed bar.
Market prior: What his draft position has historically become: the market's expectation, shown as a dashed bar. Not a model opinion. Your notes can update it; the model stays silent.
Insufficient sample: Too little college data to grade reliably, and no earlier qualifying season to fall back on. Market prices only.
Blended sample: His final college season was too small to grade alone, so the model scores a minutes-weighted blend of his last full season plus everything after it. The tiny recent sample still counts, just at its real weight.

War room

Availability: The share of 10,000 simulated drafts where the player is still on the board when your pick arrives.
Surplus: Model value minus this pick's price. Positive means he is worth more than the slot costs.
On the clock: It is your team's turn to pick. Every earlier pick is already spent.
Pick price: The career value this pick has historically returned, its price in value terms.

Scouting

Engine: Carries the offense: high usage, the primary creator.
Connector: Wins without the ball: complements stars rather than needing volume.
Co-star: A high-usage secondary star alongside a lead option.
Late bloom: A career that kept climbing after the four-year window the tiers measure.
All-Star mark: The star marks a real All-Star selection in the player's first four seasons.
Scout note: Free text you write about what you saw. The model reads it and nudges the numbers, capped so a note is evidence, never a veto.
SL-updated: These numbers fold in Summer League box scores as reliability-weighted evidence. July minutes are weighted by sample size, and the hotter the summer, the less each extra game adds — history sets that curve, so a great July informs the call but never decides it.
Prior: The model's numbers before your notes are applied.
Posterior: The model's numbers after your notes nudge them.
Comps: Closest historical college profiles at the player's position, showing the range his profile has produced.