How to Use the NBA Start Score

NBA fantasy is a per-game sport dressed up as a per-week competition. NFL gives you one game per player per week — the calculation is clean. NBA gives you 1–4 games per player, on rotating schedules, against defenses that shift in quality from night to night. Most fantasy platforms flatten this into a weekly ranking that hides the game-level information. The NBA Start Score starts there instead.

What the number means

The Score rates each player's expected fantasy output for a single upcoming game, then maps that projection to a 0–100 percentile rank within the startable pool at their position. An 85 means that player is in roughly the top 15% of startable players at their position for that game. A 40 means they're below the midpoint — a bench-level option tonight, regardless of how they've looked recently.

The percentile is position-specific and internally consistent. An 85 for a center and an 85 for a point guard reflect the same thing: a strong start relative to available options at that position. What the score doesn't tell you is how many games each player has this week. That's a separate question — one the NBA Streaming Finder answers.

Why minutes leads everything

There's a version of NBA player evaluation that leads with efficiency metrics — true shooting percentage, points per 36 minutes, usage rate. The Start Score weighs all of those. They're the second layer.

Minutes is the floor of everything in NBA fantasy. A player logging 33 minutes per game with average efficiency produces more fantasy value than a high-efficiency player getting 18 minutes in nearly every scoring system. Efficiency only generates points when it's attached to volume. A player who shoots 58% on two field goal attempts isn't helping your lineup.

The Minutes component uses a trailing 10-game exponentially weighted window, so recent role changes register quickly. A player who averaged 16 minutes across his first three weeks and has logged 29+ in his last six games carries a meaningfully different score than his season average suggests. If a backup just inherited a starting role, you'll see it in the number before most analysts catch up.

Reading the availability flags

The most actionable output on the Start Score isn't the number — it's what's sitting next to it.

Questionable applies a 15% haircut to the full projection. NBA teams manage their rosters more conservatively than NFL teams across an 82-game season. A star flagged questionable before a Wednesday game on a back-to-back carries real sit risk that a similar NFL designation wouldn't.

B2B REST RISK is the flag that matters most. When a player's team played yesterday and plays today, that's a back-to-back. For Tier 1 players — stars with documented load-management histories, typically age 30 or older with heavy minute workloads — the model applies a significant projected-minutes cut, and the score drops accordingly. A player who might score 82 for a standard game can fall into the 50s or lower under a rest risk flag.

This is the most common way NBA fantasy managers get hurt. A high-ranked player sits the second leg of a back-to-back and the manager who didn't check still started them. The flag surfaces that risk before you finalize your lineup.

The underlying criteria for Tier 1 classification — age, minutes workload, known team load-management history — are published in the methodology so the categorization is checkable, not a black box.

Per game versus per week

This is the NBA wrinkle the Start Score doesn't pretend away: a 90-rated player who plays twice this week may be worth less to your roster than a 70-rated player who plays four times.

The Score rates one game. It says nothing about schedule volume. If you're making a weekly streaming decision — picking up someone because they have a three-game week — use the Streaming Finder, which multiplies per-game score by games remaining. If you're making a tonight-or-bench call about a player already on your roster, use the Score.

Those are different questions. The right tool depends on which one you're asking.

Where it's most useful

Start/sit decisions between similar roster options are where the Score is most reliable. If you're deciding between two forwards at the same position with comparable usage, one facing a soft interior defense and one facing a stingy one, the matchup term captures that difference. If you're holding a player returning from a two-game absence with a questionable tag, the Availability flag tells you exactly what haircut the model is applying before you lock your lineup.

The Score is also useful for identifying quiet drops. A player consistently scoring in the 25–35 range — week after week, regardless of opponent — is telling you something about their role that their raw stats might obscure. That's a role problem, not a shooting slump. The Start Score sees through one-game variance in a way that box scores don't.

What it can't tell you

The Score is built on data through the previous day. Lineup decisions announced 90 minutes before tip-off — a coach resting a player who wasn't flagged earlier — fall outside what any pre-game model catches cleanly. Check injury reports before you lock in.

Early in the season, the model leans more heavily on prior-season usage rates. Players whose roles have genuinely shifted — a rookie establishing himself, a veteran whose team situation changed over the summer — carry an explicit lower-confidence flag rather than a falsely precise number. As games accumulate, the rating shifts onto live current-season data, and those flags fade.

The formula is published and every component is visible. When the Score is wrong, the components tell you which input failed — whether the matchup adjustment was too aggressive, whether the minutes projection didn't anticipate a coaching decision, or whether one game just produced variance. That traceability isn't a consolation prize. It's the point of building it this way.

Check today's ratings and component breakdowns at RankFantasy's NBA Start Score.