LUMA
LeaderboardsPlayersResearch
NBAWNBA

Methodology

Movers

percentile points

Δ = percentile(playoffs) − percentile(regular season)

Movers measures how far a player moves between the regular season and the playoffs. Both percentiles are computed within the same pool, which is the set of players who reached the postseason that year.

Using the playoff pool on both sides is necessary. If the regular-season percentile were taken against the whole league, every playoff participant would appear to fall, because the playoff field is stronger than the league.

The measure is still bounded. A player at the 97th percentile in the regular season has little room to rise and a great deal of room to fall.

The average movement is close to zero once both percentiles are taken within the playoff field. Measured across 1,494 player-postseasons from 1998 to 2026, the mean change is +0.05 percentile points. If the regular-season percentile is instead taken against the whole league, the mean change falls to -15.96, because the playoff field is stronger than the league. This is why both sides use the same pool.

The bias that remains is by tier. Players in the lowest regular-season quartile average +21.8 points and players in the highest average -20.4, because a percentile has a floor and a ceiling. Read a large positive number from a low-rated player, or a large negative number from an elite one, with that in mind.

Qualification250 postseason minutes.
CoverageEvery postseason from 1997-98 to 2025-26.
Known failure modeThe scale is asymmetric by construction. Elite regular-season players can only fall; weak ones can only rise.
ReliabilityUsable across a career or a multi-year window. A single postseason gives a split-half agreement of 0.08 and is not published.
ARCP-ARCImpactBox+ValueMinutes-adjusted ARCMoversPeak-N