Imports shine, locals decide

Most NBL players cannot be told apart from average on on-court impact, players peak at 26, and imports are the most productive players in the league but two in three are gone within a year.

01 · Impact

Most players cannot be told apart from average

Points scored say little about who moves the scoreboard. Rebuilding every line-up from 54,316 stints since 2017-18 and fitting a ridge regression that separates each player from team-mates and opponents, only 28 of 159 regular players over the last three seasons are consistently above (15) or below (13) average when games are resampled. The other 131 cannot be separated from average. Impact correlates just 0.27 with scoring rate.

Some players rank far higher on impact than on points (Kyle Bowen, Bul Kuol, Shaun Bruce), and some high-volume scorers rate below average on impact. Treat individual names cautiously, because their ranges overlap.

The model is honest about what it can do. Trained on 2017-18 to 2023-24 and predicting every game in 2024-25 and 2025-26, the best ridge setting cuts the error from 21.3 to 20.7 per 100, a 2.9% improvement: modest, but real. Without shrinkage it does worse than home advantage alone. Year to year the same players' raw plus-minus correlates 0.26 and single-season impact 0.24: one season of line-up data is mostly noise, and one-year import contracts make carry-over harder still.

02 · Ageing

NBL players peak at 26

Production rises steeply through the early twenties and peaks at 26 (range 25 to 28). At 20 a player is 6.5 points of Game Score per 40 minutes below peak, at 23 1.9, and at 25 almost there (−0.5). Players hold close to peak through 28, then lose about half a point a year: −1.8 at 30, −2.8 at 33 and −4.1 at 35. The decline is gentler than the rise, so an experienced 30-year-old (−1.8) is still more productive than a typical 22-year-old (−3.9). Minutes follow the same shape, peaking at 27 and dropping steeply after 30.

03 · Imports

Imports shine, and are gone within a year

Imports are the most productive players in the league. Per 40 minutes they post 15.0 Game Score against 12.7 for locals (18.2% more) and score 20.5 points against 16.8. On average 47.5% of each season's top 20 producers per minute are imports, though they are only about 30% of regular players.

Only 36.5% of imports played in the NBL again the next season, against 78.9% of locals. Among the top quarter of producers it is 41.9% against 67.9%. A strong NBL season is a springboard to bigger contracts overseas, so clubs rebuild their import slots almost every year. Imports aged 21 or under are now a pipeline of their own: the best by production are Alexandre Sarr (19.0, age 18), LaMelo Ball (17.3, age 18), Ariel Hukporti (17.0, age 21).

The ladder follows the locals. Across 39 team-seasons, a team's win rate tracks the quality of its local players (correlation 0.58) more closely than the quality of its imports (0.35), and how much a team leans on imports has no relationship with winning (−0.03). Locals play about two-thirds of the minutes, so their depth decides more games.

What this does not show

• Impact names are tentative. Ranges overlap, and the typical uncertainty is about ±5 per 100 possessions. Defensive ratings are noisier than offensive ones because defence leaves fewer traces in play-by-play.
• Imports are defined by nationality. Players whose recorded nationality is not Australia or New Zealand, for 2021-22 to 2024-25, the seasons where nationality is recorded for 85% or more of regular players.
• Game Score is box-score only. It credits scoring, rebounds, assists, steals and blocks, and charges turnovers and misses, but it cannot see most defence.
• The ladder result is small-sample. 39 team-seasons is enough to see that local quality matters more, not to size it precisely.