similarity.find_similar_players()

Rank candidate players by similarity to target.

Usage

similarity.find_similar_players(
    target,
    candidates,
    n=10,
    use_position=True,
)

target is a dict of career totals (see similarity_score). candidates is a Polars DataFrame with the same career-totals columns, plus a playerID and (optionally) fullName.

Returns the top n candidates sorted by descending similarity score, with an added sim_score column.