Fig. 2: Bootstraps of challenge scores.

Bootstraps (n = 1000) of the submissions for a SC1, b SC2.1, c SC2.2, and d SC2.3 ordered by submission rank. Boxes correspond to 25th, 50th, and 75th percentile, individual points are displayed beyond 1.5*IQR (interquartile range) from the edge of the box. For each sub-challenge, a baseline model using only demographic or meta-data is displayed in red as a benchmark.