Susceptibility, aggregated.

All runs rolled up into per-browser fingerprint cards: how often each attack class succeeded, across everything the range has seen. Higher percentages mean agents fell for that trap class more often.