Rolled-rarity distribution over the last 30 days vs the published tier weights. Exact low-volume counts and distributions are withheld; analysis starts only when samples reach 30.
Loading…
Methodology: actual counts come from the last 30 days of resolved bounty_pulls; expected from the current bounty_pull_tiers.rarity_weights.
Weights can drift over the window (config changes); this dashboard reports current weights. For per-pull verification of historic rolls, use the per-pull consistency checker. Distribution fit does not prove how the server selected any individual seed tuple.