Distributions

TL;DR: Pick a metric and seed to see its exact distribution shape, rarity, and normality diagnostics. Full explanation at the end ↓

Data: exact per-value seed counts over the full archive ( seeds). Bars are exact tallies, not samples. Shape readings emphasize effect size; a census-scale normality p-value would be dominated by microscopic differences. axisSum.* sums all three parallel worlds; axis.{-1,0,1}.* are single-world (0 = Main).

Full explanation and methodology

Every chart is built from exact per-value counts across the full seed archive, not a sample. Low-cardinality metrics show one bar per value; wider metrics are grouped into equal-width ranges for readability.

The summary reports practical shape measures such as skewness, excess kurtosis, and the SD-scaled Q–Q gap. A highlighted seed is placed in population context using its exact histogram position and empirical tail rarity.