Legacy Archetypes

58 archetypes ranked by meta share

Metagame concentration

How few decks Legacy is, month by month, against every ban on record. Each point is the chance that two decks drawn at random from that month's field are the same archetype (the Herfindahl–Hirschman index), shown as the number of decks a field that concentrated is equivalent to. How this is calculated

27.8

effective decks · Oct 2026

0.0360

HHI

−5.1

vs Sep 2026

11.2–32.9

range over 22 months

Lower means more concentrated. Months pooling fewer than 200 classified decks, or fewer than two events, are left blank rather than plotted — below that the reading moves more with how much data the month held than with the format. Decks the classifier could not name are excluded; each point's tooltip reports what share of the field that left.

Tier history — top archetypes

The 8 biggest archetypes by meta share, on one axis. Each line is the score its tier is computed from, so a ban or a set release shows up as the whole field rearranging.

The Y axis is the score each archetype's tier is computed from: for measured points a 60/40 blend of MTGGoldfish meta share and TopDeck top-8 appearances, for reconstructed ones the meta share our classifier derives from standings. Dashed horizontal lines mark the tier thresholds.

The shaded span left of the dashed vertical line is reconstructed from historical standings (issue #214), not measured at the time; those segments are drawn dashed, and they break rather than join at the changeover because no single reading bridges the two. The two eras are scored against different denominators — a reconstructed share is a slice of our classifier's archetypes, a measured one a slice of MTGGoldfish's curated list — so compare shapes across the line, not values. D tier is reconstructed-only — the live writer floors at C — so a line can step D→C at the changeover without its score moving.