Methodology

Which population a share uses

• Published lists: entries with a published list in the period. This is the default, because most sources publish only part of each event (MTGO, for example, publishes top finishers and a sample of 5-0 league lists).
• Top cut: only the seats in each event’s top cut. A Top 8 share is never presented as a share of the field.
• Field: only when we know the full field and have its lists. We do not show it until a source gives us that coverage.

Unknown stays in the total

Lists we cannot assign with confidence are shown as Unclassified and stay in the denominator. In a field of 100 with 20 lists of an archetype and 10 unclassified, that archetype is 20%, not 22.2%.

Entries, not decks

• Two players on the same list are two entries that share one composition. Both count.
• When a source corrects a list, the entry keeps its earlier revisions but counts once.
• When we combine events we add up numerators and denominators. We never average percentages.

Changes and missing data

• Changes are shown in percentage points (pp) against the previous period of the same length, with the same sources and rules.
• Missing data is shown as “—”, never as 0. A share with nothing to measure it on is not shown.
• We never invent start times for events that only publish a date, or results from standings.

cEDH

• Groups are the commander or commander pair, in a canonical order.
• A 4-player pod is four participations. It is never turned into six 1v1 matches. Pods of 3 and 4 are kept apart.
• Pod win rate counts draws in the denominator; with draws, 25% is not the baseline.

Win rates

Match win rate is W / (W + L + D). We do not publish win rates yet: they need complete pairings, which current sources rarely give.