Bracket Calculator Validation Report

Limits and honest framing

• Reference set is small (36 decks). This is an initial seed; we're expanding incrementally. Larger sets reduce variance. The clearest signal is at B5 (cEDH, unambiguous per WotC) and the rules-crisp boundaries; the precon band is where the framework itself is fuzzy.
• B4 ("Optimized") coverage is deliberately zero. No public source — not WotC, not the cEDH Decklist Database, not any tournament site — publishes a canonical set of "Bracket 4" decks. WotC's announcement explicitly declines to provide example B4 decklists; community labels at B4 are interpretive. Rather than synthesize B4 references and weaken our "every deck has authoritative provenance" claim, we leave the gap and document the standard we'll accept: a B4 reference must (a) link to a publicly hosted decklist (Moxfield, Archidekt, MTGGoldfish), (b) carry independent corroboration of B4 status from at least two non-affiliated sources (e.g., a tournament finish + a published primer + a community tier-list), and (c) not match B5 criteria (cEDH-tier two-card combos, fast-mana density, tutor count). Until those exist for a given deck, we don't include it.
• B3 coverage is 6 decks. WotC-published precons carrying one or more Game Changers (forced to a B3 floor by the framework) plus higher-power Universes Beyond precons — the cleanest authoritative path to B3 references. We're expanding this further.
• B1 ("Exhibition") coverage is now zero — and that is the honest state. Like B4, WotC publishes no canonical B1 decklist, and B1 is defined by intent ("winning is not the primary goal; highly thematic or substandard win conditions"), not by card choices. Our earlier B1 anchors were stock precons, which WotC's framework and independent tools (Moxfield, ScryCheck) place at B2; we re-based them rather than assert an uncited B1. A genuine B1 reference must be a deck explicitly built to prioritize theme over winning.
• cEDH decklists are canonical Moxfield primers from cEDH-DDB tier-list panels. Bracket assignment (B5) is unambiguous per WotC framework. The exact card-by-card list will vary across tournament copies — the primer is the community's reference build at last_verified date.
• The B1/B2 boundary is fuzzy by WotC's own design — and we grade precons to the authoritative default. WotC originally anchored "the average current preconstructed deck" at Bracket 2 (Core), then decoupled precons from a fixed bracket (Oct 2025: "precons span a range of power levels"; reaffirmed Feb 2026). Independent tools (Moxfield auto-bracket, ScryCheck) default stock no-Game-Changer precons to Bracket 2, and no authoritative source classifies them Bracket 1. We therefore set each stock precon's expected bracket to B2 with an in-range B1 floor (range [1,2]) and cite the basis per deck in reference-decks.json — so a B1 or a B2 verdict is in-range either way.
• "Exact" misses live entirely in the fuzzy precon band, not at the rules-crisp boundaries. The engine is 25/25 (100%) bracket-exact on every deck with a single unambiguous expected bracket (B3 forced by a Game Changer; B5 cEDH). Overall bracket-exact is 97.2%; the gap is stock precons landing B2 where our reference allows [1,2] — which WotC explicitly treats as a range, not an engine error.
• We trend slightly high vs independent tools, and we show it. Cross-checked against Moxfield's independent bracket algorithm (see the cross-tool section below), our verdicts agree within ±1 the large majority of the time but lean marginally higher on optimization-heavy decks. We publish the disagreements rather than hide them.
• The B2/B3 line follows WotC exactly — synergy is not a bracket. Per the official framework, only two-card infinite combos (plus Game Changers, mass land denial, or chained extra turns) force Bracket 3. Three-or-more-card combos and high synergy density do not; a deck with one fragile 3-card combo and no Game Changers is correctly Bracket 2. This is the single most common source of "my deck should be higher" disputes, and the engine resolves it by the rules, not by feel.