Phantom Aria Booster Display [GD04] EV Calculator

Best Evidence — Japanese collation proxy

Evidence

This calculator is based on Best Evidence — Japanese collation proxy reviewed September 3, 2026.

Sources

• Phantom Aria [GD04] official English-US product page - Bandai Co., Ltd. (accessed 2026-09-03)
• TCGCSV category 86 group 24633, sealed product 682302 - TCGCSV (accessed 2026-09-03) — identity and pricing were imported into local generated artifacts; the calculator does not call the raw endpoint.
• GD04 official English 194-printing checklist; identities only - Bandai Co., Ltd. (accessed 2026-09-04)
• GD04 photographed English 24-pack opening, 2026-05-24; European distribution - Gundanium Gateway (Sergeant Shokugan) (accessed 2026-09-04)
• Exact Normal EXB-001 and EXR-001 English identities cataloged in GD01 - TCGCSV (accessed 2026-09-04) — identity and pricing were imported into local generated artifacts; the calculator does not call the raw endpoint.
• Best Evidence — Japanese collation proxy and ordinary-pool classification - The Expected Value (accessed 2026-09-03)

Limitations

• Bandai publishes 13 physical cards per English-US pack and 135+16 card types, but it does not publish the complete ordinary slot probabilities, replacement mapping, or per-card weights.
• TCGCSV category 86 group 24633 contained 194 card products, 194 with a non-null TCGplayer Low field, when reviewed. The 198 price rows split into 104 Normal and 94 Holofoil rows. This establishes current row availability only; without a reviewed model it does not establish expected-mass coverage.
• Normal and Holofoil are marketplace subtype labels, not automatically frame/style Treatments. Every duplicate-number and finish row must be mapped to an exact SKU before it can enter a model.
• The no-chase baseline has no separate English collector product to use instead. Before approval, every +, ++, SP, Link Rare, premium Resource, and other chase SKU and branch must be enumerated and omitted entirely.
• The current English-US publisher page publishes 12+1 cards per pack but no rarity-count table.
• The photographed English wrapper gives 50 Common, 36 Uncommon, 32 Rare, 12 Legend Rare, five Special, six Token and ten Resource types. These are roster counts, not pull rates.
• One photographed English box has eight ordinary LRs, two game parallels, one foil Resource and no SP. Removing the game parallels leaves a conditional 38:8 Rare:LR marginal; it does not establish a population SP rate, seeded-box guarantee or case covariance.
• The 16-type retained Normal support pool is R-031 through R-040, T-021 through T-024 and basic EXB-001/EXR-001, visibly pulled in packs 11 and 19. Import only exact EX products 643300/643301 from group 24221. Equal per-card weights are explicit assumptions.
• Modeling policy approved 2026-09-03: every supported ordinary card in the same slot, rarity, treatment, and finish pool receives one equal positive selection weight. This is a transparent modeling assumption where the publisher or empirical sample gives pool odds but not card-level sheet weights.
• Best Evidence — Japanese collation proxy: exact English wrapper and card-price identities are combined with an owner-approved Japanese-product collation proxy because a complete English rate ledger is not published. This is an estimate, not a factory guarantee.
• Only booster-pack slot counts and rarity probabilities are proxied. Card identities, treatments, TCGplayer Low prices, sealed-product identity, language, region, release date, and any explicitly modeled fixed bonus come from the exact English release.
• The proxy is first-order and product-family-specific. No display or case correlation is inherited, and no target is used as a donor for another target.
• Within each retained English rarity/treatment pool, exact positively priced card SKUs use equal positive weight where publisher sheet weights are unavailable.
• Per pack proxy: 6 Common, 4 Uncommon, two Rare-or-Legend-Rare draws weighted 19/23 Rare and 4/23 Legend Rare after chase removal, plus one Resource/EX Resource/EX Base/Unit Token. The 24-pack English display contains exactly 312 modeled physical cards; its expected ordinary Legend Rare count is 192/23 (about 8.35), not 16.
• Replace this proxy when a complete, attributable English collation ledger or publisher rate table becomes available for the exact release.
• GD04-specific English evidence supplements the program's Japanese proxy label: one photographed European-distributed 24-pack opening has 38 ordinary Rares, eight ordinary Legend Rares, two parallel game cards and one foil Resource. The high-slot marginal conditions on removing the two game parallels, giving 19/23 Rare and 4/23 LR. This single box without SP does not establish a population SP rate or English-US factory or case guarantee.
• The English ordinary roster is 50 Common, 36 Uncommon, 32 Rare and 12 Legend Rare. Exact collector-number-to-product-ID membership excludes all 50 parallel and SP printings before drawing. Publisher checklist and wrapper counts establish identities, not pull probabilities.
• The +1 support pool contains 16 exact Normal English SKUs: R-031 through R-040, T-021 through T-024, and basic EXB-001/EXR-001 from marketplace group 24221 products 643300/643301, visibly pulled in packs 11 and 19. Retained supports receive assumed equal 1/16 weights after premium Resource removal; no support enters a Common game-card position.
• Equal per-card weights, interchangeable high-slot marginals and independent draws are assumptions. Box seeding, uniqueness, pack order and case covariance are not preserved. Draw all 312 cards before Sift or user filters; filters do not backfill. There is no topper, DON!!, purchase promo or starter bonus. Embedded pack EV is box EV divided by 24 including its support position; simulated bands are not factory-collation confidence intervals.
• No-chase redraw policy approved 2026-09-03: Curio, Enchanted, alternate-art, parallel, showcase, serialized, signature, premium-rarity, promotional, and any other card or treatment without a supported pull rate are excluded. A chase hit is redrawn from the applicable supported ordinary pool, so excluded cards occupy no modeled outcome.
• Simulation ordering is draw the complete disclosed no-chase product first, record that draw, and only then apply Sift or another user filter. Sift never changes or backfills the completed modeled draw.