Check the pod before the first draw.

One player creates a room and shares the eight-character code. Two to four players can join. Each player then adds a deck from the Current Deck saved in their browser, pasted text, a small text or CSV file, a public Moxfield URL, or a compatible public Deck Case File. A pasted list or Moxfield deck is resolved against the same card data used by the Commander Deck Audit.

The actual decklist stays in that player's browser while it is analyzed. The room receives only the commander and deck labels, total card count, a compact set of existing audit measurements, and short evidence statements. When at least two comparable audits are ready, Deckcipher measures the largest gap. The result shows one main mismatch, up to two supporting signals, and one question the table can answer together.

The comparison does not add a new hidden power score. Combo pressure, tutor density, fast mana, detected bracket pressure, interaction and protection coverage, early ramp, and optional player intent stay visible as separate dimensions. This makes it possible to understand why a question appeared and to disagree with it using the evidence on screen.

Start with the suggested question, then add any important information the audits cannot know. A compact combo may be present but deliberately avoided. A high tutor count may support a theme instead of a fast win. A commander may attract removal well before its actual deck can present a threat. Players should explain those details in plain language rather than treating the result as a ruling.

A useful Rule 0 conversation covers the pace of the opening turns, whether compact combos are welcome, how much fast mana is expected, whether stax or chaos effects are enjoyable for the group, and how consistently each deck finds its strongest line. The optional intent fields make those preferences visible without mixing them into the detected card facts.

Bracket estimates are conversation context, not official certification. If one list needs a closer review, open the Commander Bracket Calculator. If the biggest difference is answer coverage, the Interaction Checker can explain removal, wipes, stack interaction, graveyard hate, and protection in more detail.

Rules version 2026-02 (February 2026 Commander update). Last reviewed 2026-08-06.

The official bracket rules set no limit on tutors. Deckcipher shows tutor density as a consistency signal, not a bracket violation.

Compact-combo and win-speed evidence receives the most attention because it can change how early the table must interact. Tutors and fast mana are separate signals. Tutors affect how often a deck sees its strongest line, while fast mana can compress the opening turns. Detected bracket pressure describes the audit's overall neighborhood, but it never overrides the underlying evidence.

Interaction and resilience are combined only for the comparison gap. The player cards still show the source totals, and the full deck audit remains the place to inspect individual categories. Early ramp provides a smaller consistency signal. Price is intentionally excluded. An expensive deck is not automatically stronger, and a low-cost deck can still be fast, consistent, and difficult to interact with.

Optional intent answers have the smallest weight and remain clearly labeled as player statements. They matter when players openly want different games, but they do not rewrite detected deck facts. Missing card data lowers confidence. When fewer than three dimensions can be compared safely, Pod Check says the result is insufficient instead of inventing certainty.

The room code is an invitation, not an editing password. Every participant receives a separate private key stored only in that browser. It authorizes updates to that participant's seat and cannot edit another player's seat. The creator also receives a different host key that can close the whole room. Neither key is placed in the room URL.

The service stores no full decklist for Pod Check. A room keeps only the minimum summary needed to render the participant cards and run the comparison. It normally expires after about six hours. When a player leaves, that seat's private key, deck summary, audit summary, and intent are cleared. When the host closes the room or the room expires, every private key is destroyed, so the room can never be opened or read again. The room's stored summary is kept after that point for site-operator review and is deleted manually.

Operational analytics use broad categories such as participant-count bucket, room state, risk category, confidence category, duration bucket, and error category. They do not contain room codes, room identifiers, participant keys, display names, commander names, card names, decklists, or public Case File identifiers. Do not use a display name that you consider sensitive.

First, do not treat a low mismatch as a promise that the game will be balanced. Piloting skill, opening hands, threat assessment, unusual commanders, politics, and hidden strategic choices still matter. Low risk only means the checked audit dimensions are broadly aligned.

Second, do not use a partial Case File as if it contained a full modern audit. Older public snapshots may expose only three headline metrics. Pod Check labels those imports as partial and will refuse to create a confident comparison when too little shared evidence is available. Import the original decklist when possible.

Third, avoid using the result to pressure another player into changing decks. The purpose is to reveal expectations while there is still time to choose a different deck, change the group's agreement, or knowingly accept an unusual matchup. The best outcome is not identical decks. It is a table that understands the likely experience before everyone keeps seven.