MTG Mulligan Simulator

Shuffle your hand back into the library and draw seven whenever you mulligan. After deciding to keep, put cards on the bottom equal to your counted mulligans. In multiplayer Commander the first mulligan is free, so the retained sizes are seven, seven, six, five, and so on. In a two-player game using ordinary London rules they are seven, six, five, and so on.

The simulator checks the final retained cards. A six-card London hand is selected from a seven-card draw, which gives you a choice about what to bottom. It is not the same distribution as drawing six random cards. Every rejected seven is reshuffled before the next attempt. The simulation stops at your first qualifying hand or your minimum size.

See Wizards' Comprehensive Rules, section 103.5, and our mulligan rules guide. Special opening-hand effects such as Serum Powder, companion actions, and pregame abilities are outside this model.

The main result is the fraction of complete simulated sequences that found a retained hand meeting your rules. Starting-seven probability describes only the initial seven-card draw. At each mulligan depth, the chart shows cumulative success across all sequences, plus the post-bottom pass rate among attempts that reached that depth. Those denominators are different.

If no hand qualifies by your minimum retained size, the sequence is unsuccessful. The simulator does not silently turn that last rejected hand into a successful keep. Average mulligans includes those unsuccessful sequences. The 95% interval describes sampling uncertainty, not uncertainty in classifications or your strategy.

The optional seed reproduces a run with the same deck, roles, conditions, bottoming choice, and simulation count. Inspecting more examples does not change the result. For exact single-category draw probabilities, use the hypergeometric calculator.

Balanced requires two to four lands and at least one nonland with mana value two or less and a printed mana cost. Aggressive requires two to three lands, or exactly one land plus a fast-mana card, together with an early play. These are editable composition rules, not official keep standards or proof that a spell can be cast.

For a cEDH mulligan simulator tailored to your list, choose Custom and combine land, ramp, draw, tutor, and interaction conditions. An OR group can accept either draw or a tutor. An AND group can demand a mana package alongside either option. Selected-card groups can represent commander enablers or specific pieces. To require every card in a combination, add an individual condition for each card; requiring two cards from a group can otherwise include two physical copies of the same name.

Review automatic roles before interpreting a result. Deckcipher reuses its audit heuristics and curated lists; manual corrections are available. Counts indicate cards present, not mana produced or spells available this turn. Color access by turn is deferred because tapped lands, fetch choices, costs, and sequencing need a separate gameplay model. Use the Mana Base Calculator and cEDH Goldfish Simulator for that context.