Commander uses the partial-Paris mulligan: you draw seven, decide to keep or ship, and on each subsequent mulligan you draw seven again but put one card on the bottom. This is more forgiving than 60-card formats, and that forgiveness is a trap.
What you're actually deciding
A Commander mulligan isn't "is this hand good enough to win?" It's "will I get more out of seven random cards or this specific seven?"
The framework has three questions:
- Mana. Can I cast my early-game plays with what's in this hand? In Commander, "early game" means turns one through four.
- Action. Am I doing something on turn three? Turn four?
- Path to my deck's plan. Do I have any of the pieces my deck wants, a ramp spell, a tutor, my commander, a key engine card?
If two of three are missing, ship.
The land-count trap
Commander hands have more land flooding than other formats because the format is slower and the curves are higher. New players over-keep two-land hands because "I usually draw a third." This is true on average and devastating in the cases where it isn't.
A two-lander on the play with no ramp loses to mana screw 30% of the time. That's three games out of ten you walked into voluntarily.
The simplest discipline: keep three-landers freely, keep two-landers only with at least one ramp spell or strong recovery card, ship one-landers reflexively.
The bottom-card decision
When you do mulligan, the choice of which card to put on the bottom matters more than people realize. The rule of thumb:
- Keep your win conditions, you have access to fewer than the deck's 10–12% rate suggests after one mulligan.
- Bottom your situational answers, they were specific to a board state you no longer have.
- Bottom your most expensive card unless it's your commander's payoff.
When to keep a "bad" hand
There are situations where a marginal hand is correct to keep:
- You're the political-glue player at a four-player table, being present matters more than being optimal.
- Your deck has high redundancy and you have some version of your engine in hand.
- You're going to lose worse with a six-card mulligan than a marginal seven.
In tournament settings, none of these apply and you should ship. In casual settings, all three apply and you should usually keep.