Lands by Curve
"How many lands should I run" is usually answered with a number between 22 and 26 and a shrug. The tables below answer it from the other direction: given the turn your deck needs to be operating, what does the mana actually cost, and what do you get for it.
All odds assume you are on the play in a 60 card deck, so by turn T you have seen your opening seven plus T−1 draws. Every number is an exact hypergeometric probability computed at build time. See also the mana tables and How Many Lands Should You Play?
What reliable land drops would cost
This table asks: to hit your Nth land drop on turn N with a given reliability, how many lands would you need? The answers are the most useful thing on this page, because most of them are numbers no one would ever play.
| Reliability | 2th drop | 3th drop | 4th drop | 5th drop | 6th drop |
|---|---|---|---|---|---|
| 70% | 17 | 22 | 26 | 29 | 32 |
| 80% | 20 | 25 | 29 | 32 | 34 |
| 85% | 22 | 27 | 30 | 33 | 36 |
| 90% | 24 | 29 | 33 | 35 | 38 |
Read that closely, because it overturns the way most players think about mana. Hitting your fourth land drop on turn four just 80 percent of the time would take 29 lands. Doing it 90 percent of the time would take 33. Nobody plays those decks, which means essentially every Magic deck ever built misses land drops regularly, by design.
The practical conclusion is not "play more lands." It is that on-curve mana is not something you can buy, so decks have to be built to function when the mana does not cooperate.
What realistic land counts actually deliver
| Lands | 2th drop | 3th drop | 4th drop | 5th drop | 6th drop |
|---|---|---|---|---|---|
| 16 | 69.1% | 44.9% | 25.0% | 12.0% | 5.1% |
| 18 | 76.4% | 54.7% | 34.3% | 18.9% | 9.3% |
| 20 | 82.4% | 63.8% | 44.1% | 27.3% | 15.2% |
| 22 | 87.2% | 71.9% | 53.9% | 36.7% | 22.9% |
| 24 | 91.0% | 78.9% | 63.2% | 46.7% | 31.9% |
| 26 | 93.8% | 84.6% | 71.6% | 56.7% | 42.0% |
A 24 land deck, the most common count in Constructed, hits its fourth land drop on time about 63 percent of the time and its fifth about 47 percent. If your deck's plan requires five lands on turn five, it fails more often than it succeeds.
Choosing your number
Since perfect reliability is unaffordable, the real question is which land drop you are willing to build around. A workable procedure:
- Identify your critical turn. Not your most expensive card, but the turn your deck needs to be doing its job. For most aggressive decks that is turn three; for midrange, turn four; for control, turn five or six.
- Aim for roughly 60 to 65 percent on that drop. That is the practical band real decks live in, and the table above shows what land count reaches it.
- Then build so a miss is survivable. This matters more than the land count itself. Cheap interaction, a functional two or three drop, and card selection all convert a missed land drop from a loss into a slow start.
As starting points, that reasoning lands roughly where experience already puts most decks: about 20 to 22 lands for a low aggressive curve, 23 to 25 for midrange, and 25 to 27 for control decks that genuinely use five and six mana. Treat these as the output of the table rather than as rules.
Adjustments that genuinely change the number
- Cheap card selection effectively raises your land count without adding lands, which is the whole basis of the Turbo Xerox principle. A deck with several one mana cantrips can cut one to two lands.
- Mana creatures and ramp count as lands for the purpose of reaching a mana total, but not for surviving removal. Count them at something less than a full land.
- Lands that do something else when you are flooded, such as manlands or cycling lands, let you run a higher count with less risk.
- Formats matter. Limited decks with 17 lands in 40 cards and Commander decks with 36 in 99 follow the same maths at different scales. See the Commander tables and the Pauper tables.
Spotted an error, or want a table this page does not cover? Write us at contact@bopmtg.net.