How Many Lands Should You Play in cEDH?

Before I got into cEDH, I claimed that most Commander decks should play over 40 lands, and specifically, that precons should have 40 or more lands rather than 38 to teach casual Commander players better deckbuilding heuristics.

Perhaps the thing I understand least about edh players is the number of lands they play. I think 40 is a good minimum. Even if I’m wrong about it being strategically optimal, how is making sure your deck functions not a top priority for social gaming? https://t.co/DjoJ4VPkbN

— Sam Black (@SamuelHBlack) August 24, 2020

Many players took issue with this, since they play fewer lands, some taking it as criticism of their deck, and others probably because they play competitive Commander, and they know that more powerful decks have fewer lands than this and that adding more wouldn't make them stronger. Now that I've started playing cEDH, my decks typically have 27-30 lands. What's going on here? Was I wrong about lands in Commander? I don't think so.

What About in Low-Power Commander?

Low-power Commander as originally intended is a format where games shouldn't end before around turn eight and players should be expected to be able to play lots of cards that cost more than four mana. This kind of game has been described as "battlecruiser-" or "board-game-style" games. When I talk about what *most* Commander decks should do, I'm talking about decks that are built for that kind of game.

I have a theory that the average precon is stronger than the average physical Commander deck that exists somewhere in the world, that there are more kitchen table Commander players that have built decks out of cards they own in relatively small collections that are considerably less powerful than precons than there are players with stronger decks than that. This is not the world observed by more engaged players who attend stores or tweet about Magic.

In games with these very low-powered decks, players don't typically draw extra cards before turn three, and, while it's very important that they have a land to play every turn for at least the first six turns, they likely want to play a land almost every turn of the rest of the game. These are the decks that need more than 40 lands.

Why Does Low Power Commander Require More Lands Than cEDH?

Deck Composition

As power level increases, game length decreases, and decks need to have lower curves on average. As the curve drops and the games get shorter, the number of lands you need decreases.

Depending on what you count, 24-30 of my the nonland spells in my deck are dedicated to making mana, which means only about half of the mana sources in my deck are actually lands, and over half of my deck is dedicated to making mana. I believe that's more than most cEDH decks, but while cEDH decks often play land counts in the high 20s, I think they usually play 40-50 cards that make mana. Realistically, because the first few turns are so important, you need more mana early in the game than you can get by just playing one land each turn. In order to keep up properly and take enough game actions early, I think it makes sense for about half of your mana sources to be lands.

Mulligans

Another major factor here is mulligans. In casual Commander, players probably shouldn't aggressively mulligan for their best cards-; these games often use generous mulligan rules along the lines of "just find a playable hand" with the assumption that you've built your deck reasonably and you aren't looking for specific cards. In cEDH, players do aggressively mulligan for their best cards.

In my Rog/Thras deck, I'm much more likely to keep a hand that has Sol Ring, Mana Crypt, Mystic Remora, Rhystic Study, Dockside Extortionist, or Gaea's Cradle. If I keep any hand with one of those six cards and mulligan any other hand, and I'm willing to go to six cards, I'll have one of those cards 75% of the time. Sometimes I'll have one of those cards and my hand won't be keepable, and sometimes I'll have clearly great hands without any of those cards, but I'm much more likely to have one of those six cards than I am to have a card I don't want in my opening hand because of how they influence my mulligan decisions.

If I have any of those cards, I don't need as many lands as I would in other games because they either make a lot of mana on their own or draw enough cards that I'll probably find enough lands to play even with a relatively low number of lands in my deck.

Playing a land each turn, especially in the early turns, is arguably the strongest/most important thing you can do. Cards like Gemstone Mine, Mox Diamond, and Chrome Mox generate an effect that's like playing a land but cost an extra card, and they're very strong cards. This makes the case that playing a land is more important than drawing a card in the early game, for example. While there are decks that don't need three lands to win, and there are decks that can try to win on the second turn, because there's enough interaction that attempts to win on the first or second turn often won't succeed, I'd contend that roughly every deck wants to play lands on at least the first three turns as a relatively high priority.

With 30 lands, if you draw no additional cards and don't mulligan, you can accomplish this 64% of the time; however, mulligans are a huge factor here: if we assume that you mulligan any hand with zero lands or one land, you'll find a hand with two lands by the time you mulligan to five around 97% of the time, and by your four-card hand ~99% of the time. If we assume that you keep three lands if you have them and mulligan if you have fewer than two lands and you're playing 30 lands, you'll have three lands by turn three with no additional card draw over 94% of the time.

cEDH vs. Two-Player Formats

As a tournament player in two-player formats, it was hard for me to believe that it was correct to play such a low ratio of lands; this is like playing only a little over 18 lands in a 60-card deck or a little over 12 in a 40-card deck, which are both things players almost never do, so how could cEDH be so different? Most of the answer is mulligans.

To start with, the one free mulligan is a really big deal. It massively reduces the odds that you start with an unplayable hand even with ambitious land counts. With 25 lands, only ~57% of your hands with have two or more lands, but with a free mulligan, you'll have two or more lands before you even have to go down a card 82% of the time instead of 57%. Second, with four players in the game it just matters much less if you go down a card.

The natural dynamics of two-player games is that your cards are directly trading with your opponent's cards a lot of the time, and having one extra card makes it a lot more likely that you'll win, but with four players each player plays fewer cards that directly trades, and if you mulligan low you can often rely on the rest of the table to keep each other in check. Third, the best cards in cEDH are so much better than the worst cards that getting rid of a few of the worst cards in order to find more of the best cards is a great exchange, so aggressive mulligans are more rewarded.

My fear related to the third point: if you have to mulligan a lot of your hands because they don't have enough lands, then you can't afford to be as picky about using mulligans to find your best cards, but that brings me back to my original point about the six cards I'm looking for in Rog/Thras: typically, the cards you're looking for in your opening hand are cards that forgive having fewer lands, so you can actually mulligan for both of these things simultaneously.

As an experienced two-player tournament player who now has a lot of experience with cEDH, the point I most want to express to competitive players who are new to Commander is that despite how weird it looks, the math actually supports building decks with a very different approach to mana than you'd see in 60-card formats.

Do We Need to Play so Few Lands?

So you can get away with playing fewer lands, but does that actually mean it's right to do so? If you played more and just kept fewer of them or looked for land lighter hands, would that work better? Simply, no. First and most obviously, you want lands early more than you want them late, but this is much more true when you can expect to draw large numbers of cards, which is a lot more likely in cEDH than it is in typical two-player formats. Once you get some kind of strong card engine going, you'll always have a land to play, and when that happens, you want a low ratio of lands to maximize the power of the extra cards you're drawing.

The other unique consideration is that the minimum deck size is also a maximum deck size, and most decks have a lot of ways to search their library for specific cards. Every unique card you can find with your search effects increases the power of each of those effects, and there are enough effects that getting extra tutorable targets in your deck can make the deck quite a bit stronger. This means that squeezing extra spells into your deck is valuable in a way that doesn't come up as much in most 60-card formats, so you really get rewarded for minimizing the amount of space you use on lands.

So back to my suggestion to play 40 or more lands in casual Commander: a few considerations, such as the extra mulligan, do apply to that format as well as cEDH, but I hope it's clear that most of the considerations that go into playing such low numbers of lands don't apply in casual Commander. At the same time, Commander decks are built all over the spectrum in terms of power levels, curves, and nonland mana sources. There isn't a clear binary where every deck should either have below 30 or about 40 lands; that wouldn't make any sense. Really, I can easily build decks that want under 25 or over 50 lands, and there's no universal rule.

The most important point I want to make here is that players who are new to cEDH will likely have to unlearn what they know about land counts from other formats, and also that if your deck is more powerful than a precon but less powerful than a cEDH deck that the right number of lands for you could be anywhere in this range and will depend on the unique needs of your deck and texture of your games, and can even be influenced by factors outside the decklist, like the implied social contract that informs some of your mulligan range.