# How should I determine how many lands of each different color to put in my deck?

Situation: You've built your at least 2 color deck, know exactly what number of lands you want in it, but don't know how many to put in of each basic-land type.

How do pro-players figure this out? Obviously it seems like a deck that is about 50-50 color-wise would have equal numbers of each basic-land for those colors, but what if the deck is 60-40, 70-30, 80-20, etcetera? Or perhaps its a tri-color deck with a 40-30-30 or a 50-40-10 color distribution? Obviously it will depend on the mana-curve and the cards in the deck, but I'm just looking for some general guidelines.

This question addresses how many lands to put in a 40-card deck, and This one addresses the general issues of building a multi-color deck but I'm specifically interested in color distribution of lands. It would be awesome if answers could address both 40 and 60 card decks. I'm talking only about basic lands, as I believe addressing the entire issue of "What lands should I put in my deck?" is far too big for one question.

• Is your question specific to basic lands only? Or should we address cards like Glacial Fortress? Jan 19 '12 at 1:31
• @AlexP Just basic lands for this question, I figure it would be better to address non-basic lands in a different thread. :D Jan 19 '12 at 15:22

As an unscientific method for quickly throwing lands into a draft deck, I could up the number of coloured mana symbols in the casting costs of my cards, and use the results as a ratio to choose my lands. So, if my deck has 20 white mana symbols and 10 blue mana symbols, and I need to select 17 lands, I'd typically be thinking about 11 Plains and 6 Islands.

Though of course it's not quite as simple as that. You also need to factor in coloured activation costs, of course; and then, consider when you need your different colours of mana. In the example given above, if many of the white cards cost 4W or 5W, and all the blue cards cost U or UU, then evidently I need to draw my islands early and can get away with drawing plains later on. That sounds like a pretty peculiar deck, but if for some reason it did exist you could easily be considering even 9 Islands and 8 Plains.

Another example: suppose your deck is pretty much mono-white, with a red splash. Your red mana symbols may be less than 10% of your total mana symbols: you may even only have one red card! Even so, it might not serve you well to dogmatically have 1-2 Mountains and 15-16 Plains. 12 Plains should be more than enough to run your white spells off (assuming none of them cost WWW, anyway!) and you can make up the difference with 5 Mountains, to make sure you have a good chance of playing that red spell in any game you draw it.

Unfortunately I don't think there's any simple magic formula for answering this question: it's just the sort of thing that becomes more and more intuitive the more Magic you play. But one last tip: if you're playing a lot of Limited games where your mana is letting you down, consider the possibility that it might not be your choice of lands that's letting you down at all - maybe you should be trying to assemble decks that are less ambitious about the range of spells they are trying to play!

• It's not exactly simple, but this is about as close to a magic formula (pun intended) as I've ever seen: channelfireball.com/articles/… (and for Modern: channelfireball.com/articles/…). Jan 19 '12 at 0:04
• @David Zaslavsky nice article mention! Although most players seem to just pick up the skill with experience. Jan 19 '12 at 0:09
• Well explained, but don't forget that if you are running color-fixers that can influence it too. Jan 26 '12 at 22:59

Building mana bases is a mix of art and science. Fundamentally, mana is the biggest resource constraint in a Magic deck, so your land choices are just as important as your spell choices during deck design.

## Quick and dirty basic land distribution

When I'm pressed for time (e.g. in Limited) or just throwing together a rough draft of a deck, this is my approach:

• Count mana symbols, not cards. A card costing 2WWW has the same converted mana cost (5) as a card costing 4W, but you'll need to play many more white mana sources to be able to cast the 2WWW card reliably. Counting colored mana symbols in cards is a way to account for this. Some online deckbuilding tools (e.g. TappedOut) do this by default.
• "Color screw" is awful and you should take pains to avoid it. Therefore, if you have uneven color requirements, make the mana base a bit less skewed than the color distribution of your spells. For example, if you're running 70-30 red/white, your might do 60-40 mountains/plains. Most costs have a generic component, so having a few lands of your "secondary" color on the field shouldn't significantly slow down your "main"-color spells.
• By the time you get to playing your six-drops, you'll have seen something like a quarter of your deck; in contrast, you've got at most 8 cards when it's time to make your first-turn play. If you rely on a certain color heavily in the early game, play more of the matching lands to support it.

## Basic lands will only get you so far

If you have a deck that wants to play Grand Abolisher (WW) on turn 2 and Chandra's Phoenix (1RR) on turn 3, no combination of basic lands is going to allow that. You can play cards like Rampant Growth and Evolving Wilds for "color fixing", but the larger problem here is a lack of "color depth" -- the ability to make different combinations of colors with the same couple of mana sources. Building a multi-color deck using only basic lands limits your ability to reliably make use of cards with heavy color commitments, which are often some of the strongest effects for their CMC.

There are different strategies involved for 40 card decks versus 60 card decks.

A 60 card deck is constructed. This means you'll have access to a great number of dual lands and land fetches. A 40 card deck is limited. You will probably have only basic lands to work with, with may be one or two mana fixers you managed to pick up.

It's fairly easy to make the argument that the more cards in a color you have, the more you have to weigh into that color for your land base. That's very simple logic. In fact, it's almost too simple.

The first hole in that logic is that you'll have many more spells of one color in your deck than another. This really shouldn't be the case. If it was a sealed event, chances are your cards are fairly evenly balanced. You'll have 15-20 cards in each color, and more than likely you will play the two colors that both have 11 playables. You're unlikely to have a situation of 18 playables in one color and 5 in another color. If you had, let's just say, 18 red and 5 green, is it reasonable to think you had 5 or less playable cards in white, blue and black? No, it's not. Even if you did have 18 playable cards in red, you're probably still better off only playing the best 11 or 12 of those playables, and playing 9 or 10 white or black or blue or whatever color had the second most playables. If you were to bucket your cards into colors, and then sort each bucket by how playable the card is (not a scientific route, I recognize, but one that many players do) you will probably find that #'s 6-11 in one color are much better than 12-18th place in the main color. So your cards will be balanced anyway if you've built a good deck.

The second hole in that logic is that it doesn't take into account what this humble magic player finds to be the most important aspect of having multiple colored decks: color aligned cards. A color aligned card is one that has multiple colored mana symbols. Most cards in magic are in the form of `{#}{x}` where # is a number and x is a color. In other words, they're like Runeclaw Bear which is `1G` and not like Garruk's Companion which is `GG`. Garruk's Companion is color aligned to green. For it to be effective, you need to have your first two lands be forest. Playing mono green, that's easy. Playing any non-mono-green deck, that's hard. It gets even worse for cards like Leatherback Baloth and Phryexian Obliterator which are very powerful, but are very color aligned. Because most cards only require 1 mana in any particular color, you only need to worry about getting 1 of those lands casting the spell.

In the recent pre-release, I had two copies of Martial Law. Because I had a significant amount of control spells (Dispel Syncopate and friends) I felt the need to be able to cast multiple blue spells in a turn. For this reason, I had 9 islands, 6 plains, and 3 Azorius Guildgate. As it turned out, only once did I have the need to cast multiple counters, and at the same time, twice I had Martial Law and Sunspire Griffin in my hand but couldn't play them due to having only 1 source of white mana. If I could play it over again, I'd've gone 6 Islands, 9 plains to compliment my 3 gates. I had four cards in my deck that were turn 3 or turn 4 color aligned white, with no spells color aligned to blue (except for the `2WWUU` Sphinx, but since she's color aligned to both I didn't see it as a big issue. The likelihood of having 6 mana with only 1 source of either color was highly unlikely).

So here's what I do in limited:

1. I have only one color that allows for color aligned cards. Exceptions to this are if I have an incredible amount of land fixing, or if there is a huge benefit to be gained (for example, Trostani, Selesnya's Voice or Jarad, Golgari Lich Lord. I would also consider making exceptions if I were in a situation where I pulled, for example, Murder and Deadbridge Goliath.) Remember how my pre-release didn't technically have blue-aligned, but I treated it as such because of how important I thought it was to cast two blue spells in a turn? (That's basically the same thing as having a blue aligned spell.) Trying to align to two colors was disastrous for me in that tournament. I went against a primary tenet of my deck building strategy, and I paid a hefty price (0-3 first three rounds, until I finally WTF'd.)
2. Assuming I have no mana fixing, I will go with 9/8 basics for no color aligned cards. The 'winner' (the color with 9) is the one that has more cards in that color. I will go with 11/6 if I have 1 or 2 color aligned cards. I will go with 12/6 if I have three or more color aligned cards. Note that 12/6 adds to 18, not 17 as normal. Because there's so many color aligned spells, I can afford to play more lands. Why 'afford'? Because a color aligned spell in my deck will always be more powerful than a non color aligned spell of the same cost. If it wasn't significantly more powerful (compare Runeclaw Bear and Garruk's Companion again) I wouldn't align.
3. If I have mana fixing that is either mana-free (like dual lands) or is very cheap (like Traveler's Amulet or Transguild Promenade) then it takes away from the biggest mana pool. For example, if I would have gone 12/6, and I get 3 fixers, it instead becomes 9/6 with 3 fixers.
4. If I'm going 3 colors, I do not use color aligned cards. It becomes more difficult for me to set any hard and fast rules because I don't often go three colors in limited. This will probably change this block, since RTR is so color-happy, and perhaps I'll re-evaluate this answer after doing more RTR.

With constructed it's very different. I do not use formulas for constructed mana pool determination. I start with something that seems decent and I play test the deck against my other decks or against friends. With a constructed deck you have the luxury of playing the deck many, many times before you take it to a tournament. For this reason, you do not need to resort to using formulas which are always guesses to what your mana pool needs. The only way to really know if you have a good mana pool is to play test your deck many times. You don't time to do that in a limited environment, so you are forced to use a formula to approximate your ideal mana pool.

Also in constructed, you have access to a wide variety of dual lands and mana fixers. The key is to get the right mana at the right time. When you're play testing, keep track of how many times you want to play a particular card, but can't, and why you can't. Perhaps it's because you don't meet requirements of the card. This happened to me with metalcraft a lot in the Scars block: what was the point of playing a card whose benefit was only worth the mana if I had three artifacts out and I didn't have those out yet? That might mean I need to scrap the deck, or it might just mean I need to do some tweaking. If a card, game after game, can't come out on turn X where X is it's CMC, (and you want it to... you don't usually want to Murder on turn 3...), you either need to fix your mana base or scrap the card. Both are good or bad options depending on the situation, and you can't make hard rules about that. It's that decision making that separates good deckbuilders from great deckbuilders.

• "You're unlikely to have a situation of 18 playables in one color and 5 in another color." It's about more than playables, though. Which cards work together? If I open 16 cards that are good for a weenie aggro strategy in white, why shouldn't I play all of them? Now instead of just looking for "playables" in my other colors I'm prioritizing based on access to aggression and removal. Apr 29 '13 at 17:22
• As noted in the answer, the determination of a playable card isn't a scientific exercise. But if you have 16 good cards for white weenie, you still have at least 7 more spells required for another color. You're (generally) better off not going 16-7, but rather more like 13-10 or 12-11. In otherwords, of those 16 cards, three of them have to be the least synergetic. It's very very likely spells 8, 9 and 10 in that other color will be better than spells 16, 15, and 14 in white. So that's why you shouldn't play all of them. Will you have times when they are? Yeah, it can happen, but unlikely. Apr 29 '13 at 17:56
• I don't think there's a very strong basis for the generalization that you're making. Sealed and draft pools can easily be very skewed, especially in sets where certain archetypes have very strong color associations. Think about infect in SOM block Limited. I think there's really no reason to assume that red card #11 is going to be better than green card #15 consistently enough to make "don't skew your colors" a rule. Apr 29 '13 at 18:22
• Then we'll need to agree to disagree. I've found it to be case enough to make it a "rule of thumb". There's a huge number of other factors, too, of course - double-colored mana cost (e.g. Garruk's Companion), your curve, meta games, etc. Just like I don't use formulas to determine my mana pool, I don't use formulas to determine my card composition either. But I do use it as a rule of thumb and have found that avoiding color skew leads to more wins and a higher average power level. YMMV. Apr 29 '13 at 19:44

There is no definitive formula. Although you specifically ask about color combination it is just part of the bigger problem that is land distribution in a deck.

As @thesunneversets also mentioned, this is the sort of thing that (at least so far) has no simple answer. It is a skill that is refined as you play more and more. Even so, some important questions you have to ask yourself:

1) How many cards of each color (obviously);

2) How many cards that cost more than one colored mana - i.e. if some cards have RR in their cost you have to get two red mana at the same time before you can play those cards.

3) How many one/two casting cost cards of each color - i.e. if all your one drops are red cards, you probably want to make sure you have red mana in your first turn;

Highlighting #3, I would also say this isn't so much a problem of the number of cards in a deck (or format) but a problem of general game plan.

Example: a UR Storm deck in modern only needs red mana to win, but generally packs more blue mana just because it's game plan is to start the game playing cheap blue draw spells in the first turns. All the library manipulation also assure the deck sees the 1/2 red mana sources it needs in a single game.

Another example: Richard Bland's top 8 decklist in GP San Diego 2011 features only 2 really powerful cards that need red: Kessig Wolf Run and Devil's Play. Using Traveler's Amulet and Caravan Vigil he could have two extra "virtual" Mountains, without stressing his manabase that much. Plus, both are late game winning cards, so a single Mountain to enable the strategy seems more than enough.

Hall of Famer Frank Karsten wrote the definitive article on this question recently. I refer you to that article for the details. The gist is that how many lands producing that color you need depends on the casting cost of your cards (a WW card for example is harder to cast than 1W card), and how soon you want to cast them.

Practical example: Adrian Sullivan's Jeskai control deck that won GP Milwaukee. This is a control deck, so not everything needs to be cast on curve, but some spells certainly should be available on demand. For example,

• Deafening Clarion - being able to cast this on turn 3 could be the difference between winning and losing against an aggressive deck. This is a 1RW card. If you check Frank Karsten's table for 60-card decks, you find that to cast a 2C card, you need 11 sources of that mana. So to cast Deafening Clarion on turn 3 consistently, you need 11 white sources and 11 red sources. Adrian Sullivan's list has 16 red sources and 12 white sources, which is more than enough. (This analysis is superficial because some cards produce both red and white mana, and this double-counts them - see the Niv-Mizzet section below for a more sophisticated analysis.)

• Ionize - the thing about this card is that it's simply a 3-mana counterspell. Doing 2 damage is not really relevant for control decks, so why isn't Adrian Sullivan running Sinister Sabotage? The answer is mana. From Frank Karsten's table, to cast a 1CC card consistently on turn 3, you need 18 blue sources. The deck has 17. You could still run Sinister Sabotage, but you would have to tweak the mana base, removing one of the red/white lands for blue lands. This isn't cheap and means you'll have to re-evaluate the other cards in the deck.

• Niv-Mizzet, Parun - at RRRUUU, this is easily the hardest card in the deck to cast. Frank Karsten explicitly treated this card in his article:

Niv-Mizzet, Parun. From the tables, we know that a 3UUU card needs 16 sources of blue and a 3RRR card needs 16 sources of red. So I would recommend 17 blue sources and 17 red sources for Niv-Mizzet. (I would give an analogous recommendation for Golgari Findbroker.) Although I didn’t list CCCCCC cards in my tables, you obviously cannot afford a single source that isn’t either red or blue.

• The deck does have enough blue sources, but is short one red source. There're two reasons the deck gets away with this. First is that the deck doesn't need (and often doesn't want) to cast Niv-Mizzet on turn 6. Ideally the deck casts Niv-Mizzet when it's able to protect it, which means it wants to cast Niv-Mizzet even later in the game. Second, the deck also has Treasure Map which produces treasures that help cast Niv-Mizzet.

If you try using the table yourself you should quickly find that manabases are pretty sensitive things. If you have only basic lands, the chances of you being able to cast your spells - especially your low-mana spells - on curve drops quickly. This is why 40-card limited decks are usually two colors, and why one thing pros pay attention to when they look at a format is what dual lands are available. If you have only basic lands, you'll probably have to play a mono-color deck, possibly splashing for some high-cost cards in the second color only (or have lots of mana fixing).

Edit: as of time of writing, there's a fantastic tool at mtgoncurve which calculates how likely you are to cast any card in any deck, given your mana base. It's not perfect - for example it doesn't realize that Flower // Flourish is effectively a land - but it's still a great simulation for how much you can strain your manabase. You'll want to identify your key cards that you always want to cast on time, and try to fit enough lands and colored sources such that the number in the first column is high, preferably at least >90.