NOTE: This article is from 2018 and is based on outdated assumptions. Please use the 2022 update instead!
The creation of a solid mana base is one of the most important aspects of deck building. When choosing the lands for your deck, it is important to know, for example, how many sources of green mana you need to consistently cast Tarmogoyf on turn 2. Or how many sources of black mana you need to consistently cast Liliana of the Veil on turn 3.
To answer these types of questions, I once wrote a frequently referenced article filled with useful mana tables. Today, I have an update to that four-year-old article. While many of my recommendations have stayed the same, I have tweaked the underlying calculations, leading to small changes.
Before describing my methodology in detail, I will start with a quick summary of the results that may be useful as a quick reference. Sorted from low to high, here is how many colored sources you need to cast your spells consistently on curve. For the mana cost, “C” stands for an arbitrary colored mana symbol.
|Mana cost||Example spell||Number of colored sources required|
|60-card deck||40-card deck||99-card deck|
|5C||Drowner of Hope||8||6||14|
|4C||Through the Breach||9||6||15|
|1CC||Liliana of the Veil||18||12||29|
|CC||Lord of Atlantis||20||14||33|
Note that some numbers differ from my original article. For example, I now only recommend 16 sources for a 2CC card, whereas this used to be 18. This is caused by my updated assumptions. My new recommendations also match successful tournament decks more closely. For example, a glance at recent Grand Prix Top 8 deck lists reveals that top-performing decks typically run Vraska’s Contempt off of 16 black sources.
To explain the methodology behind the above table, let’s start with a recap.
How did I originally define “consistently cast”?
The first thing I did in the original article was to define what it meant to “consistently cast” a certain spell on a certain turn. As I argued, due to the randomness of your draws, 100% consistency is practically unattainable. Instead, I proposed that it is sufficient to have access to the required mana 90% of the time on the play.
This notion of 90% consistency, which was mostly based on experience and intuition, was then sharpened further. I argued that it should be based on keepable opening hands only. After all, you would mulligan no-land or all-land hands in practice. Moreover, the consistency number should be conditional on drawing a certain amount of lands. If your red-black deck can’t cast Goblin Chainwhirler on turn 3 because you only drew two lands, then that’s caused by not drawing enough lands—it doesn’t mean your Mountain/Swamp ratio was off. My calculations took these aspects into account.
Since this meant that the analysis of the mana base became intertwined with the mulligan strategy and the numbers of lands, I made the following assumptions (which are retained in this updated article):
- A 40-card deck contains 17 lands, a 60-card deck contains 24 lands, and a 99-card deck contains 40 lands.
- The only colored mana sources are lands, i.e., there is no Noble Hierarch or Aether Vial.
- You mulligan any 7-card hand with 0, 1, 6, or 7 lands, any 6-card hand with 0, 1, 5, or 6 lands, and any 5-card hand with 0 or 5 lands. You keep all other hands.
For decks that are merely close to satisfying these assumptions (say, decks with 20-28 lands and 0-8 non-land mana sources) my tables remain useful as solid rules of thumb.
For today’s article, I tweaked my approach and definitions to make the analysis more relevant in several ways. Here is the list of changes, as incorporated in my simulation code.
1. The Vancouver mulligan rule
Since my original article was written, the Vancouver mulligan rule was introduced, which allows players to scry 1 after completing one or more mulligans. This can yield a small increase in mana consistency.
I will assume that in case of a mulligan, we scry toward the required colored source. So if you’re wondering how likely it is to hit a green source for Tarmogoyf on turn 2, then this probability is now based on the assumption that any green land will be scryed to the top after a mulligan and any other card (both off-color lands and spells) will be scryed to the bottom.
2. More detail on the actual probabilities
My original article only gave the information that if you want to cast a Tarmogoyf on turn 2 in a 60-card deck with at least 90% consistency under certain assumptions, then you need at least 13 green sources. But I didn’t tell you whether the difference between 12 and 13 sources was 83% vs. 90% or whether it was 89% vs. 92%. Such information is valuable and insightful, and I will provide it today.
3. Higher consistency required for your top-end
Desiring 90% consistency for every converted mana cost is overly rough. In my experience, it’s a good number for cheap spells, but I would like to have a higher consistency for my top-end.
Consider a 1-drop creature on turn 1. If you are unable to cast it because you lack an untapped source of colored mana for that creature, then that’s not the end of the world. Often, you might then use your first turn to play a tapland instead, possibly even one that can produce the right color on turn 2. If not, then you may draw into the colored source soon enough and fit the 1-drop into your curve on turn 3 or 4. Either way, it’s no big deal, so we don’t need the highest level of consistency.
But if we look at expensive spells, then an inability to cast them due to drawing the wrong lands could be much more impactful. For example, imagine the last card in your hand is Lyra Dawnbringer, but you are unable to cast it because you control four Forest and one Plains. Then that’s a fully wasted turn, and potentially a game-losing disaster. Likewise, if you’re a control deck but can’t cast the game-changing Cleansing Nova on turn 5 because you didn’t draw at least two white sources, then that can easily be the difference between a win and a loss.
For this reason, I want to increase the consistency requirements for spells with higher converted mana costs. Specifically, I will require 90% consistency for 1-mana cards, 91% consistency for 2-mana cards, 92% consistency for 3-mana cards, and so on, up to 95% for 6-mana cards.
4. More adequate conditioning on the number of lands
In my original analysis, for a card costing N mana of a certain color, I determined the probability of drawing at least N colored sources by turn T conditional on drawing at least N lands by that turn. For example, for the whole set of spells with two white mana symbols in their cost (Serra Avenger, Wrath of God, etcetera), I determined the probability of drawing at least two white sources as a function of turn T, conditional on drawing at least two lands by that turn. This turn T did not have to equal the card’s converted mana cost. This general approach gave good guidelines for a wide group of cards, whether you cast them on-curve or off-curve, but this conditioning was maximally relevant for cards like Serra Avenger, Daybreak Coronet, or Runed Halo—cards without generic mana cost that you might cast off-curve.
In hindsight, my choice to use the same approach for Serra Avenger and Wrath of God was too simple. To cast Wrath of God, you need at least four lands, among which you may find the required two white sources. For a 4-mana card, it’s more relevant to condition on drawing at least four lands than on drawing at least two lands. Since mana bases are typically built with the aim of casting spells on curve (i.e., on a turn equal to the spell’s converted mana cost) I adapted my definitions to reflect that.
Hence, in the updated analysis, for a card costing N mana of a certain color with converted mana cost M, I consider the probability of drawing at least N colored sources by turn M on the play conditional on drawing at least M lands by that turn (after mulligans). So I will assume that you cast spells on-curve. This means that for Wrath of God, I now consider the probability of drawing at least two white sources by turn 4 on the play conditional on drawing at least four lands (after mulligans).
In large part due to this change, the analysis under my updated assumptions now generally recommends fewer colored sources for double- or triple-colored cards of 4 or more mana. For example, I now recommend 16 white sources for Wrath of God, whereas this was 18 in my original article.
The numbers: 60-card decks
Okay, now on to the results! The following table will provide the probability of casting a spell with a certain mana cost on-curve on the play, taking into account mulligans, conditioning, and other assumptions as described above.
Click to enlarge.
In bolded green, I highlighted the minimum number of sources to consistently cast a card on curve (where “consistently” means 90% probability for 1-drops, 91% probability on the play for 2-drops, and so on)—those numbers resulted in the table you saw at the beginning of this article.
The numbers: 40-card decks
The numbers: 99-card decks
How to deal with gold cards?
To deal with gold cards, I split up the mana costs in parts, and then increase all mana source requirements by one. I’ll explain my process via three examples in the context of a 60-card deck:
- Emmara, Soul of the Accord. From the tables, we know that a 1G card needs 13 sources of green, a 1W card needs 13 sources of white, and a CC card needs 20 sources of that color. Hence, for Emmara I would recommend 14 sources of green, 14 sources of white, and 21 different sources that can produce green and/or white.
- Vraska, Golgari Queen. From the tables, we know that a 3G card needs 10 sources of green, a 3B card needs 10 sources of black, and a 2CC card needs 16 sources of that color. Hence, for Vraska I would recommend 11 sources of green, 11 sources of black, and 17 different sources that can produce green and/or black. (I would give the same recommendation for Assassin’s Trophy, provided you don’t plan on casting it before hitting four lands.)
- 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.
This reason for increasing all mana requirements by one (which I did not yet give in my original article) is that for gold cards, you need all colors to be present. To illustrate, suppose that you would be 90% sure to have a white source on turn 2, that you would be 90% sure to have a green source on turn 2, and that (for expositional ease) these two probabilities would be independent. Then, you would only be able to cast Emmara, Soul of the Accord 90% * 90% = 81% of the time on curve.
Although this number would be higher in a more precise model that takes into account that these two probabilities are not independent, the basic effect would still be present. To avoid it, my rule of thumb is to increase the individual requirements by one. This is an imprecise hack, but it’s based on the results of a quick simulation, where I found that for 8+ dual land mana bases, increasing a gold card’s mana requirements by one generally brought the numbers close to the consistency I like to see.
As a final note: these adjustments are only necessary when the colored mana consistency of both colors might be an issue. If you’re splashing Justice Strike in a mono-red deck where each land produces red mana, then that’s equivalent to splashing a 1W card. Likewise, if you’re splashing Teferi, Hero of Dominaria in a deck with 18 blue sources, then that’s pretty much the same as splashing a 4W card.
How to deal with other unusual mana costs?
For convoke or delve cards, you should imagine the typical amount of lands that you expect to tap to cast the spell. Then I would use those lands to represent the card’s mana cost for mana base recommendations.
For convoke, this depends on your creature base and how many removal spells your opponent may have. For delve, this depends on how quickly your deck can fill the graveyard. But as an example, you might treat Gurmag Angler as a 1B card in a deck that rapidly fills its graveyard, and you might treat March of the Multitudes as a 3GW card in a deck with plenty of white creatures.
For hybrid spells, you need to have enough combined sources of either color in the hybrid cost. For example, if you have four Watery Grave, five Island, and four Swamp, then that’s 13 sources for Discovery // Dispersal on turn two. Perfect.
How to account for fetchlands?
I usually consider Verdant Catacombs, Flooded Strand, and the like as a full mana source for any color that they might be able to fetch. Assuming that you can fetch shock duals, they are the perfect mana fixers.
For 2-color decks, I would also count Evolving Wilds as a full mana source for both colors.
For decks with 3 or more colors and heavy color requirements in more than 1 color, Evolving Wilds isn’t fully reliable because you have to make a choice. For example, an opening hand with Kitchen Finks, Liliana of the Veil, Chandra, Torch of Defiance, Forest, Swamp, Mountain, and Evolving Wilds is much worse than the opening hand in which Evolving Wilds is replaced by Savage Lands. For a deck with these mana costs, I would roughly count each Evolving Wilds as 2/3rds of a source of each color. Sometimes science is more art than science.
I would roughly count each Field of Ruin as 1/2 a source for any color of basic land in your deck, but only for spells with converted mana cost 5 or higher. This approximation is based on the fact that you may not have time to activate Field of Ruin in the early game and that your opponent may not even control a nonbasic land.
By the way, when building mana bases for Guilds of Ravnica Standard or for Modern, you should not skimp on basic lands. After all, your opponent might have Field of Ruin.
How to account for taplands?
For turn 2 onward, all sources count. This is a simplification, but as long as you don’t have an insane number of taplands, it adequately captures the color consistency of your mana base.
So what’s an acceptable number of taplands? Clearly, you should avoid putting too many of them in your deck because they might adversely affect your chances of curving out. In aggro decks with lots of 1-drops, I would generally recommend you avoid taplands like Selesnya Guildgate or Evolving Wilds completely, though one copy is acceptable.
In decks with few or no 1-drops, you can easily spend your first turn playing a tapland, so it’s more reasonable to include a few. But even for midrange or control decks, I would recommend running no more than 4 Selesnya Guildgate or Evolving Wilds. Otherwise, the risk of drawing multiples becomes too high. If you can’t reach the required number of colored sources without running 5 or more such taplands, then you might want to rethink your deck or simply add more lands. Note that this only refers to 100% taplands without any useful abilities—lands like Temple of Abandon, Shambling Vent, or Memorial to Folly are fine to run in slightly larger numbers.
I normally don’t count shock duals like Temple Garden as taplands. But for control decks, paying too much life may be an issue, so if you plan to build a control deck with 8 shock duals, consider counting that set of lands as 2 taplands for the purpose of figuring out if you have too many.
I sometimes count checklands like Sunpetal Grove as partial taplands, but this depends on your mana base. For decks with 3 or more colors especially, you should keep an eye on the number of lands with the right basic land type. For detailed probabilities, I refer you to this article, but I can give some general guidelines:
- In a deck with 16 or more Plains, Forests, and/or Temple Gardens, you can count Sunpetal Grove as a full untapped source, at least for turn 2 onward.
- In a deck with 12 lands of the right basic type, you can count each Sunpetal Grove as 1/8 of a tapland.
- In a deck with 8 lands of the right basic type, you can count each Sunpetal Grove as 1/4 of a tapland.
- In a deck with 4 lands of the right basic type, you can count each Sunpetal Grove as 1/2 of a tapland.
In Guilds of Ravnica Standard, checklands will usually enter the battlefield untapped from turn 2 onward. The reason for this is that we have shocklands, which have basic land types. As a result, even 3-color decks may be able to incorporate 10-12 checklands, as long as there are not many other taplands.
How to account for mana creatures or card selection spells?
If (contrary to what I initially assumed) the deck contains non-land mana sources as well, the numbers in my tables should still act as good guidelines, but you can count some of these non-land mana sources as (partial) colored sources as well.
For cards with converted mana cost 2 or more, I would generally count fragile mana producers (e.g., Llanowar Elves, Birds of Paradise, or Noble Hierarch) as half a colored mana source for each color they can produce. “Bolt the Bird” has been a saying since the early ’90s, so you have to account for the fact that your mana creature may not survive.
For cheap card draw spells (i.e., cantrips costing 1 or 2 mana), I would count them according to the fraction of your deck that can produce the right colored source. For instance, if you run 15 Swamp in a 60-card deck, then Warlord’s Fury can count as 1/4 of a black source, at least for black cards of converted mana cost 2 or higher. If you run only 10 Swamp in a 60-card deck, then Warlord’s Fury would count as 1/6 of a black source. All of this is assuming that you can consistently cast the card draw spell on turn 1.
You could argue that card draw spells should count for less because you don’t always have the time to cast them, but you could also argue that they should count for more because you may draw into another card draw spell to dig further. For convenience, let’s just say that these two effects cancel each other out and that my way of counting is a decent general guideline.
Scrying or surveiling is not as good as drawing a card. For example, you sometimes have to dig for another effect than a specific colored source. Also, there are diminishing returns. As a rough guideline, I would count cheap scry or surveil effects as follows:
- A cheap scry 1 effect in a deck with 16 black sources counts as approximately 0.2 black sources.
- A cheap scry 1 effect in a deck with 8 black sources counts as approximately 0.1 black sources.
- A cheap scry 2 effect in a deck with 16 black sources counts as approximately 0.3 black sources.
- A cheap scry 2 effect in a deck with 8 black sources counts as approximately 0.15 black sources.
Let’s consider the most popular card draw spells in Modern right now: Opt, Serum Visions, and Faithless Looting. Suppose that you can cast them consistently on turn 1, and that you run 15 or 16 black sources. Then, I would approximately count Opt as 0.45 black sources, Serum Visions as 0.55 black sources, and Faithless Looting as 0.5 black sources (disregarding the flashback), at least for black cards that you plan to cast on turn 2 or later.
Assuming that you have enough green sources to consistently cast them on curve, I would count Farseek, District Guide, or Gift of Paradise as full sources of all colors that they could yield, but only for cards of converted mana cost 4 or higher. They won’t help you cast your off-color 2-drop on turn 2. But they can help you cram the best mid-to-late-game gold cards of multiple guilds into the same deck.
Flower // Flourish in a Selesnya deck counts as a full white and green source for cards costing 2 mana or more. But it also basically counts as a tapland—it’s like Selesnya Guildgate in spell form. As I mentioned before, avoid an over reliance on taplands, especially in decks whose creature curve starts on turn 1.
Example 2-color mana bases for Guilds of Ravnica Standard
Let’s start with a sample mana base for a Selesnya deck:
This is an evenly split mana base with 16 green sources and 16 white sources. This means that you can run Nullhide Ferox and Ajani, Adversary of Tyrants in the same deck. Emmara, Soul of the Accord also fits, and you’re really close to consistently supporting Conclave Cavalier. But there are restrictions.
The above mana base cannot yet support History of Benalia or Thrashing Brontodon. If you would like to consistently cast 3 mana double-colored cards on turn 3, then you need 18 sources of that color. To go from 16 to 18, you could replace some basic lands with Selesnya Guildgate or Flower // Flourish, increase the land count, and/or rely on non-land sources like Llanowar Elves or Adventurous Impulse. You could, of course, stubbornly stick to 16 and accept a lower level of consistency, but I would consider that greedy.
The mana base also cannot fully support Pelt Collector or Legion’s Landing. If you would like to consistently cast 1-drops on turn 1, then you need 14 untapped sources of that color. Since Sunpetal Grove doesn’t count, the evenly split mana base only has 12 sources of that color. And it would be even worse if you had replaced some basics with Selesnya Guildgates.
Another is to accept a lower level of consistency. If a 1-drop is still a reasonable play on turn 2, then sticking to the above mana base is acceptable. Honestly, running 1-drops is doable with 12 sources. My tables indicate that you’ll have the right untapped colored mana source on turn 1 with 86.3% consistency. Not ideal, but not terrible either. Indeed, there are Pro Tour winning aggro decks that found success with less than 90% colored mana consistency on turn 1. In the end, mana consistency always has to be weighed against the pain it can cause, and my tables should be viewed as useful guidelines only.
But my main suggestion for the 1-drop conundrum is to choose one color and skew the mana base toward that color.
This mana base can adequately support Llanowar Elves and/or Pelt Collector. Indeed, it contains 14 untapped green sources, not counting Sunpetal Grove, which is enough to cast green 1-drops consistently on turn 1. Since Sunpetal Grove does count for turn 3, you also have the 18 sources needed to support Thrashing Brontodon.
The list has 14 total white sources, including Sunpetal Grove, which is enough to cast a turn-2 Emmara, Soul of the Accord or a turn-5 Lyra Dawnbringer consistently. But you won’t be able to consistently cast Legion’s Landing, History of Benalia, or Leonin Warleader on-curve. I have seen several brews with both Pelt Collector and History of Benalia, but I am skeptical since this really stretches your mana base.
Steel Leaf Champion, which requires 23 green sources, does not yet fit in the above mana base. You can get there by increasing the land count, by relying on non-land sources like Llanowar Elves or Adventurous Impulse, by replacing a few Plains with Unclaimed Territory, Selesnya Guildgate, or Forest, or by accepting a lower level of consistency. But in a deck with Steel Leaf Champion, I would restrict myself to splashing only single-colored cards for turn 3 or later. For example, splashing Knight of Autumn off of 11 sources is doable in a properly constructed Steel Leaf Champion deck, but you should take care before stretching the mana base further.
Likewise, in a Goblin Chainwhirler deck, it’s possible to splash Aurelia, Exemplar of Justice with four Sacred Foundry, four Clifftop Retreat, one Boros Guildgate, and one Plains. But stretching the mana base further would be greedy. For example, if you want to broaden the white splash with Swiftblade Vindicator, then that would come at the cost of Boros Guildgate (which can screw up the curve for an aggro deck) or Unclaimed Territory (which is far from ideal for a non-tribal deck) or extra Plains (which reduces the reliability of turn-3 Chainwhirlers). I generally favor decks that don’t strain their mana base as much.
Next, here is a sample mana base for a Dimir deck:
There are only 13 untapped blue sources for Thoughtbound Phantasm and only 17 black sources for Blood Operative (assuming that you have something like three Thought Erasure, three Lazav, the Multifarious, and two Discovery // Dispersal, which altogether count as two additional black sources for 3 mana cards). That’s slightly below where I would like to be, which is worrisome, but there is no way of improving the sum of these numbers without adding extra lands. Maybe going up to 25 lands by adding a Swamp might be best in a deck with surveil to avoid mana flood in the late game.
If the creature types had more overlap, I would have included Unclaimed Territory as an extra mana fixer, but it doesn’t fit the creature base of this deck. Unfortunately, this exercise teaches us that if you want to cast both Thoughtbound Phantasm and Blood Operative on-curve, then you have to accept some colored mana inconsistency.
Example 3-color mana bases for Guilds of Ravnica Standard
Here is a sample 24-land mana base for a Grixis deck:
This list has 10 red sources, 18 blue sources, and 16 black sources. Hence, you can run Vraska’s Contempt, Sinister Sabotage, and Nicol Bolas, the Ravager in the same deck. If you increase the land count, you may even be able to fit in more restrictive mana costs or avoid the Evolving Wilds.
Since the shocklands have basic land types, there are enough lands to support the checklands. There are 12 lands (including Evolving Wilds) that can support Dragonskull Summit, and the same is true for Sulfur Falls. Drowned Catacomb even has 14. So I would count the whole group of checklands in this mana base as somewhere between 1 and 2 taplands. Adding the pair of Evolving Wilds, that’s between 3 and 4 taplands total, which is high but still acceptable for a midrange deck.
This list has 10 white sources, 18 blue sources, and 16 black sources, which is the analog of the Grixis deck. But since Azorius and Orzhov are not supported in Guilds of Ravnica, you only have one shock dual. As a result, I needed Meandering River to reach the required number of sources. This leads to an increase in the number of taplands. Moreover, since we only have one shock dual, it’s less likely that the checklands will enter the battlefield untapped. All in all, this mana base is akin to having 7 taplands, which is far too many.
Realistically, you would build this mana base with 26 or 27 lands. After all, if you cut an Meandering River but add one Island and one Plains, you retain the same amount of sources but decrease the number of taplands needed. You would also include several cheap card selection spells to find your colors. Since these can act as colored sources too, you can reduce the number of required taplands further.
Even then, it’s clear that some color combinations are better off in Guilds of Ravnica Standard than others. For 2-color combinations, Izzet, Dimir, Boros, Selesnya, and Golgari are the only ones that currently have a shock dual. It’s not impossible to run the other five guilds, but you have to either accept more taplands, restrict your color requirements, or exploit Unclaimed Territory.
For 3-color combinations, Grixis, Jeskai, Sultai, Naya, and Abzan are the only ones with two shock duals, at least until Ravnica Allegiance is released. I expect that these color combinations will thrive while the others will remain substantially weaker for several months. Having better mana bases is highly correlated with tournament success.
My updated analysis has given useful guidelines for the number of colored sources required for a solid mana base. The table at the beginning of this article is a handy reference that takes into account the Vancouver mulligan and other updated assumptions.