In Development – Building a Deep Mana Base

There is a concept in military and business circles of the Tooth to Tail Ratio. Rather than being a metric for ancient crocodiles, this is the ratio between the active combat forces (“Tooth”) and support services (“Tail”) in your military. On the business side, it can be used as a distinction between R&D plus sales and all the other staff – IT, maintenance, and so forth.

The casual observer tends to think that the military is mostly “tooth.” Sure, there are those dudes on the aircraft carrier who wave at Tom Cruise while Kenny Loggins plays in the background, but it’s really all about the F-14s, right?

We often think about mana bases in Magic in the same light. Sure, they’re a necessary part of the deck, but we’re inclined to just write “and 25 lands” and then move on to the cool stuff, the “tooth” of our deck – the spells.

Except that those F-14s required some fifty hours of maintenance per hour of flight time. Your mana base may not require quite that much attention, but if you turn your brain off and just pile some lands in there, it’ll be very much like being Tom Cruise without all those dudes on the carrier deck.

It won’t end well.

Today I’m going to plunge into the world of mana bases with a particular interest in the idea of color depth versus color coverage — that is, not just which colors your mana base supports, but how well it supports them. I’ll start by explaining a bit about both ideas, then move on to discussing how they impact deck design and deck tweaking, and how they explain why modern RUG decks can be hard to play correctly.

Color coverage versus color depth

I came to this week’s topic as I was trying out a Standard W/B/G (aka “Junk”) build. I liked the deck, but I noticed that it had a curious tendency to falter on its mana despite having reasonably robust color access. With a mana base of 25 lands, my color access looked like this:

In support of a black splash for removal spells in an otherwise W/G deck, this seemed reasonable. Compare that color distribution, for example, with an example W/U/B Caw-Blade deck:

Clearly, their color coverage is somewhat better, but is that the only difference?

Well, no. The real issue here is not the difference in color coverage, but the difference in color depth. Let’s take each of these concepts in turn.

Color coverage

Color coverage is a pretty straightforward idea. How many lands do I have that can provide each color I want to use? Clearly if we’re under some cutoff, we either need to alter the mana base to suit our needs or play ramp spells or mana producers to give us access to the colors we need.

For example, here’s the color coverage for Josh Silvestri’s recent U/W Caw-Blade list:

…and here’s the color coverage for A.J. Sacher’s red-splashing Caw-Blade list:

Just operating from this point of comparison, it looks like the U/W/R deck has managed to slide in access to the color red at little to no expense in terms of access to white and blue. In fact, it even has one more white source than the U/W list.

But that’s color coverage, and as I’ve emphasized above, it’s not enough.

Color depth

Where color coverage talks about how many of our colors we have access to, color depth refers to how much access we have to those colors simultaneously.

Consider the very basic example of the following pair of highly theoretical mana bases:

Simple R/G


Dual R/G


These mana bases have the same color coverage profile:

But you can probably tell, at a glance, that they’d play out very differently – and this is what was ambushing me in that Junk mana base earlier. The difference between these two mana bases is in the color density in those “dual” lands.

Whenever I play and use a Wooded Foothills or an Evolving Wilds, I have to choose to select either a Forest or a Mountain. Although these lands give me access to both colors, they have an effective “color density” of one – at the end of the day, they give me one color.

In contrast, when I play and use a Copperline Gorge, I have access to both colors for that turn and every turn afterward. Gorge and Ravine have an effective “color density” of two – each turn, they give me access to both colors.

How color depth impacts our building and refining

The idea of color coverage versus color depth may not be especially profound on an intuitive level, but I’d bet that you’re not consistently tracking it during your actual deck building…or, perhaps more critically, when you modify an existing design.

Tools for mana base building

Over time, I’ve found myself doing more things with spreadsheets – including in Magic territory. Spreadsheets help automate certain aspects of deck design so that we can catch issues like accidentally modifying a mana base until it no longer works. This might come off as too much effort for the job, but it’s really a matter of a little up front investment to make things much easier on yourself as you go along.

My current mana base spreadsheet looks like this:

It’s still a pretty manual affair, but the essential point is that I can type in the lands I want to use along with a few associated numbers, and it reports back to me about color coverage and color depth.

“Number of lands in each color” reports on color coverage. In the example mana base, it lets me know that I have access to 18 white sources, 16 blue sources, and 12 red sources. If you’re having trouble finding that part, it’s the first line with colored entries, below the big list of lands.

“Density in color” reports on color depth. It tells you, effectively, how often you have access to that color in an overlapping manner with another color. Higher is better.

You can skip this part if you don’t care, but here’s what density actually reports on:

Density in color = ( sum of all ((# of Land N) x (color density for Land N)) for all land types) / (total land count)

In other words, if your mana base were entirely composed of dual lands, your density in each of the two colors would be “2.”

So a low density in a color means we often have to choose to have that color or have some other color. A high density in a color means we frequently get to have both that color and some other color at the same time.

I’m going to use these numbers in our discussion going forward, but if you don’t feel like monkeying with spreadsheets, don’t worry – the principles apply no matter what, and I’ll end with an easy “rule of thumb” test for the same thing.

Color depth in action

Earlier, I ran out the color coverage numbers for Silvestri’s U/W Caw-Blade and Sacher’s U/W/R Caw-Blade. As a reminder, here are those numbers one more time:

Now, take a look at their density in those colors:

What do these numbers tell us?

First, we can see that the U/W mana base tends, more often than not, to simultaneously give us access to both colors. This makes sense – it runs twelve dual lands. The U/W/R list gives us simultaneous access to white and blue slightly less often, but it’s still reasonable. Again, this makes sense – it runs eight duals lands, which is not as impressive as the pure U/W list, but is still pretty good.

However, the color density for red in the U/W/R list is an abysmal 0.4.

But we know that A.J. did well with this list. So what gives? Is this a useless stat?

The practical significance of color depth

That low color density stat for red in Sacher’s Caw-Blade list means that if he wanted access to red, he had to choose not to have access to white or blue. In contrast, the higher values for white and blue mean that he frequently didn’t have to choose between white and blue when he made a land drop, getting access to both.

This makes sense in light of the deck’s actual mana base:

U/W/R Caw-Blade mana base


The access to red in this list comes from Mountains, Arid Mesas, Evolving Wilds, and Scalding Tarns. That’s one basic land and three types of fetch. Thus, your choices related to red access are:

Play a Mountain
Choose to fetch a Mountain instead of an Island (Scalding Tarn)
Choose to fetch a Mountain instead of a Plains (Arid Mesa)
Choose to fetch a Mountain instead of an Island or a Plains (Evolving Wilds)

…and this works pretty well for the list because the red cards it’s trying to cast are:


Access to a single red is plenty if you’re Bolting and Sparkmaging all day. That Inferno Titan is a little suspect, except it’s a one-of and it comes in at six mana, meaning you can probably afford to choose a Mountain over and Island or Plains by the time you’re at your sixth land drop.

But what you really couldn’t do in this deck would be to run [card]Goblin Ruinblaster[/card]s. Sure, the color coverage is decent, but the lack of density in red means that either:

1) You’re never kicking them


2) You’re going to accidentally cut off access to some of your other cards

The first case is obvious – if you have one red, you can’t pay RR to kick Ruinblaster. However, the second case is the real point. A deck like this one frequently offers you the choice to pick access to red over access to blue or white. That’s the point of the color density stat – it tells us, roughly, whether or not we have to pick access to that color over access to any other color in our mana base.

Say you cavalierly stick some Ruinblasters in your U/W/R Caw-Blade deck, and on your third turn you crack an Evolving Wilds for a Mountain so you can Ruinblast your opponent’s Valakut. If that leaves you with these lands:



…then you’ve cut yourself off from these cards:

4 Jace, the Mind Sculptor

If you had an Island instead of that Plains, you’d be cut off from:


Neither one of these choices is very hot, and they provide a compelling argument for not running Ruinblasters (as, indeed, people don’t).

Color depth in deck design

The case for keeping track of color depth during deck design is pretty obvious. You want to make sure you’re building a deck that can actually cast its spells. A mana base like the U/W/R Caw-Blade list above can support double white and double blue simultaneously, but can’t carry the burden of a lot of double red – not without screwing up how the deck plays.

Here’s the color density for that Junk mana base I was talking about earlier:

…and suddenly, the problem is clear. Even though I wasn’t attempting to cast any spells with more than a single black in the cost, the mana base as I’d built it required too many choices between its constituent colors. I frequently had to choose black rather than one of the other colors, and all too often had to choose between green and white as well. The end result of all those choices was a deck that, in playtesting, often found itself locked out of a second green or a second white, cutting off access to a whole swath of spells.

So our general rule is that, absent other forms of mana fixing, we want to make sure we aren’t trying to cast spells with double or triple color in their casting cost if we have a low density in that color.

It’s an intuitively obvious rule on the face of it, but it’s all to easy to confuse color coverage for color depth and find ourselves playing a deck that can handle all sorts of mana costs…just not all at once.

Color depth in deck tweaking

Whereas mana bases earn a lot more of our attention when we’re designing a new deck, they’re way down on our radar when we’re tweaking one. Would you have tossed Ruinblasters into that Caw-Blade deck? Maybe. It can make a lotta red, and it already has an Inferno Titan, right?

In tweaking a deck, it’s easy to accidentally “creep” the design into a state where it no longer has enough color depth to support all the cards you’re stuffing into it. This is very much like how we sometimes screw up the curve in a deck as we attempt to rejigger it to fit the ongoing metagame.

The lesson here isn’t any different than the one about deck design, except that you still need to stay vigilant even if you’re just updating an existing deck. The temptation will always be there to toss in a few extra cards to shore up a matchup, and you may not realize that you’re actually disrupting the flow of the deck.

…or, alternately, it may inform your tweak. Is it critical that you cast those kicked Ruinblasters in your Caw-Go deck, and no other card will do? Then perhaps in sideboarding, you’ll take out either those Jaces or those Days and Gideons, since you know that you’re going to get stuck choosing between Ruinblasters and white cards, or Ruinblasters and blue cards – so why not get ahead of the game and just preemptively ditch one side of that problem?

Color depth in generating play complexity

As a final note, some very effective decks have curious color depth stats. Here’s the mana base for a recent RUG list:

RUG Control mana base


Here’s its color coverage:

Decent enough. Here’s its density in each color:

So, did playing RUG decks ever strike you as hard? These density numbers show some of the reason why. This deck requires choice after choice between color access options…and then wants to cast Garruk (GG), Jace (UU), and activate Raging Ravine (effective RR). As the pilot of the RUG deck, you need to properly sequence your plays such that you get to access the right color combination at the right time – make a mistake and you’ll find yourself cut off.

…and this version of the deck plays Explores, but no Oracles (it’s Bobby Oleksy’s list from the Memphis Open).

The color depth on this list kind of sucks. But the deck is otherwise tremendously powerful – so if you can stay on the ball and sequence your color access properly in each and every game, then you’re going to do well.

In other words, being “color shallow” is not automatically bad…but it does present more opportunities for operator error.

A caveat and a rule to conclude

In closing out this discussion of color depth, there are a few things to mention.

First, I made a throw-away remark up there about “absent other forms of mana fixing.” This is a big deal, of course – if you have reliable ways other than your mana base to get color access, or you can just have “more mana base per turn” than normal via acceleration, you can get away with skimping in terms of color depth…at least to an extent.

Second, if you don’t care for spreadsheets, then you can just keep in mind that “more duals is better” in colors in which you want to play double or even triple-color spells. As a corollary, you should be suspicious about the stability of your mana base if you find you’re leaning mostly on fetchlands to get all your colors.

Finally, if you do like spreadsheets, you can get the one I’ve described today by clicking here.

Hopefully, either the spreadsheet or “rule of thumb” approach will let you tweak your decks to your heart’s content without accidentally locking yourself out of a critical color.

magic (at) alexandershearer.com
parakkum on twitter

31 thoughts on “In Development – Building a Deep Mana Base”

  1. Very good article. Back when I cared a lot, I did very similar things to optimize mana bases in excel and it made a world of difference. This is the kind of thing people need to be reading to get better at magic.

    Awhile back I criticized the site multiple times for lacking content (And from a place of caring), I rescind said comments. The recent video and article content has been very good, beyond my expectations. More care just needs to be paid to times when the players are busy, I suppose.

    TSG drafts don’t cut it.

    Again, very good.

  2. Straight to my ‘Excellent Theory Articles’ bookmark folder. I wonder if this might be the reason Wizards has decided not to have enemy-colored dual lands…

  3. excellent read. people often forgot that a good mana base is the back bone of constructed game. well done.
    a minor nitpick, too many non-basic lands (dual lands or whatever) will kill you especially when there is a horde of goblin ruinblaster running around. you need to balance the ratio of your basic and non basic lands.
    well this article is not talking about constructed play anyway, just a reminder 😛

  4. Hi there Mr. Shearer!

    I am an avid reader of your articles here at Channelfireball; in particular I especially enjoyed your articles which focus on configuring good mana bases for any constructed decks.

    I really like your previous article regarding mana base which was titled Which Lands Can I Run; and I do think that your current article complement your previous article very nicely.

    I am interested to acquire if you do not mind the Excel version of your Mana Base Tools spreadsheet. 🙂

    Really a nice article, Mr. Shearer!

  5. One question which comes to my mind after reading your article above, Mr. Shearer.

    If in general to have a deep Mana Base it is considered quite good to have as many on color dual lands as possible (take for example the 12 UW duals to support a traditional UW Control), I quite sometimes get a feeling that I should use less duals since it is not rare that my duals came into play tapped at crucial moments.

    What do you think regarding this matter, Mr. Shearer?

    Thank you for your insight.

  6. This article tells me Wizards should print enemy-color duals more often, if nothing else. Flavorfully, it makes a bit of sense why they shouldn’t exist. But this is more than just a game for a lot of people, and arbitrarily cutting off access to better manabases just because green and blue don’t get along seems mundane and unfair.

  7. Great article, but you could go even further with this sort of analysis. “Probability of turn 4 Jace” must be a useful stat to know about a deck and it would be really interesting to see how that compares between the UW and UWr versions. If you’re interested in colour depth you could do something like “Probability of supporting both turn 4 Jace and turn 4 DOJ”.

    What I’m trying to get at is it’s not just about how your mana averages out, it’s about how likely it is to let you cast the spells you need to cast.

  8. Thanks for posting this. A very informative read on a topic I never actively thought about!

  9. Some more coverage on the issues of crantrips and fixing would have made the article more complete but otherwise it was a very good article. For example a deck might need high coverage of its fixing/cantripping colors to allow it to access it’s other colors, but depth is not as important if it does not have heavy color commitments in those colors at higher mana costs.

  10. Great article. This is the type of stuff that helps a player take his game to the next level. When you’re dealing with fractions of a percent, play testing often isn’t enough to determine the best mana base. I use a slightly different method but the concept is certainly very sound.
    Big Thumbs Up!

  11. As usual, I loved the piece. Great analysis, good way to think about this.

    The thing about RUG, which is a deck I like a lot, is that it is indeed tricky to play, but you REALLY want a Cobra on turn 2. Not just for acceleration, but for color fixing. When playing against RUG, killing the early Cobra is critical. A huge proportion of the games with RUG are decided by whether or not the RUG player gets to untap with a Cobra on turn 3 or 4.

  12. Very usefull article, specially to realize that i should give up in trying to splash red in faeries for lightning bolt. I have tried all kind of mana bases but never got a good result, especially when you want ot cast things like cryptic command, vampire nighthawk, consume the meek, mutavault, creeping tar pit, thoughtseize first turn or lightning bolt.
    My last approach, anyway, was this:

    4 Secluded Glen
    2 Creeping Tar Pit
    3 Crumbling Necropolis
    4 Blackcleave Cliffs
    2 Darkslick Shores
    2 Cascade Bluffs
    2 Sunken Ruins
    3 Island
    4 Mutavault

    I actually didn’t have problems with the thoughtseizes and the nighthawks, the cryptic comands must of the times were ok but not always. But found myself lacking of red for several turns. So then i thoguht removing creeping tar pits but they are a must in this deck. So i think i will have to find better ways of beating the sword and the mirran crusader (the stag too) that aren’t lightning bolt. What do you think?

  13. Pingback: Channel Fireball: In Development – Building a Deep Mana Base

  14. Thanks as always for the comments, here and over on GiftsUngiven. I’m glad this article had some useful ideas for so many of you, and I hope the simple tools I’ve put together help some of you with your own deck design and optimization.

    @stillmoon – Agreed, and that, like the availability of other mana-fixing options, represents a pressure that lives outside of this kind of analysis. We always start with “what can my lands DO?” and then move on from there. That’s why Kamigawa block play kind of boiled down to “red decks, white decks, and four-color Gifts with Kodama’s Reach and STE.”

    @MIkhail – The potential impact of duals on tempo is a concern; that’s why the current version of the spreadsheet tool, for example, explicitly wants to count your ETBT lands and estimate how often you’re going to be stuck with one or more in your opener. In the past, I’ve also counted the M10 duals as “half” of an ETBT land; that feels not-quite-right, and if I feel incredibly bored or nerdy, I might want to build in a component that rates how “tapped” they are based on how many of them you’re running and how many activating basics you have.

    @Michael – True; right now, I’m sort of ballparking that concept in saying “Hey, you’re deep in white but shallow in blue” and things like that, without trying to ascribe a hard value to “how frequently will I be able to hit 2UU by turn 4?” It would be feasible to do that calculation in some manner, but I was honestly just bridging the gap of “geez, why does this mana base suck so bad in practice?”

    @oyzar – Yeah, I think fixing is at least a whole additional article, if not a series. It layers so much complexity on top of this topic.

    @Greco – I’m glad it was helpful for you already. In terms of beating Swords and Crusaders, I’d direct you to LSV or Josh Howe, both of whom pilot Faeries way better than I do.

  15. Three more issues in color access, one subtle and one purely metagame point – two reinforcing the point that duals are better than fetches in color consistency terms, and one leaning somewhat the other way.

    First, a fetch needs to still have the chosen target available as a basic land. If you already drew 3 mountains and only have 3, you can’t crack an Arid Mesa for red. You might think this would never be an issue because you already got 3 red in that case. But effects like Spreading Seas exist, so it still can arise and matter. Moreover, in ramp decks that use many basic land tutors you can run out of targets quite rapidly, between Cultivates and fetches etc.
    Cracked fetches also shift the remaining color density of the deck, especially for small splashes. This can also be complicated by mill effects e.g. getting hit by 2 Body and Mind sword hits that happen to take all the red targets for your Scalding Tarns.

    The second issue is about effects that shut off fetches or make them harder to play. Right now in Standard that means Leonine Arbiter, and there can be further effects in Extended or for corner case cards (like Archive Trap).

    The issue leaning the other way is Tectonic Edge, which can blow up duals but can’t touch a basic or the fetch that gets it.

    A further point to examine is how vulnerable or robust your mana base is to land destruction effects disrupting what you are trying to do. I’ve been color screwed in a mono White deck in recent standard, in part because I took the single color as an invitation to add more spell like effects in my land base, adding tectonic edges and emeria sky ruins.

    In a game against a RUG deck, the ruins were targetable by enemy tec edges; some of my plains got hit by Spreading Seas, and my own tectonics were available but not providing White. I was kept off double white the entire game, through a land destruction storm of 2 spreading seas, 1 acid slime, 1 tectonic edge, and finally a Frost Titan keeping a plains tapped.

    Naturally that is an outlier. But the point is you cannot assume that the opponent will ignore your lands. The ability to play 1 mountain via a fetch does not guarantee you can play all your splashed single red cards. Because a single Arbiter, Seas, or Slime might just cut you off.

  16. Pingback: links for 2011-03-21 « Blarney Fellow

  17. Pingback: MTGBattlefield

  18. Pingback: links for 2011-03-22 « Blarney Fellow

  19. Pingback: links for 2011-03-23 « Blarney Fellow

  20. Hi,

    Really interesting article, I like building mana bases and this is another tool that can help me. I have a question thought about your spreadsheet. In your example Seachrome coast has a ‘W Density’ of 8, I think this should be 4. If not can you explain why it has to be eight.

    kind regards,

  21. Hey, I just read your article, is very interesting but I have some ideas about it:
    1) Since density is a way to calculate the % of something in a solution, it shouldn’t be mayor than 1 for a given color, unless you have something really odd like a land that can produce 2 mana of a single color without consuming any mana on itself.
    2) Any kind of fetchland doesn’t add to your density in either way, because you won’t be getting mana out of it, you will sacrifice it and change it for a mana producing land.
    Think of the chance to produce any color if you had all your mana sources in play, the fetchlands would’t make any diference in the mana you are producing.
    3) Since fetchland don’t add to your mana production, they don’t change the mana density on the battlefield, they do change the availability of it.
    In the example you have 12 sources of red mana availability, but only 3 red mana producing sources
    4) I propose another way to check mana density on decks:
    1- You count all the lands or sources able to produce one color
    2- Repeat the process for every playing color in the deck
    3- Count the amount of lands that produce mana or sources, do not count fetchlands in this total because the don’t produce mana.
    4- The density of each color is the total of the color in you get in “1-” divided by the total of all the sources in “3-”

    With the previous process you get this numbers for the datasheet example:
    Your density for:
    W: 66%
    U: 61%
    R: 16%
    Your mana availability is:
    W: 66%
    U: 59%
    R: 44%

    With the same process applied to R/G fetch vs RG dual you get this numbers:
    Mana density :
    G: 50%
    R: 50%
    Dual lander:
    G: 66%
    R: 66%
    Mana Availability:
    R: 50%
    G: 50%
    Dual lander:
    R: 50%
    G: 50%

  22. Pingback: In Development – All Your Victories Begin Here : Magic: The Gathering – Strategy, Singles, Cards, Decks

  23. Pingback: » In Development – Lily Bring the Bling

  24. Pingback: MTG Quality | Alexander Shearer : “Building a Deep Mana Base”

  25. Pingback: » In Development – Lily Bring the Bling

Comments are closed.

Scroll to Top