This is another area that I believe needs general improvement. Building manabases is not about fudging things, it is about numbers. A rock-solid manabase will be one of your greatest assets during a long tournament, as it greatly reduces the chance of your deck beating itself.

So, how do we go about constructing a manabase? The first step is simply to determine how many lands you need. To do this, you need to determine how many consecutive land drops you need to hit. A good way to start is to look at average casting cost (ACC). Take the converted mana cost of each non-land card in your deck and average it. This is your average casting cost. If you are building a deck in a computer program, it may have a feature that does this for you.

You normally have to hit land drops up to your ACC consistently (although things like Eldrazi can mess this up). This means that if your ACC is 5, you have to hit your first 5 land drops at minimum. After looking at your ACC it's time to look at your deck composition. Are most of your critical spells at or below your ACC? If so, then it is probably fine to use your ACC as a target. If not, you should probably increase your target number to include more of your crucial (framework) spells.

So now that you have a target number, you can determine how many lands you need to run. Here is a chart for probabilities of hitting your first N land drops (on the play):

If you are on the draw, the percentages are simply a little big higher. This chart should show you that you probably aren't running enough lands.

If you consider 70% to be "consistent" then you need to be running 22 lands to hit your first 3 land drops and 26 to be hitting your first 4. If you consider 75% to be consistent you need to up each of those numbers by 1. You'll note that the effect on probabilities of adding an additional land is roughly linear, although it begins to plateau in the high eighties. This is where the conventional wisdom of running at least 24 lands comes from. That number gives you a reasonable chance of hitting your first 4 land drops, which will cover the ACC of most decks.

As far as limited is concerned, the same principle applies, although each land has a more drastic impact on probabilities. Remember that 16/40 is the same as 24/60, 17/40 is 25.5/60, and 18/40 is the same as 27/60. Thus you can see that adding a land to a limited deck is like adding 1.5 lands to a constructed deck, which makes sense given the ratio of total cards. You can hedge this a little bit by running an extra card, which will drop your percentages by about 1 for land drops 2-3 and 2-4 for 4-6. This is useful if your deck has a lot of redundancy and you want to avoid mana flooding.

People who know me will tell you that I am far more prone to run 61 or 41 cards than most, and this is the reason. I believe that having the proper mana ratio is very important, and running an extra card is a great way to hedge against drawing too many lands. Picking up that extra four percent of not drawing your sixth land is something to consider when you are building your deck.

So now that you know the number of lands you need, it is time to determine the color balance. If you are mono-colored, it's trivial, but if you are not, this is a simple process to help you determine the proper balance. For most two-color decks this is very easy and will likely be intuitive, however, for more complicated manabases it is an invaluable approach.

1. Count the number of mana symbols of each color in your deck.
2. Count the number of cards of each color in your deck.
3. Create a ratio of both mana symbols and color; see how closely they match up.

If your mana symbol ratio and card ratio are close, then it's easy, just pick one and go with it. If they are far apart, then you have a few very color intensive cards that are probably driving the card ratio up. You have to determine how important it is to cast those cards on time. If it is very important, you should skew your manabase towards that color, treating it as if there were more cards in it. The exact specifics of how much to skew will come with experience.

When you have your mana ratio you then know what percentage of your manabase must produce each color. For example, if your mana ratio is 5 blue vs. 3 white, then 5/8 or 62.5% of your mana sources should be blue. By contrast 3/8 or 37.5% of your mana sources should be white. These are, of course, minimum numbers.

Why do I say "mana sources" instead of lands? Because of the way I treat dual/multi lands. What you need is a percentage of your TOTAL mana production.

Here are some guidelines:

• Dual/multi lands count as 1 source for every color they produce, so Savage Lands counts as a single green, red, and black source.
• Fetchlands (Misty Rainforest, Terramorphic Expanse) count for ¾ of a source if they are fixing for 2 colors, 2/3 if they are 3, ½ if they are 4, and 1/3 if they are 5.
• Lands that have special conditions need special rules. For example, Tainted Field is probably only good in a deck with at least 12 Swamps, and likely counts as both a white and a black source in those decks.
• Colorless lands count for a single mana source, always.

After you have a total mana source count (which will almost certainly be higher than your total number of lands due to multicolored lands), you should make sure that your color ratio matches your mana source ratio as closely as possible. Your mana ratio can be skewed by cheap, color intensive spells as follows:

Note: C refers to a generic colored mana.

In order to cast C 70% of the time on turn 1 you need 9 lands of that color that come into play untapped.

In order to cast C 85% on turn 1 you need 14 lands of that color that come into play untapped.

In order to cast CC 70% of the time on turn 2 you need 17 lands of that color.

In order to cast CC 85% of the time on turn 2 you need 21 lands of that color.

In order to cast CC 70% of the time on turn 3 you need 15 lands of that color.

In order to cast CC 85% of the time on turn 3 you need 19 lands of that color.

In order to cast CCC 70% of the time on turn 3 you need 22 lands of that color.

In order to cast CCC 85% of the time on turn 3 you need 26 lands of that color.

In order to cast CC 70% of the time on turn 4 you need 13 lands of that color.

In order to cast CC 85% of the time on turn 4 you need 17 lands of that color.

In order to cast CCC 70% of the time on turn 4 you need 20 lands of that color.

In order to cast CCC 85% of the time on turn 4 you need 24 lands of that color.

The reason this is number of lands and not sources is because you need to draw different individual lands that produce those colors. This should show you why the Shadowmoor filter lands are so strong, since they count double for casting hard to cast spells.

The last thing to talk about is whether or not lands come into play tapped. Here are some more numbers:

With 8 ETBT lands you have a 65% chance of drawing one in your opening hand and an 82% chance of drawing one by turn 4. With 10 lands those numbers rise to 74% and 89%. With 12 lands they go all the way up to 81% and 93%. This obviously affects your mana curve. As you can see, if you are running 8 or more ETBT lands, the vast majority of the time one of your first 4 land drops will be a tapped land. You have to take this into account when planning your curve.

That's basically everything, but as an exercise in building manabases for decks with complicated manabases, this is the deck that I took States almost a year ago:

[cardlist]

[Spells]

3 Cruel Ultimatum

4 Path to Exile

4 Lightning Bolt

3 Day of Judgment

2 Negate

4 Double Negative

4 Esper Charm

3 Ajani Vengeant

1 Into the Roil

2 Mind Spring

[/Spells]

[Creatures]

2 Baneslayer Angel

1 Sphinx of Jwar Isle

[/Creatures]

[Sideboard]

2 Negate

2 Mindbreak Trap

1 Bogardan Hellkite

3 Wall of Denial

3 Flashfreeze

4 Tidehollow Sculler

[/Sideboard]

[/cardlist]

Using lands from Shards block, M10, and Zendikar (the set), what manabase would you build for this deck?

Critical stats (maindeck)

Mana symbols

• 21
• 27(28)
• 13
• 17

Cards

• 16 White
• 17 Blue
• 7 Black
• 14 Red

The card in the sideboard that affects the manabase the most is Tidehollow Sculler, since it is far better when cast early. Why don't you try building the manabase and see what you come up with? As a spoiler, I will tell you that in an 8-round event my manabase only hurt me in one match, and that I nearly beat Spread'Em with this deck. I did make last-minute changes to the deck at the event site, but the manabase (which I did not change) was built for that list.

Chingsung Chang

Conelead most everywhere and on MTGO

Khan32k5@gmail.com