Received Wisdom: Knowing Your Mana Counts
by Roberto Castro-Mahoney

Most people learn to play Magic: the Gathering from already existing players, receiving from these seasoned veterans some rather standard advice on different aspects of the game. On the whole, this is usually a good thing. And assuming these mentors have some semblance of play skill, they can be an invaluable resource for new players on tricky subjects like mana curve, land ratios, splashes, and cards to play around.
However, many players get these different pieces of information in their head without ever truly understanding the “why?” For instance, do you understand why you play 17 land sources in your Limited decks? Do you know why you split your mana 10-7 or 11-6? Are you making an informed decision or just going on what you were taught as law back when you first started playing? For instance, you know to run seventeen land, but how do you balance them? If you had the above cards in your draft pool, how many Plains would you want to run? How many Swamps?
What I’m going to attempt to do here is explain a bit about the logic behind these generally accepted practices. While this will still be “received knowledge,” hopefully you’ll know where the numbers come from, and why these choices are generally accepted as being correct.
For starters, how many mana sources does a deck need? Or more importantly, how many does it want? These concepts may seem the same at first glance, though seasoned players will recognize an important distinction. This graph below is designed to show the likelihood that you will see a certain number of mana sources of a given type by the turn listed, based on total sources available in your deck.
*Note that if you’re on the draw, shift every figure to the left once.
I’ve highlighted the key points on this graph. Moving now to consider Scars of Mirrodin in particular: True Conviction is a card which most decks would want to hard cast as soon as possible, though to do so requires some heavy mana requirements – six mana in total, and at least three white mana. Now obviously running twenty lands is a bit much in the overkill department, as its doubtful your deck is top heavy enough that you’d need to hit six lands every single game. Instead, look at when you’ll want your triple white requirement. As we can see above, 10-11 white sources is about right if you’re looking to have WWW ready to go on turn six.
Now that covers a card which wants to be windmill slammed onto the table. But now think about something like Volition Reins. This card is much less important on the earlier turns, as you will usually want to save it, waiting for your opponent to drop a bomb worth stealing. With Reins, because we don’t necessarily need to play it on the turn it is first available, you can go as low as 9 sources of blue mana.
In Scars of Mirrodin draft, most players have been running 16 lands, since mana Myr (and the occasional Horizon Spellbomb) often raise the number of mana sources in a deck to 18 or even higher. In addition, this format has a surprising abundance of small creatures, so it’s often important to devote maindeck slots to as many little dorks as possible in preparation for the trading game. With so many decks sporting breakaway plays, having a few extra action cards goes a long way into winning. (Note that most other formats advocate a 23/17 spell/land ratio.)
A goal of most decks in this format should be to have an average of four lands on the table by turn four. This ensures that (even on bad draws) you have at least three lands. Moreover, the powerful cards in the format usually start at four mana (see: Oxidda Scrapmelter and Skinrender). Also, it is beneficial to make sure that you don’t often have to wait longer than turn eight to reach your six-drops. In the deck where you need to absolutely reach five mana you might run as many as 18 or 19 sources (again, counting Myr).
Now, lets look at double-costed colored cards and how many sources we need to hit them on curve. There are four-cost cards like Skinrender[8], Glimmerpoint Stag[8], the three-cost Kembas Skyguard[9]. Then of course are the six-cost bombs of this format, such as Sunblast Angel[7]. In brackets beside each card I have the [minimum] color mana-source count needed to reliably hit those cards on curve. A card like the Angel, I never want to wait for so I’ll try to get the eighth source in if I can – and you should evaluate your cards in a similar manner. If a card is sitting in your hand waiting for a mana source, are you upset or is the card not yet relevant to the board position anyhow? Which cards need to come out, and which can wait in the wings?
Lastly we can examine splashes. How many sources do you want in your deck to splash Oxidda Scrapmelter and Arc Trail? In a two-color deck you can round up heavily and play 11/6 since there’s no reason to shortcut. However, if red is your third color, you are going to want 3 sources for a two-card splash. Why three? First, note that with only 3 sources you are more-likely-than-not to have one in your opening hand (check out turn 1, sources 3)! If that isn’t enough, by turn seven when you’ve seen a third of your deck you have an average of .98 sources of the splash color.
What about other formats? If you think back to Shadowmoor, which had cards like Wilt-Leaf Cavaliers, these were three-drops that you desperately wanted on turn three, no matter what. 12 sources is right for hitting the curve on turn three, but think where 5 sources of the other color leaves you. I often went down to 11 sources for Cavaliers, etc. because it wasn’t worth gimping my second color down below 5 lands. While five was enough to usually have colored mana on turn three, it was not worth making six cards less likely to hit for the benefit of just one. You can be sure though, that my decks with three copies of Ashenmoor Gouger had their 12 sources. But there, reliably playing a four-power threat on turn three was central to my strategy. With a one-of Wilt-Leaf Cavalier, I would not of been so aggressive with my sources.
When you build your deck, look for the key cards that you will depend on for tempo. You should try to maximize your threats per turn while making sure you can play all of your cards. Keep it in mind too, while drafting, that trying to play Kembas Skyguard and Skinrender in the same deck strains your mana. Identify your important tempo plays and play the right mana split to make sure you reliably get there!

















Excellent article man !!! Excellent !!!
I put the chart on my phone and took it to draft. Great to have. Thanks.
Great post
Aura, do you have any questions?
I’m happy to see this was received well. With many of our readers returning from school for December and January I wonder if there’s more content like this you want? Feel free to suggest columns and visit the forums to post about whatever.
What are you stupid Roberto? His name links to a spam page. Stop responding to spammers.
Very useful, especially since Scars is splash heavy and has some very good double-cost cards and a few triple-cost bombs. It can also be applied to artifacts for metalcraft.
I’d like a little more insight into the math behind the chart, if anything. I’d like to apply it to ratios like 17 lands/41 cards, as well as 60 card decks for Standard.
Don’t play a 41st card unless you’re playing against mill and then you should add 42, 42, … 60, 61 etc.
The numbers are obviously rounded to the nearest tenth digit. It’s simple math that gets them. Each figure represents the average “pull” of lands based on turn and total sources. But yes I went up to 20 so that colored sources could be read as well as total lands. Also how many artifacts you can plan to have by turn 4 when you drop Chrome Steed, etc.
Obviously the option of actually getting them on the board depends a lot on their curve. Some decks don’t mind playing Rusted Relic on turn 3 and then activating on turn 4. Others might rather have it come into play as a 5/5.
@Chris what? Both of them were spam? Haha shitttt.
Great article, Berto. Good stuff to consider.
Your chart also shows the effectiveness of siding out a land when you plan to be on the draw (assuming that you’re not messing with your curve otherwise). The diagonally adjacent units are nearly identical.
Great article ! Would love to see more basic explanations on how to draft / deckbuild with a decent theory behind it ;-) Good work
I think this graph is misleading.
The probability that you will draw exactly 5 lands by turn 5 is 26.2% if 18 lands are played. The probability to draw exactly 5 lands by turn 5 is 27.2% if 17 lands are played. Is that not what this graph was meant to show?
Yeah I’m not sure if there is any value in that chart at all. I’d like to hear what value you guys think it has.
Berto, your odds of having drawn a red source in your opening hand if you are playing 3 is 37.4%. I don’t think you made this graph correctly at all.
That last post is wrong your odds are actually 44.8%. Whatever still less then 50.
ChrisT you are right about the probability of reaching that lands count. This is averages and the -6% you find comes from 6% chance of having multiple sources in hand. The graph is averages. Detailing a graph that shows precise likelihood of drawing your 4th land on turn 4 is much more detailed than a simple X,Y scatterplot that this one would presumably come out to.
You are correct that getting precise likelihoods is harder. Without precise likelihoods, however, the words likely and reliable lose their meaning.
You are wrong to use the word likely so loosely. In this article, you say that to be likely to cast Wilt-Leaf Cavaliers on turn 3 a deck would want 12 sources of white or green. You also mention, in the article, that to reliably cast sunblast angel on turn 6 requires 7 white mana sources.
Your odds to have 3 or more white or green sources by turn 3 is 55% if you play 12
and
Your odds to have 2 or more sources of white mana by turn 6 is 69% if you play 7.
However,
Your odds to have 3 or more white or green sources by turn 3 is 62% if you play 13
and
Your odds to have 2 or more sources of white mana by turn 6 is 59% if you play 6.
this is very good for you, ybg :)
Deckulator is a probability calculator you can use to answer questions about the probability of drawing card combinations. See the link for a graph of probabilities of mana per turn for different deck designs.