General Advice on Smoothing

[Ruien]

Imagine that I have a fairly balanced cube when draws go right, but sometimes there are non-games due to mana/color-screw and/or flooding. It is also the case that higher-curve decks face proportionately more variance in that they are more likely to be mana/color-screwed or flooded, although it's still a universal problem.

Next, suppose this hypothetical cube intentionally does not run much fixing -- let's say a couple of Terminal Moraine. There's a strong focus on two-color decks while 3-color is encouraged only as a very light splash (and even then would be atypical). Finally, double-color spells should still incur a risk of color-screw, so that 1MM != 2M.

Are there any good ways to help smooth this experience and significantly reduce non-games due to screw/flood while at the same time not making it significantly easier to run decks across a broader spectrum of colors? Put another way: can we reduce mana variance in a specific situation and do so without side-effects?

Cycling seems like a good place to start. Maybe a generous helping of monocolor cycling lands could solve flooding, but it seems a bunch of taplands would cause other issues, nor does it really help a two-lander opening hand. Maybe scry effects and scrylands? Maybe add "cycling M" to basic lands and add a free mulligan for every game?

Or maybe it's better to come at this question from the other direction: suppose I want to design a cube from scratch which binds tightly to two-color identities as a function of the mana base, while maintaining strong out-of-guild interactions which can only be "bought" at the price of adding risk to that mana base. Thus a two-color choice is "safe" -- ideally zero chance of non-games -- while three-color is perhaps more powerful but riskier, and four-color would be impossible to pull off reliably (both due to higher chance of color-screw).

Does anyone have any thoughts on how this might be accomplished? Even out-of-the-box solutions are welcome. This is mostly a brainstorming exercise and something I'm curious about.

 

[anotak]

idk your list but you could look into upping your count of cards with heavy hybrid mana investment. or twobrid mana.

 

[Onderzeeboot]

Lands like Evolving Wilds are a good starting place, I think. In addition, smoothing mechanics like cycling, scry, and morph can help.

 

[Jason Waddell]

I think in general you will always be reducing the risk of color screw with an increased density of fixing lands.

As for flooding / screwing (purely in terms of the number of lands), a little bit of card filtering can go a long way. Especially in non-blue, things like:

 

[Knifethrower]

Cards with more than one Mana cost, such as split cards, kicker, monstrosity, etc help with flooding.
Also cheap permanents with activated abilities, flashback cards, etc.

Off-color kickers or split cards allow splashing with low risk.

An across-the-board discard theme allows pitching lands, fatties, greedy splashes, that aren't working in the moment.

Most environments with close to zero fixing are safe for two colors and sometimes reward three color splashes.

I think you really just want the first sentences.

 

[japahn]

On the subject of how to push for 2-colored, smooth decks, my strategy has been running lots of 1CC and 2CC cards, very few multicolor cards, and plenty of dual lands.

More generally, I've found a more forgiving mulligan rule and a slow environment are a combination that greatly reduces non-games.

 

[Onderzeeboot]

On the subject of how to push for 2-colored, smooth decks, my strategy has been running lots of 1CC and 2CC cards, very few multicolor cards, and plenty of dual lands.

1CC cards as in Glorious Anthem-style costs? It's true that including a lot of double mana symbols in casting costs makes splashing harder. Murder needs a dedicated black deck, Dark Banishing can be splashed for.

Also, be aware that CC cards (like White Knight) are hard to cast, even in two color decks! It's almost never correct to include CC creatures, unless you support monocolored decks, have alternate ways to cast them (e.g. Bloodghast, or it's correct to not play them on turn two, I feel. In any case, the more 1C creatures you include over CC creatures, the smoother your cube's decks will run.

 

[Rowan_CB]

I've also had success promoting primarily 2-color (but sometimes 3-color) decks, primarily by eliminating lands and artifacts that provide immediate five-color fixing (e.g., no vivid lands, no prophetic prisms, etc.) Not running fetch lands helps as well.

 

[Knifethrower]

I have another undeveloped concept called the fast cube. The idea is to shave about 2 turns off the average game, but still have good balance between agro, control, and synergy driven decks. Part of this includes shifting and /or compressing the Mana curve by 1. Does having the whole cube be slightly cheaper help with Mana screw? I'm not sure, but I think so. Many of the cards one would choose for this cube probably have a stronger color commitment.

 

[Onderzeeboot]

Does having the whole cube be slightly cheaper help with Mana screw? I'm not sure, but I think so. Many of the cards one would choose for this cube probably have a stronger color commitment.

The opposite, actually. The cheaper the card, the earlier you need a specific color, the more lands you need of that color. In other words, the slower the format, the easier it is to splash for an extra color. I'll link to my favorite author on the subject, and let him do the explaining (And I get to link to an updated article as well, wooo!)

 

[Chris Taylor]

Karsten is the truth by the way, great advice in that column

 

[ravnic]

For mechanics, Morph and Cycling help, especially Landcycling
Having a lot of colorless playables, widely useable ones, is also a good step and further helps with an eventual artifact theme
I also like my smoothies with Banana and spinach

 

[sigh]

I don't have the time to find it, so you'll have to imagine a link to a counter article that says it's really not as intense as Karsten states, because he misses the variable that you have to have the card in your hand in order to need to cast it. So Karsten basically assumes you always have the card in your opening hand or similar, which is very conservative.

But good premise on his part.

 

[Onderzeeboot]

I don't have the time to find it, so you'll have to imagine a link to a counter article that says it's really not as intense as Karsten states, because he misses the variable that you have to have the card in your hand in order to need to cast it. So Karsten basically assumes you always have the card in your opening hand or similar, which is very conservative.

But good premise on his part.

Well, naturally he answers the question: "How many lands do I need to run to cast this card on curve." I don't think the chance to actually have that card in hand should play into his calculations. Of course, if you only splash a single black one drop in your deck, his calculations seem off, but in reality I think you shouldn't look at the 1 cmc column. You can't reasonably expect that card to turn up in your starting hand, so apparently casting it on turn one is not a priority. And he does acknowledge that premise, that he's calculating land distributions on the premise that you want to cast your card on curve with a certain reliability. Even if you splash for four Swords to Plowshares, you shouldn't treat it as a one drop for determining the number of lands you need to run. If you treat StP as a four-drop (for example), the number of required white sources in a 60 card deck drops from 14 to 10.

Note: You can actually flip the numbers around if you want to. If you want to splash a third color in your 40 card limited deck, six lands of that color will usually let you cast (single mana symbol) spells of that color on turn 5.

 

[sigh]

I absolutely disagree on "I don't think the chance to actually have that card in hand should play into his calculations." If it's not in your hand, you don't have to cast it. That's intrinsic to how a deck runs, and therefore intrinsic to the actual reliability of the manabase. He's overestimating the mana color needs by not accounting for that, no matter how much someone "acknowledges" it.

I'll link the article later that adds that variable. Providing the results for something like one black 1CMC card, but with actual math, not estimates. The same math can be applied to a 2 color deck with a 13-10 color split, etc. etc.

I've also talked about it before in some other thread about mana bases, so an industrious individual could probably find it.

 

[Rasmus Källqvist]

I absolutely disagree on "I don't think the chance to actually have that card in hand should play into his calculations." If it's not in your hand, you don't have to cast it. That's intrinsic to how a deck runs, and therefore intrinsic to the actual reliability of the manabase. He's overestimating the mana color needs by not accounting for that, no matter how much someone "acknowledges" it.

I'll link the article later that adds that variable. Providing the results for something like one black 1CMC card, but with actual math, not estimates. The same math can be applied to a 2 color deck with a 13-10 color split, etc. etc.

I've also talked about it before in some other thread about mana bases, so an industrious individual could probably find it.

I think I have to agree, it's not hard to argue that this is a case of conditional probability (having the card and being able to cast it).

 

[Onderzeeboot]

I feel like Karsten's numbers are more useful by ignoring the probability you're drawing the card. They're two different odds that are both important, but I feel your mana base should be based not on the probability that the card is in your hand, but by what turn you want to be able to cast the card if you end up with in your hand. Like, the question I feel we should ask ourselves isn't "how likely is it that I will have both the right mana to cast Wrath of God on turn four and actual Wrath of God in my hand?", it should be "what turn do I want to be reasonably sure to be able to cast Wrath of God in case I draw it?" The answer to that question can be guided by your likelihood of drawing Wrath of God, but it shouldn't be dictated by it. You could decide that you "always" want to have the right mans on turn four in case you draw the Wrath, or you could reason that you don't have enough space for that many white sources, so you're okay with casting it on turn six and figuring out the required mana distribution from there. In any case, if you do account for the probability of Wrath of God not being in your hand by turn four, and lowering the required number of white lands accordingly, my gut feeling tells me you will also increase the percentage of games in which you do have Wrath of God in hand but don't have the right mana on turn four.

 

[Knifethrower]

I agree that Karsten's number are conservative. For instance, I successfully run three color decks in limited with fewer sources then his article suggests, both since I have multiple cards in hand and can play one if I can't play the other, and also because I don't expect to draw all my cards on curve, especially splash cards.

I imagine these numbers are most useful in constructed, where being able to cast cards like wrath of God or splinter twin on curve reliably will make or break games. In some of these situations the fail state of not having the card in hand may not even be worth giving full weight in the mana base calculations, if you are going to lose a significant percentage of those games simply by not having the card.