Limited Design: 3/2 For Three is the Place to Be

If you drafted Guilds of Ravnica one hundred times, you probably would be repeating almost identical deck patterns after the first twenty or so. Guild and tribal formats just tend to work that way, where you have a very set number of viable strategies and each of those has fairly scripted synergy plans you are building for.

If you played one hundred matches of Guilds of Ravnica draft, you probably wouldn't be bored. The game play was challenging, flexible, unique, all the best parts of limited Magic.

While the strategies and spell mix of Guilds of Ravnica certainly played a part in this, limited is at its core a format of creatures. Guilds of Ravnica excelled here, allowing a wide range of creatures to see play and be relevant across entire games.

All the set had to do to start this was place the basic replacement level creature at three mana 3/2.

But why three mana 3/2? Why does that base sizing create this natural dynamic game play that a different size doesn't? The best way to explain is backwards, looking at why other sizes just don't work the same way.

Good Limited Gameplay

Before we start, it's probably best to define what I'm talking about with "dynamic game play". The two things that should be maximized are:

-In game decisions are constantly unique to the exact context of the game, matchup, and cards.

-More individual cards are playable, allowing for a wider mix of effects to react to matchups or plays.
-Less cards/creatures are obsoleted due to being made not worth a card of battlefield equity, or forced down a text box road of being "artifacts/enchantments you can Doom Blade".

This is what our base creature size should be trying to maximize.

2/2

For a very long time, 2/2 was a much more important size in Limited. When people talk about creatures just being better now, this is part of why.

Dominaria ended up being a format heavy on 2/2s, likely because it tried to capture this classic Magic feeling. While many people consider Dominaria an amazing format, it doesn't take much to look at it and see why having a 2/2 base size doesn't lead to dynamic game play.

Simply put, 2/2s suck. It's extremely hard to design other creatures at any cost that don't turn your 2/2 into blank cardboard. If you do this by making the three and four drop creatures into 2/2s, then the size above them just does the same thing because no one ever dies to a stupid 2/2. If you try to upsize the mid-curve creatures but still let them trade with a 2/2, you are right back to a 3/2 format.

When your format is centered around 3/2s, there's a lot more room to design creatures across the curve that don't immediately get obsoleted. Two mana 2/2s and four mana 3/3s both can play a role in the same combats as your three mana 3/2s, and from there a variety of uneven sizes like 1/3 or 1/4 can also do different things.

The fact that 4/4 is the size jump that bricks a 3/2 also has a lot of tempo implications.

Dominaria worked because used the 2/2 backdrop to ignore creatures and make wacky things more interesting, but that's hard to repeat for a number of reasons. Dominaria got away with literally putting chapters on cards to promote this, but at the same time even professional players would just forget about lines of text on cards because of it.

Marginalizing so many cards in a format also has some not great effects. Either you create a format where the best three spells in everyone's deck are too good for the other twenty to deal with and most things are irrelevant, or you can create a format where the answers to those three cards are too good and there isn't a real way to win the game. Either way, you promote a format where no one attacks, you sit there drawing cards, and entire colors start getting written off because they have less answers and less card draw.

Selesnya was a near unplayable color combination for the first decade of Limited Magic because of this, and blue was always the best color. It was cool when Dominaria did it for the first time in a while, but it gets old quick.

3/3

If a 2/2 gets bricked, what about bigger things?

On the scale of things, a 3/3 format is actually pretty good. The set that immediately came to mind was Return to Ravnica, where the combination of the populate mechanic and 3/3 Centaur tokens made that the key size of the format. But 3/3s aren't as dynamic as 3/2s on an in game, tactical level.

Yet again, the 3/2 format promotes a wider variety of creatures being combat relevant.

The correct response to a 3/3 is just a creature with 4 toughness. The correct response to a 3/2 can be a 2/2, but a 2/2 in a 3/3 world is just not going to be a combat relevant creature until it double blocks. More of your creatures are judged on their non-combat utility, reducing the range of game text and cards that matters.

That all said it isn't horrible to make a bunch of 2/2s with reasonable non-combat abilities, or make two drops with flying or some other utility, so as Return to Ravnica demonstrated things can still work out.

A 3/3 format also shares an issue with a 2/3 format, so let's jump over to that.

2/3

Why not all 2/3s? This is why.

When every creature has more toughness than power, the natural game state is nothing happens. For Constructed players, you may remember this peak of board stalls in the Catacomb Sifter, Reflector Mage, Sylvan Advocate era, and draft is no different.

You control two 2/3s. They control two 2/3s. The correct play is basically to never attack. The best case scenario is having a combat trick that trades... for a 2/3. Then they can play a replacement 2/3 and you are back to the start.

When each player controls two 3/2s, attacking is a decision of resources. Do you want to make an exchange? Does your 3/2 have an ability that makes you want to do that like Spinal Centipede? Does your opponent have cards that are advantageous if they control more creatures? If they don't block, are you winning the race?

You control a 4/4. They control two 2/3s. Welp, guess you can't attack without that trick. You control a 4/4, they control two 3/2s. This is a good attack. A smaller increment in creature sizing matters more, allowing for more diverse design.

As I mentioned, 3/3s can share a bit of this. To attack profitably into multiple 3/3s, you need a 6/6 creature to not get double blocked. The games can easily bog down until someone just has a larger raw number of creatures than the other. Remembering Return to Ravnica, this could often happen where games became about counting 3/3 pennies and shreds of paper rather than trying to figure out leverage on attacks through the entire game.

Another hidden feature of the 3/2 baseline also shines through in double blocks: the 3/4 reward. Let's flash back to Eldritch Moon. I consider that format to be easily one of the best of all time (just below Innistrad), and could easily write a full length thesis about everything that can and should be replicated to make other formats as good as it. For now, let's consider this the first chapter in that paper and just stick to the 3/2 sizing that the Eldrazi Horror tokens set.

Faithbearer Paladin is an unassuming card. It's a moderately sized creature, about as big as white gets, and it has a single evergreen ability that isn't evasion. However, because it is a 3/4 in a 3/2 format it has huge ripple effects.

First off, the 3/4 sizing in a 3/2 format is the minimum sizing required to stall anything in the 3/2 scrum. This includes the 3/3s that are largely still just brawling at parity against 3/2s. This minimum jump allows it to interact differently against the stock +2/+2 combat trick sizing than a 4/4 on defense, and means that double blocking when the 3/4 attacks is profitable. You are allowing some of the board stall creating things to exist, but pushing the threshold to break the stall down and pushing the cost to create it up from the 2/3 metagame.

The double block part is where the incentive comes in. See, Faithbearer Paladin is very much a card you want to attack with, but the sizing makes you need to try to figure out ways to make that work.
A deck leaning on Faithbearer Paladin would put more stock into combat tricks, auras, or random Human boosts than you would expect for a deck also playing a five drop to stabilize. Normally you would think of wanting these payoffs as a property of smaller creatures like Scroll Thief, but just by the numbers you have more ability to create this design at all sizes.

X/1

If less toughness than power allows for more dynamic sizing to matter, why not drop all the way down to one toughness?

In a 3/1 oriented format, you can't print anything that leaves a lingering token, or damage, or anything like that. You can't create fractional value, because that fraction always rounds up to a full card. 3/2 is also a nice point because it's reasonable to accumulate these fractional advantages and cash them in for a full card. You can even print 3/1s in the format, knowing that even if they trade for a 3/2 the reduced toughness is an actual drawback because fractional edges exist.

The reverse of this is also true as you move up to a 3/3 or 2/3 format. The larger the format toughness baseline, the more stray tokens feel like Chuck-E-Cheese where 100 of them that you spent all day winning trades for a king size candy bar. 

Three Mana

I've spent a lot of time stating why 3/2 is the right size, but I also specified three mana.

If you make your format about two mana 3/2s, bad things happen. Like Gatecrash. While Gatecrash isn't the worst draft set of the modern era (that "honor" goes to Avacyn Restored), it is very close.

You can only really get hit with a 3/2 a handful of times before you die. This is also why 4/2 sizing isn't great, not just because it trades too far up but because the return on tempo plays and Falters just becomes too high. Think about how easy it was to kill someone with Piston-Fist Cyclops and Maximize Altitude in Guilds of Ravnica, and we have seen similar things repeat over and over with cards as bad as the 5/1 Incurable Ogre. The puzzle should be "how do I want to interact in combat", not "how do I avoid the declare blockers phase".

3/2s don't have that issue, but at two mana they have the tempo issue. This occurs even if all other sizes shift with the 3/2s, as the turning point is the turn you get to make multiple battlefield relevant plays. If you produce multiple attackers on turn four and can follow that up with multiple plays, your opponent just dies. At three mana, the turning point for multiple great plays is often turn five, and specifically five lands. More cardboard has to be invested on mana and not combat, meaning there's less chance of running turns of multiple plays that automatically bury someone.

Four mana 3/2s have the inverse issue, where it's too hard to make multiple plays. If you miss a turn of making a play, you are just too far behind because there is no place to make it back. Also, if you invested too much into 3/2s and get bricked by a 4/4 there's no good way to make a tempo positive play and increase your advantage. 

The more expensive your small creature baseline is, the better larger creatures are at stopping it because it costs about the same one turn of mana to produce something regardless of size. The less expensive, the worse trying to block with something bigger becomes because it is too easy to get punished over and over for spending a full turn on a play instead of a partial turn.

Another subtle issue with two mana 3/2s is that it is hard to create a lot of abilities that are reasonably costed for the body. One of the payoffs of a wider number of cards being played is the diversity of effects makes games unique, and if all your 3/2s have basically no game text that's lame. This is why Core Set 2019 fell flat, as even though 3/2 was a key sizing they all felt nearly vanilla. Woah this Hired Blade has FLASH, so you see you pay three mana and then... it's just a stupid 3/2 with no abilities on the battlefield. With Whisper Agent at least flash and surveil was a baked in synergy to the entire Dimir plan of surveil payoffs and other instants to cast, and basically every other 3/2 had some cool ability.

The one problem with three mana 3/2s is an echo of Constructed. It is hard to print a really good cheap removal spell like Shock for Constructed purposes and not have it absolutely dominate games of Limited because you trade one mana for three mana. Dead Weight is absurdly good in Guilds of Ravnica draft, and I don't know if you could print something like Shivan Fire at common there. There are ways around this via higher rarities, but it's definitely something that needs balancing thought on both sides of the coin.

What Else Works?

Listen, if every format was a 3/2 format eventually it would get a bit stale. You can print a ton of different effects, but sometimes it's nice to have a Dominaria where the base principles feel completely different for a breath of fresh air. Or, like I said, so you can print a good cheap removal spell and not cause draft nightmares. You can really construct a format with another creature size and its limitations in mind, but there's another trick.

The other viable option is to make sizing largely dynamic. You can't have static, boring games if creature sizing is extremely variable turn to turn on the same piece of cardboard. This is often tied to a mechanic: morph, landfall, exert, or so on. I guess also transform, but I've ranted enough about Innistrad and Eldritch Moon for one day.

You may notice that this covers some of both the best and worst formats. Khans of Tarkir was a huge success with morph, and Hour of Devastation utilized exert to great effect. On the other hand, Amonkhet's exert was very non-interactive, landfall in Zendikar was largely the same, and recently in Ixalan most of the tribal payoffs that made the format so lopsided were tied up in dynamically resizing creatures of that tribe. What's the difference between the two?

One of the easiest ways to make sizing dynamic is via a temporary boost. The key to balancing this is making this boost exist without making blocking impossible. This was the Zendikar and Amonkhet issue, where the creatures blocked as 1/1s but attacked as 3/3s making the format a race. The sizing also got a little too big for too low cost, pushing things into the tempo and falters being too important category we talked about with two mana 3/2s.

Hour of Devastation changed exert to let blocking be harder, but not quite as impossible with random ground creatures. Rhonas's Stalwart and Khenra Scrapper gained less toughness and more manageable evasion than Gustwalker's flying and Hooded Brawler's size. Your decisions along the game's curve became more about figuring out how to present or push into a large range of potential, difficult, but viable blocks.

Ixalan took a more permanent approach to sizing, investing multiple cards of resources for a large payoff. The problem here is balancing the reward of investment, and the cost in case your opponent has a direct answer. It was a little too hard to answer a boosted thing in Ixalan, allowing it to just run away with the game. The flip side is the previous 20 years of Magic, where it was just too risky to attempt this and end up down cards. Rivals of Ixalan worked on this by providing smaller boosts, better removal, but also returning some value immediately from the enhancement. Squire's Devotion was more vulnerable in the format than Mark of the Vampire in Ixalan, but between the immediate lifelink and token you could balance the risk-reward equation and make the case where the creature died pallatable.

Morph is an odd one. It reliably works for these purposes, making a 2/2 format early where the 2/2s are live late. But it has a lot of rules baggage associated with it, especially in tournament settings (obviously I would point that out). 

Morph can also dominate a few key interaction points of the format, so much that some basic effects get overly magnified or ignored. It makes sense this would happen because morph bottlenecks a portion of game play through the three mana 2/2. 

Take the meek and miserable Grizzly Bear for example. Sometimes in morph formats it is unbelievable because it is a cheaper 2/2 in a world of 2/2s that largely trade (Onslaught). Sometimes, it is unplayable because every morph is secretly a six drop, and every deck just wants to play morphs and aim to draw more six drops later on (Khans of Tarkir). This doesn't change per game, it's just locked in on a format level. The same applies to combat tricks, which either are great at killing three mana creatures or irrelevant because morphs are just a temporary phase of the game. Or to morphs turning face up too cheap, which is why Khans of Tarkir had strict design rules about unmorph costs. This design also has the cheap removal issue, where Shock is the best thing ever in a morph format. Morph is a fine tool to use, but it also has its own issues. 

Hopefully this post has been useful for thinking about Limited on a system-wide level. A simple decision about typical sizing has massively branching implications for every card in the format. Next time you look at a new format, build a new cube, or I guess if you are really lucky sit down at your Wizards of the Coast office in Washington, put these ideas to good use.