Wednesday, August 22, 2018

As Garfield Intended

I’m a big fan of variant formats. Not only do they add some spice and variety for long time players, but they also allow for more flexibility when scheduling Magic events with your friends (I’m always looking for more limited formats that don’t require 8 people in the same place for multiple hours). When I saw the most recent episode of Enter the Battlefield about Richard Garfield and the Alpha Playtesters, I became inspired. The “original” Magic format they described, full of surprises and built around trading, sounded like a lot of fun! It got me thinking: how can we best recreate that original Magic environment using modern technology and design sensibilities? In this article I’m going to explore the best way to play Magic “as Garfield intended”, literally.

Let’s start by going over the goals of this theoretical format, which I’m going to call “Garfield League”. We’re trying to emulate Magic as it was played by Richard Garfield and the original alpha playtesters, while making sure to prioritize fun over nostalgia and novelty. Based on the video, and previously available common knowledge, Garfield League at its core is what we today call a Sealed League: players use pools of cards to construct 40 card decks, and play each other in a (mostly) non-structured fashion over a long period of time. However, Garfield League introduces a number of unique factors:

  • Card pools are small (60 cards total, instead of today's standard of 90)
  • Basic lands are in limited supply, and included in the card pool
  • Players are completely unaware of what cards exist (or could exist) in other player's pools
  • Trading between players is allowed and encouraged, including/especially trading basic lands
  • Ante

Type P

Before we get started, I’d like to briefly mention Type P, a fan made format (called a “perpetual sealed league”) that features some similarities to Garfield League. You can find the full rules of the format here. Personally, I think Type P is a little too old school (originating in 1997). It features a lot of slightly convoluted rules that I wouldn’t expect players to remember without a reference sheet (which is never a good sign if you’re trying to convince people to try something new and weird). Still, it’s sort of a precursor to what I’m covering here, so I thought it deserved a shout out.

Card Pools

Using a smaller pool of cards (with trading) is perhaps the thing that most sets Garfield League apart from regular old sealed league. A small card pool that includes a limited supply of basic lands means your deck building options start out very restricted (and honestly quite poor), and will only improve through active trading with other players. Watching players’ decks transition from piles of cards to well-tuned machines over the course of the league seems to me like one of the big draws to the format. That being said, we have to be careful to not overdo it. A pool that’s too small gives players no options to start with, and can lead to an extended period of unfun, underpowered games.

In the Alpha Playtesters feature, Barry Reich describes the original playtest pools as 60 cards (divided evenly between colors), which included basic lands (also evenly divided), with which they’d build a 40 card deck to play with. If we want to copy this 60 card pool exactly, assuming we use modern booster packs, we’ve only really got two options: 2 boosters with 30 basics, or 3 boosters with 15 basics. The former option feels a little too anemic to me: you’ve only got 28 non-land cards to build with, and just one bad rare can be a huge hit to your future prospects. The latter offers you just barely enough basic land for the 17/18 land standard, which all but guarantees your mana base will be unplayably horrible until you trade with other players who aren’t also using the specific basics you’re looking for.

Once we break away from the hard 60 card limit, our options get a lot better. My personal pick would be 4 packs, 30 lands (90 cards total). Having played a few "small sealed" casual events myself, 4 packs usually results in "decent enough" three color decks. 30 basics means 6 lands of each type, so your "best" mana base is going to be 6-6-6 anyway. Wonky three color piles should be expected in round one. However, it would only take one instance of trading unused lands and spells with another player to turn your deck into an average "two-color with a splash" sealed deck, and from that point forward trading can be focused on hunting down specific cards to improve your deck's weaknesses or match-ups.


Long time Magic players often mention “the good old days” before spoilers, when you basically had no idea what to expect when sitting down across from your opponent. Personally, I enjoy going into games prepared, but I can certainly appreciate the novelty of experiencing true surprise, and it’s definitely a key part of what makes Garfield League special.

There are a few ways to try and simulate the experience of "playing blind”. For some people, simply playing a normal limited environment while avoiding spoilers is enough. If everyone can actually make it to release day spoiler free (I know I certainly can’t), then you’ll be treated to a blind experience that also benefits from all the design work put into balancing each set’s limited environment.

That’s the ideal, but I know there are many of us out there that can’t miss a single day of spoiler season, let alone skip it all together. For us, there are two options I see. One is the “chaos sealed” option, where everyone receives random booster packs from a variety of different sets. For players without encyclopedic knowledge of Magic sets, this can do a pretty good job of obscuring the contents of each player's pool. If you really want to mix it up, you could even pre-open all the packs, sort them by rarity, and then redistribute them. The chaos method loses out on a well-balanced and well-tested limited environment, but the diversity of themes and mechanics allows for much more discovery (both within your own pool, and through trading with opponents).

Finally, there’s the cube method: you create a curated list of cards completely unseen by the other players, and create packs/pools out of that. This offers you an extreme level of control over the environment, and what’s more: the maximum level of surprise for your players as they have no idea what to expect. The downside is that you will know every single card in the league. However, that’s what Richard Garfield himself did, so maybe you’d enjoy that sort of thing (note that the "chaos draft resuffle" method also has this issue). If you really wanted to avoid having one person with full knowledge, you could even go “potluck style” and have everyone bring in a stack of their own cards to shuffle together at the start of the league.

My choice would be the chaos sealed method. As a long time cuber, I've struggled for years at finding the right "prep work to fun output" ratio. Simply acquiring a variety of boosters is a lot easier to manage than the other methods, while still providing a sufficient level of surprise for my taste (especially after one or two rounds of trading have happened). Something worth mentioning is that unlike chaos draft (which in my experience often flattens out into being about generically powerful cards with the occasional cool theme deck), chaos sealed usually gifts each players with two or three synergistic combos by virtue of them receiving an entire pack from the same set (in draft that pack is split among all the drafters). Because of this, chaos sealed decks often start out just a tad more synergistic than chaos draft decks, and this can really help drive players towards seeking out synergistic cards to trade for as the league goes on.


Trading offers an interesting and nostalgic method of “card pool building” not seen in most limited formats, and it can be fun to watch the “card economy” evolve over the course of the league as players learn how to properly value their cards in a metagame with lots of hidden information. As mentioned above, Garfield League card pools start small and are purposefully hard to build coherent decks out of, so trading is an essential part of the experience (including trading for the right basic lands!)

However, trading can create some complications. The easiest way to “regulate” trade is to have no restrictions at all: players can trade any number of cards for any other number so long as they agree. This is nice and simple, but you could run into situations where players have drastically different pool sizes due to lopsided trades. This can create problems with players keeping track of how many cards they should have, or certain players ending up with too few cards to play or trade effectively. Type P has some rules to address this, but I find them a bit too fiddly for my tastes. And who knows, maybe your group trusts in the power of the free market (or thinks making poor trading decisions should have consequences).

Limiting trades to 1-for-1 trading system helps to regulate pool sizes, but it becomes hard for players with weaker pools to “trade up” for stronger cards with nothing to offer. You can always try to work things out under this system (trading four mediocre cards for one great card plus three useless cards), but it might end up being too restrictive.

Since I don’t have a PHD in mathematics, I can’t really say what method of trading would result in the most fun, but my pick would to start with unrestricted trading and see if any problems arise. Since a Garfield League would probably have a normal sealed league life cycle of a few weeks or months, the chances anyone will find themselves in serious long term trouble due to bad trades is pretty low. Unrestricted trading can certainly be predatory to newer or less experienced players, but I would say Garfield League is not a good format for beginners in the first place, especially if you're playing with strangers or a friend group that's particularly cutthroat


Finally, there’s the elephant in the room: ante. Ante wasn’t even mentioned in the ETB feature, but most players are aware that it existed in Magic’s earliest form. Ante has a lot of problems that I won’t get into here, and was removed from the game with good reason. However, ante was originally formulated as a way to facilitate cards transferring from player to player, encouraging trading. Personally I think you can easily play Garfield League without ante, as the format naturally has a lot of incentives to trade cards around. Type P also offers alternative ante rules, though again they’re a bit prickly.

Rather than ante, I think I’d probably try out a system where trading can only happen after a match is completed. This way, players are encouraged to play games with other players in order to acquire new cards, but aren’t forced to lose cards for losing matches. This method would act a bit similar to the casual limited format "Continuous Draft", where players leave each match with a different pool than they came in with. However, it also addresses the biggest issues I've had with continuous draft: you're not forced to (potentially) give up your favorite cards each time you sit down, but instead get to build your strategy over time.

As an extra bonus/incentive, I would allow the winner of the match to peruse the loser’s deck and sideboard (offering trades for the cards they find appealing), but let them keep their own pool a secret. This doesn’t give the winner any tangible advantage during the trading process, but allows them to keep more information hidden from the rest of the players. I've ran a copy of Deal Broker in my cube since it was printed, and can attest to the importance of information when it comes to trades. Offering up a premium removal spell for trade sounds a lot more inciting when the other player doesn’t know about the second, much better card hiding in your sideboard.

To recap, here's what I'd recommend for running your own Garfield League:

  • Each player receives 4 booster packs from random sets, and 30 basic lands (6 of each type)
  • Players build 40 card decks (using only the cards in their pool)
  • Players schedule and play matches with one another on a casual basis
  • After the completion of a match, the winner can look through the loser's deck and sideboard, and offer any trades they wish to make
  • Trades can be comprised of any number of cards (including basic lands)

Garfield League is clearly still in the “early idea” stage, but hopefully I’ve given you some ideas about where it could go. It combines a lot of interesting ideas I've tried separately, but seeing all of these variants interact with one another can create something truly special. Finding new ways to play Magic is one of my favorite parts of the game, and I’m eagerly awaiting the chance to try out this nostalgic twist on a familiar format.


  1. Good stuff. Thanks, Chris!

    I messed with a similar trading-based Limited format some years ago: Swap. In that one-and-done environment, limiting/focusing trading was crucial, but it's easy to imagine how open trading could work fine in a long-term even like your league.

  2. Are you suggesting that the loser should never be able to propose trades? That seems like a good way to ensure bad pools stay constantly bad.

    For example: suppose my manabase sucks, but my opponent lucked out with a good manabase (+ some fixing). I want to trade for basic lands in specific color X, and I'm even willing to give up spells from color Y to do so. How is my opponent going to come up with the idea to make this trade while combing through my pool, other than treating my pool as a separate sealed deck building exercise and divining which color I'd benefit from playing more of and which color I wouldn't miss playing less of?

    Also, would players play multiple games against the same opponent within the same league?

    1. There's certainly the possibility that a player could end up with a pool that has literally nothing of value to other players. I suppose this is one situation where ante is useful, but I think the small starting pool size means you'll almost never start out in that state (as in, even the players with the best pools will be seeking out good filler in their colors).

      The specific situation you describe is easily avoided because the players are having a conversation, not purely a transaction. If I lose, and my opponent is looking over my stuff, I can easily say "hey, you see what I've got, and I'm looking for [this type of card], got any offers?"

      In my experience I've always played the same opponent multiple times in the same league, but that's because I've only played sealed leagues with somewhat small player counts. You could avoid rematches if you've got enough players, but I think Garfield League would lend itself better to rematches than most other formats, since your deck could be significantly different from the last time you faced off.

    2. That makes sense, though in that case I would hope you would not be cultivating too cutthroat of an attitude in the league, i.e. "If I trade you a card you want for your deck, you might beat me, so I'll just decline to make sure your deck continues to suck". It seems like trading is supposed to be an important part of the "format", with the goal that everyone should be able to gradually improve their decks, and hopefully have more fun games toward the end than at the beginning.

      I'm also going to assume that the pack distribution works like GP Chaos Sealed, where everyone is guaranteed at least one pack from a "good" set (Modern Masters) and one from a "bad" set (Battle for Zendikar, Theros).

      Regarding spoilers: if this is a format for experienced players, it makes sense that everyone should have a rough idea of what the entire (potential) card pool does. Having to repeatedly pause a match to read cards is not my idea of fun, especially if you're playing with strangers who are not inclined to be charitable or patient.