Wednesday, August 27, 2014

Experimental Limited Format — Swap

I’ve had this idea for a new Limited format for years now. It’s partially based on the trade mechanism from Andromeda and partially on a vague idea about how players could build a Limited deck in a way analogous to how starting players build Constructed decks: by opening a few packs and trading cards.

No card designs follow.

Here’s the core idea: Each player opens 3 booster packs; players trade with each other; players build decks using the cards they kept and traded for; they play. That second step is the critical part that requires work and will make or break this format. It could be dirt simple and have no rules behind it: Anyone can trade anything with anyone else, until everyone’s done, or time runs out. The reason I’d like to avoid that is a lot of players hate binder trading and/or are terrible at it. Like me. (It’s also not a terribly clever display of game design, but that’s awful motivation.)

I would be remiss not to mention the possibility of using Pit trading, but that's just as chaotic as free trading with the added insanity of lying and betrayal, so I'm just going to set that idea aside. In this lock box. On the bottom of the sea floor. Of Venus.

I was attracted to Andromeda’s trading because it lets multiple players trade multiple cards in a structured way that’s pretty interesting on its own. The player whose turn it is sets out 2-4 cards, one at a time, and each other player sets out a card of his own whenever she does, of any color except the colors the active player has revealed. She chooses a pile of cards to keep and the choice of what to take next passes to the guy she chose from. Players are effectively bidding for the first player’s lot, or for another player's bid.

I tested that once years ago, and it didn't go great. It wasn't awful, it worked, but it wasn't awesome. I wish I knew where my notes were for that, but my recollection is that it was too slow and too rigid. Cards have wildly different values in Magic where they are abstractly identical in Andromeda, so forcing trading at 1:1 is limiting. Also, I tried breaking the trading up pack-by-pack so it would be more like a Draft, but trading some white cards away in pack 1 and 2 only to open a white bomb in pack 3 (while still valuable to trade) would be frustrating, so doing it all at once ala Sealed will give players a bigger picture of what they have and what they need.

Here's what I'm thinking now. The active player chooses any number of cards from her pool and reveals them. The other players reveal any number of cards from their pools. The active player may choose to abort the trade if she doesn't want any of those offers. The other players take their bids back in this case. (This is so that players can't purposely undersell the active player, or ignore her offer in favor of others.) Hopefully this doesn't happen, in which case she chooses any offer and takes that. The person whose offer she took then chooses any offer and takes that. And so on. Each player will be the active player the same number of times. How many? I'm really not sure yet. I'm guessing twice around. The number should be small to keep things moving and force players to make the most of each round of trades.

How many players? I'd like to test initially with 4 or maybe 6. It would be super if the format could scale well, but I'm guessing it'll skew heavily in favor of 8- and that would be fine. The world needs more Limited formats for 3-7 players. Likely, the number of rounds should scale with the number of players so that the total number of trade turns remains fairly constant, making the format consistent in length regardless of the crowd. I'll try 12 turns minimum. So, 3p=4 rounds, 4-5p=3, 6-8p=2.

What do I call this format? The most obvious name is Trade, except that's already got an important meaning in Magic, so I'm leaning toward another short, descriptive name like Swap or maybe Swap Meet. I'm tempted by a fun/flavorful name like Bazaar, but Swap fits the pattern set by Draft and Sealed much better.

I hope to try this out soon, and I'll let you know how it goes and what I learn. If you'd like to try it out, I'd love to hear what your group thinks of it.

Parameters, Technicalities and Variants
Skip the rest unless you're interested in really nitty gritty design stuff. This is mostly for my reference, but totally open for discussion.

Players open packs (less than Sealed) and trade cards with each other to make 40 card decks. This activity should take an hour or less. How to improve your initial pool in this way should be intuitive. There should be no ways to abuse the swap format that can't be easily judged. Collusion is an obvious concern, but I also don't want to see players getting locked out of trades either intentionally or incidentally.

Usually, a player with 36 cards from three boosters will want to trade away three colors to get 23-24 cards of a two colors. Ideally, mono-colored and poly-colored decks won't be terribly hard/unusual, but let's focus on two-color decks. 14-16 of your starting 36 will be in the colors you choose (more if trading isn't so easy that you can ignore your broadest colors in favor of your bomb colors, which I'd push to support a little more). That means you'll be looking trade for about 10 cards and able to offer about 20, which makes me optimistic that trading quantity for quality (and vice versa) will be a thing. Given 12 total turns to trade, you can trade 2:1 ten times, fail to trade twice, and be perfectly on target. That sounds like a lot of wiggle room. I'm more concerned about it being too much than too little, but we'll see.

10 total turns, maximum: 2p: 5 rounds, 3p:3, 4p:2, 5p:2, 6p:1, 7p:1, 8p:1. (total rounding: 12)
10 total turns, minimum:  2p: 5 rounds, 3p:4, 4p:3, 5p:2, 6p:2, 7p:2, 8p:2. (total rounding: 16)
12 total turns, maximum: 2p: 6 rounds, 3p:4, 4p:3, 5p:2, 6p:2, 7p:1, 8p:1. (total rounding: 11)
12 total turns, minimum:  2p: 6 rounds, 3p:4, 4p:3, 5p:3, 6p:2, 7p:2, 8p:2. (total rounding: 9)
15 total turns, maximum: 2p: 7 rounds, 3p:5, 4p:3, 5p:3, 6p:2, 7p:2, 8p:1. (total rounding: 15)
15 total turns, minimum:  2p: 8 rounds, 3p:5, 4p:4, 5p:3, 6p:3, 7p:3, 8p:2. (total rounding: 12)

My Draft group usually does a rare redraft, which is great because we don't pay extra for prizes, and players have no incentive to pick a card worth more on eBay than in their deck. When we do Sealed, we just keep what we open and forget prizes, because the redraft would take twice as long (and we often have more tied records because we usually have 4-6 Sealed players). We could go either way here since it's a small number of packs, but given the fact that we're trading cards, it seems more natural that players would just keep what the end up with. That will make it a Thoughtseize a pretty weak card to open for purposes of winning (because it's worth in a Limited deck is reduced, and you're unlikely to trade it away for something more useful), but I'm pretty okay with there being some balance between opening helpful cards and valuable ones. Note that this introduces another skill factor: knowledge of card prices. A lot of Limited players don't know Constructed card prices and that can be a real disadvantage when drafting for keeps.

You could play multiplayer, but I don't think any kind of team play would work, because team members might trade exclusively and/or abusively with each other.
We could restrict players from offering cards of the same color as any in the active player's bid. We could restrict the active player to offering cards of a single color. We could let the active player state what it is that they're most interested in. Each of these would add a form of complexity and limit player's flexibilty, in return for adding focus that might make trades more impactful. The second idea is the simplest and I'm guessing I might end up there, but I'm inclined to start with the third idea. If it works, it's super elegant; my concern is that it could slow proceedings down immensely for a lot of groups.

We could make non-active players reveal their offers one at a time, but that sounds slow as dirt. All choosing and revealing simultaneously is an option, but I'm calling it plan B. First I'd like to try asynchronous bidding. Each player can bid for the active player's lot at any time until everyone is done, even adding cards to their bid after seeing other's bids. I could potentially allow removing cards as well, but that sounds like a slippery slope toward unending bidding. We could time bidding and cap it a minute; we could let the active player count down from N; we could say when the second to last person locks in their bid, all bids are locked; or we could leave it open and just yell at slow players. I'll try that first, but I suspect some external end will be necessary. If 'table talk' is allowed before bidding, is it allowed during? That could slow things down or cause loudness, but it might be hard to prevent once any table talk is allowed.

If trading proves faster and easier than I predict, the format could potentially support a mere two packs per player. I doubt it, but you never know. If it's slower and harder, we could easily go to four packs and Swap would still fill in a useful gap between Draft's 3 and Sealed's 6. You could totally Cube Swap, though I imagine that could be a brain burner what with the much tighter power band.

What do you think?


  1. The way i read it, you're turning draft into a weird variant of team sealed.

  2. I always loved leagues where trading was allowed. This is like the same experience concentrated into one evening, so sounds very interesting to me.

  3. Got to Swap with four people last night. It went well, but not great. Promising, but in need of work. Like all game designs.

    The good news: Even with just four people and three packs each, the mechanic of trading cards to improve our decks worked. We all built good and interesting decks, comparable to the power level of Draft/Sealed decks. The act of trading was interesting and the process had rewarding moments.

    The bad: Swap asks bigger questions of swappers than Draft does of drafters. Choices in a Draft are atomized card-by-card. You're making larger strategic decisions as you go, but there's no real deadline to settle on a strategy. Sealed asks a single huge question, but gives you the entire time to figure it out, with no cost to exploring solutions and changing your mind. Swap can allow players to change their minds, and the sets of cards you offer don't have to be big. In fact, I did change my mind profitably (going from BG to GR), and there were some small card exchanges, but it is possible for neither of those to happen, and based on this one data point it seems like players will naturally steer toward fewer+larger trades.

    Why is that a concern? Because it ups the stakes on and complexity of each discrete game action (choosing a set of cards to offer in trade), making the format either harder for new players or longer.

    Two of my players were very new to drafting (fascinating to me that established drafters were less interested in trying something new and different). One of them was a little slow to act, and one of them was mind-numbingly, punch-me-in-the-face-to-keep-me-awake slow. Yes, I'm sure they'd be faster next time. Yes, I'm sure more experienced Limited players would be faster. No, I can't offload all the blame to them. Because of the number of permutations possible in trades, and because there it may appear there's no recovering from a suboptimal decision, there is more room for analysis paralysis than in Draft.

    The upside is that there is depth and skill involved. The experienced drafters not only swapped much faster, but saw multiple opportunities the others seemingly didn't. (Trading away 'invested' colors, or trading black cards for different black cards, to name two.)

    We played allowing full tabletalk, so that anyone could say anything. That worked well. It meant that individual trade turns took longer, but likely reduced the number of trades needed considerably. Next time, I will try swapping with no tabletalk. I expect that to take more trade turns, but for them to each be faster. Then I can judge which seems better in the long run. Our group only need about 2 turns each, 8 turns total, but I'll try 12 total again when we use no table talk because I think that will require more trade rounds.

    We played so that we kept all the cards in our pool. That made sense to me going in because of format's inspiration in card trading, but all the reasons and benefits of rare re-drafting from a Draft apply here too, and I'm eager to try it with redraft next time. It's pretty hinky when players trade in-game-value for out-of-game value.

    1. I'm also really eager to try it with 6 now, to see how different that is from 4. I expect it to make for better options in trades on average, as well as making cross-trades (trades between non-active players after she has taken her first choice) more common. It shouldn't take any longer if we don't allow tabletalk (which sounds more appealing the more players we add—not many enjoy shouting over one another).

  4. Also, it's fascinating to me how much Magic players fear the new. For a group of people who don't just tolerate, but crave and demand new cards 4+ times per year, the vast majority of Magic players are so convinced the way they play is the only good way that they will not even give passing thought to multiplayer sets like Conspiracy, new formats like Swap, or other games at all. Obviously (and thankfully) there are tons of exceptions, but as someone who's interfaced with every type of gamer, the only group I've seen as insular and close-minded are old-school war gamers.

    That's not a value judgment, as many of the folks I'm thinking of are really nice people, skilled players and fun opponents. They just don't have the time to find better things, and that makes me sad for them. Because local maxima.

  5. Oh, I love Pit. Let me go get that lock box:

    Each player opens a Cube pack of 10 cards. Players swap back and forth for 5 minutes. Trades can be made face up or facedown. Repeat this process 3 times. Then players make decks.

    I think adding the time limit would fix a lot of the issues involved, as well as creating the fun trading floor aspect of pit. I'll have to try it out.

  6. Swapped again last night. This time, with 8 players. The format still needs work, but as an experienced drafter who's done this twice now, I can see a horizon of things I haven't yet mastered, which bodes really well for the depth of the game. For example, I thought I was red-green until the very end, but once I laid out my card pool, I realized I was so mono-red-aggro that I wouldn't even splash for my Soul of Zendikar. If I'd realized that sooner, I surely could have traded it for several good red cards to make that deck even better.

    It took longer than I want it to when we're done, but it took less time with 8 than it did with the particular group of 4 we had last time, despite their being a lot more potential trades. I have confirmed my suspicions that Swap with its current trading method isn't ideal for 8. While I could just say it's a format for fewer players, I think there's a better solution that will improve the game for all sizes.

    The Andromeda trade, while interesting and clever, just isn't suited to large groups, or to trade pools with such vast differences in value. It's also not intuitive. I'm going to look for another trade mechanism, and I'm going to focus it so that you only ever trade directly with the 2 or 3 players closest to you. That will limit neck-craning, and talking past people, as well as allowing for more simultaneous action. That should speed things up considerably.

    I vastly preferred rare redrafting. I opened two planeswalkers in my three boosters, and was able to trade Jace and a handful of solid blue cards for a great set of green and red cards, while keeping Chandra to myself.

    Not allowing table talk had pros and cons. With 8 people, there would have been way too much negotiation and it would have gotten loud if all table talk were allowed. On the other hand, while it's possible to nonverbally communicate what colors you want by offering away at least one card of each color you don't, it's not really possible to do that for creatures, or removal, or whatever else your initial pool substantially lacked. If I were continuing with this trade mechanic, I'd look at one-way communication: The active trader gets to state explicitly what she wants, but everyone else responds only with their offers.

    1. Let's talk about the next trade mechanism.
      • I want it to focus players on only trading with their neighbors, to ease their mental scope, and to keep them from physically straining to see what's on offer.
      • I want to make as much trading happen simultaneously as possible, while keeping some structure to the process.
      • I want trades to be mutual. Whether it's two-way or three-, I want each player to agree to that exact trade, so that players feel like they have more agency, and so that players can't get screwed by loopholes or assholes.

      So let's say that each round, every player reveals up to 8 cards. They're all revealing simultaneously, and can add more as they will. You can agree with any of the players closest to you on a swap. You could potentially swap with the player on your left and then swap her pile with that of the person on your right, if you wanted to. You can tell your neighbors what you'd like and what you have to offer, but you can't deal directly with anyone else. Once everyone is done, we start a new round. We do this at most 8 times, probably 6, and that's your card pool.

    2. Just wanted to say that I don't have much to contribute, but I'm watching the updates here with interest.

  7. Ran Swap last night with 6 with a new trade mechanism: Each player could trade without limits, in real-time, with the two players next to them. We set a timer for 30 minutes, and most of the trading was done in 23. There was one four-person train trade made in the last minute that was pretty awesome. The facilitator of that trade went mono-black and went 2-1. All of our decks were pretty sweet, again.

    Four of the players (incl myself) had played last week and so we compared thoughts on the much changed trade. This way was faster, and more natural/intuitive, but required more of the trading skills one might use outside of a game, and arguably made trading from a bad initial pool harder. I'm not sure if it's coincidence or correlation—I think the former—but it seemed to me harder to change from the colors you're initially strong in.

    Notably, I was definitely blue, and wanted to be blue-white because I had two Pillars, two Triplicate Spirits and a Paragon. Both my neighbors also wanted to be white-x, and despite considering the possibility of trading all my white to put me strongly into another color, it didn't happen.

    Also of note, the neighbor to my left traded away some good blue cards before ever glancing my way. That seemed like an unnecessary risk on his part (I might have given him a better deal), but apparently he was responding to a rushed trade between the player on his left and that player's other neighbor. I hadn't really expected anyone to rush trades, and I would expect players to resist such tactics, but maybe that's valid. Which might be more good than it is bad; not sure.

    The swap went well and was a lot of fun. I've gone from optimistic there's something here to confident. I'm not convinced we've found the best possible format, though, and want to try a few more trading mechanisms.

    Next, I want to try another round-based system with formal rules, but through the lens of the neighbors-only restriction, to see how that fares.

    1. Might be a while though, since I'll miss draft night next week, and then we'll have the new set which I expect players to want to draft for a good while before getting experimental again.

  8. How to Run a Swap (as of 10/18/14)
    Gather 4-8 players, preferably 4 or 6. 3 or 9 might work but probably not well.
    Sit around a table so that each player has one neighbor to the left and one to the right.
    Each player opens 3 packs.
    Each player can trade any set of cards with the either of her two neighbors, any number of times, as often as they agree.*
    Once everyone is done trading, add basic lands and build 40+ card decks. Play three swiss rounds or whatever your group prefers.
    Agree beforehand whether you will keep everything you end with after trading, or the group will re-draft rares and foils by record at the end. I recommend the latter.

    *Alternately, players make trades in structured rounds. First, each player in an odd-numbered seat simultaneously makes a lot of any number of cards of one color and/or colorless cards. Each other player may then make a lot, perhaps having discussed, and then each player may trade with the player to her left or her right or not at all. In the second round, the even-numbered players make their lots first. Continue for 8 rounds total, or until everyone is done trading.

    The only real variable right now is how exactly the trading works. I'd love to hear whether you think it is too slow, too restrictive, unintuitive, or anything else that strikes you.
    Have fun!