# Can bridge-style duplicate play work for Magic and other CCGs?

As I understand bridge, you form teams with four players even though only two play at a table. The idea is that there are two identical hands. Two players in your team play North-South with one of the hands, and the other two players play East-West. This removes luck as a factor because if you are very lucky with one hand, opponents will be equally lucky on that hand. Both you and the opponents are going to win the hand, but if your team are the better players, then you should on average win the hand by a larger margin than opponents.

Will this work for Magic: the Gathering and other CCGs? Here's a sketch of how it might work:

• You register for a tournament with two players and two copies of the same deck. For Magic, this would be 150 cards (60 cards + 15 sideboard per deck).
• The cards are numbered and shuffled by a computer, resulting in a random sequence of 75 numbers. Then an arbiter manually stacks the decks, removing cards that are in the sideboard.
• One of your team exchanges decks with the opponents. Now you have the exact same match playing out for both players on the team.
• To deal with cards that shuffle the deck, each time the deck is shuffled, have the computer generate another random sequence, and remove the first card that is searched for. For example if I have four copies of Steam Vents and I search for one of them, the first copy is removed in the random sequence generated by the computer, and the rest of the deck is stacked following the sequence. The sequences are shared in both matches to make sure both decks are still stacked in the same way.
• In the same way, we remove all numbers corresponding to cards not in the deck from the random sequence. This handles sideboarding even if both players don't sideboard the same way.
• Absolutely no communication with teammate once the match starts, since it can reveal which cards are in opponent's hand.
• This can provide some way to tiebreak draws. For example if our team's deck has a strong matchup against opponent's deck, and I win 2-0 while my teammate loses 1-2. We win the match because the net score is 3-2. Alternatively, if I win 2-0 and my teammate loses 0-2, but I win in 10 turns (summed across both games) while my teammate loses in 7 turns, then we lose the match since opponents arguably played better (not ideal since one can reasonably decline to win to play around certain cards, but it is a way to break ties).

Some obvious consequences for this could be:

1. Deckbuilding is less important since the effect of a good deck is washed out. One could show up with a deck of 60 Plains for example and still play an equal match against the opponent. However in this scenario you will need to pilot the opponent's deck as well as they can, and presumably they are going to be more practiced with their deck, putting you at a disadvantage.
2. You need to learn more than one deck, since you'll need to play with more than one deck in a tournament. However this doesn't seem fatal: it's arguable the most skillful player is the one that can play most decks at high level.
3. This might not work in limited, unless teams draft together and then acquires a second copy of the same deck from somewhere.

Presumably if this works it'd work best for digital tournaments since it removes a lot of manual work. On the other hand, perhaps it won't work at all because there's some fatal flaw in the setup. Can anyone see anything that might go wrong?

• Isn't #1 already a dealbreaker? A team of goldfish could show up with a deck of plains and be guaranteed to win exactly half their matches, no matter how good their opponents are. That sounds like a broken game to me. Nov 8, 2021 at 3:58
• @ArcanistLupus how can you be sure they will win half their matches? The setup forces the matches to be on equal footing, but it doesn't guarantee a win for either team. If one plays badly, one is still going to lose (c.f. last bullet point). Nov 8, 2021 at 4:00
• how, pray tell, will a deck with 60 plains and nothing else beat an opponent with a regular deck playing badly? Nov 8, 2021 at 4:05
• That's still a degenerate game where one player is trying to win against a goldfish quickly instead of playing a real game, and the other player is playing nothing at all. And it still favors the player that knows their opponent is a goldfish in game 1. Nov 8, 2021 at 4:08
• @Allure Moreover, without open deck lists, deliberately designing decks around tricking a player who doesn't know its content into piloting it poorly would become a major strategy. Nov 8, 2021 at 7:47

Not sure this is an answer, but it certainly isn't a comment. Sentences in [brackets] are ignorable side-notes.

In team-of-four bridge events, including the world championships, the two tables in the match are set up so that each team puts one pair N-S at their "home" table and E-W at the other team's home table. Then the same hands (6 is a "short" match, 24-30 is a "long" one, 50-60 is a "day-long"; the finals of the world championships are 192) are played at both tables, and the scores for each pair on the team are added together. If the sum is positive, you "won" that board; if negative, you "lost" it. [Most of the time, the scale of the win on each board is important; but that's not relevant here]. Like Magic tournaments, there are hundreds of other tables playing (frequently) the same hands, but only the two tables for your team and your opponent matter to you; the results of all the other matches do not affect your result, just the standings.

[In pair events, the same hand is played multiple times and your results are compared to every other pair who played that hand the same direction (N-S or E-W). The more you beat, the better your score on that board. The reason team-of-four is usually considered more prestigious is that "the field" can be really random. Not quite an accurate comparison, but consider "swiss pairing" with "you will face 3 of the top players in your 9 rounds, and 6 others at random. Matches are set at the beginning of the event and will not change based on results." (also, usually, you won't know the results of anyone else until the end of the event, for what that's worth)]

The difference is that in Bridge (and duplicate Scrabble, and Poker, etc.) the only variables outside of playing skill are the cards, owned by nobody and the same at each table (and the same 52 cards on each deal). It's a matter of skill and judgement (and bidding system, which is kind of but almost totally unlike deckbuilding) as to who does better with each deal presented to them at random.

I guess in Magic, you could get two players to bring in two copies of a deck, and have them shuffled and duplicated and one given to the "away" player of the other team (let's just not do this in Legacy, at least without proxies, eh?), and play two matches, "our decks" and "opponent's decks". You'd end up with either two wins, two losses, or a win and a loss (okay, draws, 1-0 vs 2-0 vs 2-1 wins, ... ignore them) scoring I assume 3-0-1 points as normal.

You'd bring one player that is really good at piloting the deck their team brought, and one player that is really good at piloting "anything they decided to bring", against the deck you brought. I assume decklists are provided to the away player before the match so they at least know what cards are in their deck?

(now, if your "home" deck was a Shared Fate build...?)

Anyone who brought "60 plains" would lose their home match, and likely win their away match, drawing every round and placing well down the rankings. And wasting everybody's time as well.

Of course, the "pattern" of the deal is static in bridge (players have access to their entire unplayed hand at all times) in a way that fails at Magic as soon as the first shuffle takes place - so that the "computer shuffle to ensure the decks start in the same order" will remain relevant on average, what - two turns? So that doesn't lend itself to duplicate play much either (duplicate Scrabble has a fixed (for each match) tile drawing order; duplicate poker plays the same cards in the same order, but with no shuffling until the next hand).

A gauntlet, where everybody is paired against the same AI-controlled opponent, playing the same deck (but a different "same deck" each round) might have some of the duplicate style. But "competent Magic AI" is right up there with "spherical object of constant mass and uniform density": good for theory, in practise doesn't exist.

Deckbuilding is part of playing a CCG. Playing a game of Magic with a deck you had no control over is like playing a round of Bridge where someone else did the bidding for you.

Any proposal for playing a duplicate CCG involves at some point playing a deck assigned to you by the format, created either by your opponent or by the competition's hosts. And that's fundamentally anathema to the concept of a CCG.

• While I agree with this in general, there are a lot of Magic events where you play decks you have "no" control over, especially on Arena. The whole idea behind Jumpstart was "take two precon halves at random from 30 or so, mash shuffle them, and play." Events like the Arena Worlds event played along side the WC, where people got to play with one of 7 former worlds-winning decks are not ubiquitous, but not unheard of either. And as a casual player (years ago), I've always enjoyed "hey, can I borrow a deck?" - I personally hate deckbuilding. So, not anathema. Just not "normal". Nov 11, 2021 at 1:11
• @Mycroft Ah, but in every case but a few explicitly random (and highly noncompetitive) events, you always get to choose your preconstructed deck. Even when you say "hey, can I borrow a deck", you're making a choice to play whatever you're given. Jumpstart (at least on Arena) explicitly has you draft your deck halves. You still have agency over what you play. Nov 11, 2021 at 2:24
• Even choosing to play blind is still a choice Nov 11, 2021 at 2:27
• Some of the Magic Invitational tournaments included “duplicate sealed” events — players are given identical pools, but deckbuild separately.
– Marq
Nov 28, 2021 at 10:05

One of the things that makes duplicate bridge work is that, once the cards are dealt, the random factors are mostly fixed. The decisions people make in the middle of the game don't tend to change which unknown random factors are important and which are largely irrelevant. No matter what decisions people make throughout the bidding and play of a hand, if one on sequence of player decisions it is important which player is dealt the Queen of Hearts, on a different sequence of player decisions it will likely still be important who is dealt the Queen of Hearts.

This is in contrast to Magic or any other games where random factors enter the game not all at once and possibly at different times depending on player decisions. It could turn out in Magic that the order of 2 cards halfway down the deck is crucial in some games and completely irrelevant in others, because some decision made on the first turn meant that those two cards were drawn simultaneously (or tutored) in one game and in different turns in another.

• In a spade contract that can be won by cross-trumping, the QH may be quite irrelevant, while in a heart contract it may be vital, and in a NT contract significant but not vital. Or one pair might use it as an entry making it vital, while a pair taking a different line, or played from the other direction, the entry potential may be irrelevant. All that said, I think the overall point is largely valid. Nov 9, 2021 at 16:29
• any tutor changes the context of the deck too much for any stabilization between games, especially in the first few turns if you don't know the match up since it might change the color of lands you want to grab. If I know it will be mostly about controlling the stack over going fast could be the difference between getting blue source or a red/black source depending on my cards. May 16, 2023 at 16:56

Duplicate bridge exists because bridge is a highly random game - if you don’t, then the deal determines who wins far more than the play. The deal is zero sum - opponents rarely have evenly matched hands.

TCGs are totally different: each player makes their own deck, and they’re positive sum - both decks can have the best cards (equally). That’s most of the skill! A good deck maker who successfully predicts the meta and knows how to play the deck reasonably well will do better than someone who gets the meta wrong or has an inconsistent deck. Duplicate style takes this away.

TCGs are also much more tolerant of ‘gambling’ randomness. Bridge grew up in a time where it was a problem to be linked to gambling. It made a huge effort to differentiate itself from any gambling ties - making it as much a game of skill as possible. TCGs on the other hand are all about gambling - they’re fine with the randomness. Making a deck that reduces randomness of course is also a good thing! You’d actually massively change the balance of the meta this way. Consistency would go out the window!

Further, it’s not possible to take TCG play and remove all of the randomness. You have to shuffle during the game sometimes - Pokémon in particular but all of them have some shuffling typically. So even if you had both players play pre set decks, they’d add randomness the first time they tutored a card. Taking out shuffling the way you describe fundamentally changes the game - particularly Pokémon where there are plenty of elements that involve setting the top card and other elements where you shuffle the hand into the deck.

Finally, the folks who put on the tournaments - either the TCG publishers or folks they sponsor - wouldn’t want this, because the point of the game is to sell cards - having ‘fair’ matches means you don’t have the incentive to build more decks and spend more on cards.

• I do address that ("Taking out shuffling...").
– Joe
Nov 8, 2021 at 16:48

First, duplicate bridge is played in teams of two because bridge itself is played in teams of 2. It is a feature of the game, not the duplicate tournament style, so there is not necessarily a reason to have teams of two here.

In addition, the tournament setup described here does not really reflect duplicate bridge: in duplicate bridge, in different rounds pairs play different hands against different opponents, instead of playing the other side of the same hand against the same opponents. This would be more directly reflected by pairing up decks at the beginning, and having counter-rotating pairs of player play matches with the corresponding pairs of decks. Then each player's performance per round would be compared with other players playing the same deck.

Second, this proposal does not accomplish the same degree of sameness in game options that duplicate bridge does. Even with shuffling being fixed in the way described, the game is practically random from the point of view of each individual player because they have to make choices based on hidden information, and those choices can influence their future options, which can cascade into wildly different games.

For example, say the opening hand contains a Scalding Tarn. Player A plays and activates it on turn 1, and shuffles. Player 2 plays a different land on turn 1, draws a second Scalding Tarn on turn 2, and plays and activates both fetch lands in the next couple of turns. Now the two players will likely see a wildly different sequence of draws for the rest of the game. Those different draws can easily result in the players interacting with their opponents differently in ways that result in their opponents having different options and needing to make different choices.

In short, the outcome of the game can still depend more on luck than skill. In this case, the luck is simply with whichever player happened make choices that would be advantaged by the shared random sequence, instead of whichever player got the better random sequence.

This doesn't even account for other sources of randomness in Magic, such as coin flips, dice rolls, and random choices. How would you even try to align player outcomes for cards like Capricious Efreet or Dementia Sliver?

Third, your proposal of breaking ties based on number of turns in game is fairly arbitrary, and has no connection to any existing way that Magic games are scored. That could easily have a bad effect on deckbuilding or game play as players try to optimize for that scoring condition.

• About 1: I'm pretty sure duplicate bridge is played in teams of four, not two. Half your team plays one half of the hand and the other half of your team plays the other half of the hand. Dice rolls can presumably be handled by creating another random sequence. Finally I don't think Magic has any ways to break ties in otherwise-drawn matches right now. Are you aware of any such ways? Nov 8, 2021 at 5:09
• I can't find anything about duplicate bridge being played in teams of 4. Wikipedia says "In duplicate bridge, a player normally plays with the same partner throughout an event. The two are known as a 'pair'." There's nothing about multiple pairs in a team playing different halves of the same hand. Fixing the sequence of die rolls doesn't account for the different values of the same die roll depending on the different choices made. That's why I brought up those specific cards. Nov 8, 2021 at 5:40
• I see, there is a variant of duplicate bridge in which players play on teams like that. From other things I've heard and read, I don't believe it is the common tournament style. I don't actually play bridge, though, so I don't know for sure. Nov 8, 2021 at 5:48
• Duplicate bridge is commonly played as both pairs and teams ("Teams" means four or more, typically 4 active at a single time with sometimes some "bench" pairs who swap out). The highest level of play is primarily teams (of 4), i.e. the Bermuda Bowl, but probably most tournament play is pairs.
– Joe
Nov 8, 2021 at 16:44
• @murgatroid99 Most bridge tournaments include both pair events (two players playing as partners) and team events (4 or more players on a team, exactly 4 playing at any given moment). Many players consider team games the better test of skill, and for that reason most championship matches use the team form. Club games typically use the pair form. Rubber (non-duplicate) bridge normally is played by pairs. Nov 9, 2021 at 16:21

You are also not grasping the reasons why people prefer tcg over poker/bridge. There is an inherent unfairness to tcg that you don't have in casino card games. If you where a legacy elves player queuing into a miracle deck then no matter what you did you will not win more than a third of your matches. Proper assessment of the meta is crucial we call this favourable matchups.

This why I enjoy MTG over poker. Although I have played a fair bit of both. Tcg gets new cards added all the time. These games are in a constant state of flux. Although metas these days gets solved awfully quick. Your success at tournament tables are determined to great extent before the first hand is dealt.

Trying to change this fundamental aspects of magic the gathering is simply not going to work.