With the release of Core Set 2020, we'll be introducing a new mulligan system for all competitive Magic formats. This new mulligan was tested at Mythic Championship II in London as well as on Magic Online and will work as follows:
103.4. Each player draws a number of cards equal to their starting hand size, which is normally seven. (Some effects can modify a player's starting hand size.) A player who is dissatisfied with their initial hand may take a mulligan. First, the starting player declares whether they will take a mulligan. Then each other player in turn order does the same. Once each player has made a declaration, all players who decided to take mulligans do so at the same time. To take a mulligan, a player shuffles the cards in their hand back into their library, draws a new hand of cards equal to their starting hand size, then puts a number of those cards equal to the number of times that player has taken a mulligan on the bottom of their library in any order. Once a player chooses not to take a mulligan, the remaining cards become that player's opening hand, and that player may not take any further mulligans. This process is then repeated until no player takes a mulligan. A player can take mulligans until their opening hand would be zero cards.
Essentially, each time you take a mulligan, you draw up to seven cards, then put a number of cards from your hand equal to the number of times you have mulliganed this game on the bottom of your library in an order of your choice. Your starting hand will still be down a card for each time you mulligan, but you'll always get to select that starting hand from a choice among seven cards. Unlike the current ("Vancouver") mulligan, there's no scry after you decide your starting hand.
This rule goes into effect with the Core Set 2020 rules update. See below for details on when the new rule will be implemented on tabletop, Magic Online, and Magic: The Gathering Arena.
The goal of this new "London" mulligan is to make games where one or more players mulligan more competitive, especially in cases where players mulligan an unequal number of times. In particular, greater selection of the starting hand will reduce the number of "non-games" where a player's deck is unable to function due to not having a reasonable mix of lands and spells.
We understand that some aspects of this change are bound to be debated, as is the case with any rules change. A big part of what makes Magic such an amazing game is that it's constantly evolving and improving. We, as designers, strive not only to provide new card content but also to continue to update the game's rules and systems to provide the most enjoyable and competitive experiences we can. I'd like to share some insight as to why we're making this change and what we expect the results will be.
Magic's mulligan system has changed several times throughout the game's history. From the early "no land/all land" mulligan, to the "Paris" mulligan, and more recently to the "Vancouver" mulligan, the goal of each change has been to give players a better chance of having a reasonable opening draw leading to a competitive game where either player might win.
As more and more games of Magic are being played these days across both tabletop and digital platforms, we've gathered data that shows that even the current Vancouver mulligan isn't doing as good of a job as it could be in providing a competitive starting point for both players. A player who mulligans once against an opponent who keeps seven cards, in general, is at more of a disadvantage than we're comfortable with. The situation gets even worse for a player who mulligans two or more times more than their opponent. We also dislike that some of those games play out in uninteresting ways where it's clear that the game was nearly over before it began based on starting hands alone.
Over the past year, R&D has been actively discussing and testing ways to improve how games begin. We've tested the new London mulligan internally for more than six months and are pleased with how it closes the gap between a player who mulligans and an opponent who doesn't, and also how it greatly reduces the number of games where a player's deck and strategy simply don't function at all. In this sense, we think of the London mulligan as being a "stronger" mulligan in the player's favor as compared to the Vancouver mulligan.
Testing the Mulligan
Of course, in strengthening the mulligan, there could be such a thing as going too far. A mulligan that's too strong and gives too much agency over a player's opening hand could make too many games play out the same way. There's also a risk that combo decks could abuse a very strong mulligan to much more reliably assemble a combo early in the game. Or that aggressive decks always having their best draw, or control decks always having the right answer, could change metagame balance too dramatically. Plus, the stronger the mulligan, the more often players will choose to use it, which increases time spent shuffling and making decisions before the game can begin.
These are the risks we knew we needed to look out for while testing the London mulligan both internally and publicly. Our hope was that it would be enough stronger than the Vancouver mulligan to make a difference in reducing the number of non-games, but not so much stronger as to have detrimental effects on gameplay as a whole. We still want mulliganing to be something a player prefers not to do if they have a reasonable hand, we just want it to be a little less punishing when a mulligan is necessary.
While we'd been very happy with how the London mulligan was playing in Standard and Limited, we knew that the biggest risks would be in nonrotating formats, where the card pool is deeper, games are generally faster, and the best draws from combo decks are stronger. That's why we chose Mythic Championship II (in London) with its Modern Constructed rounds and open decklist policy as a competitive testing ground. If the new mulligan was too strong, that's an environment in which we believe we'd see it break.
What We've Learned
By and large, what we've seen is that the upsides of the London mulligan outweigh the downsides and risks. Mythic Championship II featured a diverse and healthy Modern metagame, both in terms of what decks players brought and which decks performed well. While some elements of the risks above showed up to small degrees, like more frequent mulligans, overall, we did not see what we felt were large systemic problems arise in the metagame or in gameplay. That combos were assembled a bit more frequently was largely counterbalanced by sideboard cards appearing more often, and it appears as though the metagame was able to self-correct.
That being said, make no mistake we do expect metagames to change somewhat with the introduction of the London mulligan. Any time a game system introduces changes like the new mulligan rule, it affects some decks differently than others. The question is whether the metagame will adapt to those changes and come to a new, healthy equilibrium. So far, all signs point to yes, including for Modern.
We expect the London mulligan to be almost all upside for Standard and Limited. In those formats being down a card in the starting hand is enough of a penalty that we don't anticipate players mulliganing much more aggressively than under the current Vancouver mulligan. It'll just make for more enjoyable games when mulligans do need to happen.
We understand that players have had some concerns about Eternal formats, particularly Legacy and Vintage. We agreed, and gathering data on those formats is part of why we tested the mulligan on Magic Online. From what we saw, the metagames were able to adjust even better than expected. One characteristic of Legacy is that there are very efficient one-for-one answers to most threats, which has the effect of making raw card quantity important. That in turn means choosing to mulligan aggressively to a particular card or combo is more of a cost, and so we weren't seeing that strategy be very successful. Vintage is harder to gather data on due to smaller sample sizes, but we did not see any alarming imbalances jump out. For example, the win rate of dredge did not change markedly even with a greater chance of finding Bazaar of Baghdad in the opening hand. Again, this is perhaps counterbalanced by other decks also more reliably finding their sideboard cards.
Our approach for all formats will be to allow players and metagames to adjust and to gather data and feedback before considering any banned and restricted list changes. Trying to be predictive with changes isn't the right approach for us or for the community. To be clear—we don't currently have any specific cards or decks that we expect to have to target with B&R changes, though it is possible that need for change emerges down the road.
Some members of the community have suggested introducing the London mulligan on a format-by-format basis, but, as designers, we feel strongly that having one unified mulligan rule for all of Magic is the right path forward. Many Magic players try out a variety of different formats over their lifetime, and having inconsistent mulligan rules across those formats would be a substantial added complexity cost. From a technical standpoint, it's also a consistency benefit for both the tabletop comprehensive rules and digital rules engines to have a unified mulligan rule that exists "above" the subdivision of different formats, rather than as a separate rule within each.
We'll be rolling out the London mulligan in all tabletop and digital expressions of Magic with the release of Core Set 2020. Starting with tabletop Core Set 2020 Preleases on July 5, the new mulligan will be used for all play. It will become officially reflected in the comprehensive rules with the M20 rules updates on July 12. Stores may opt to enact the new rule for regular REL or lower events (including the Modern Horizons Prerelease) before the official rules change on July 5 to give players the opportunity to test out the new procedure. Competitive REL events cannot use the new rule until the official change.
MTG Arena will test functionality of the London mulligan in a special event starting June 7 and will fully adopt it for all play formats with the release of M20 card content starting on July 2.
Magic Online will begin using the London mulligan on July 2 after servers come up from scheduled downtime.
We hope you all enjoy playing with the new mulligan as much as we have internally. We understand that change can be hard to digest, especially for a community as deeply invested as Magic's. But in the end, the goal of this change is to make Magic more fun, and we have confidence that it will!