Mulligans

Posted in Latest Developments on August 7, 2015

By Sam Stoddard

Sam Stoddard came to Wizards of the Coast as an intern in May 2012. He is currently a game designer working on final design and development for Magic: The Gathering.

Mana problems are the bane of many players' experience with Magic. At the same time, dealing with mana issues is one of the most important things about making Magic what it is, and I believe it's integral to making the game successful. While the mana system can add a lot of frustration, it also gives us a considerable ability to balance cards and ways to let players gain advantages in deck construction. Players are forced to ask question like "Should I play a greedier deck and get punished for my lands more often, or play something more consistent but risk losing to a greedy deck that hits all of its colors?"

This is one of the core questions in all Magic formats. That being said, I believe the highest levels of frustration come from the pregame procedures, and when it feels like you never really have an opportunity to get started, due to either not drawing enough land, drawing too many lands, or drawing lands of the wrong colors. The goal of improving our mulligan procedures is not to make it much easier to play greedy mana bases, but to increase the chances that decks that have made reasonable deck-building decisions are able to play a real game of Magic every time they shuffle up.

Cosi's Trickster | Art by Igor Kieryluk

One of the frustrations with early Magic was that you only had the opportunity to mulligan zero or all-land hands, so having something like six spells and a Kjeldoran Outpost basically translated into "you lose the game." Not a very fun pre-game procedure. The first mulligan rule change was tried out at Pro Tour Los Angeles (for Limited) and ended up being kept for Pro Tour Paris, from which it gets its name. It was a big change, but one that was needed.

The Paris Mulligan gave players the ability to throw back un-playable hands for the increased chances of having something playable. Players still took a healthy hit to their win percentage, but it was much higher than basically zero, as was the case in the example above. This was a huge improvement, and I think one of the things that led to competitive Magic being successful. Successful enough that we haven't made any major changes to it in eighteen years. Just because we haven't changed it, though, doesn't mean there isn't room for improvement.

Continuing Quality Improvements

Just because the current mulligan rule works, however, doesn't mean it is the best one possible. Every once in a while R&D likes to look at the things we do and try some new things out to make sure that we couldn't be doing better. The mulligan rule is no exception. Mulligans and what they represent are important to Magic. We want players to be rewarded for choosing things like the correct curves, land counts, and mix of spells. And the randomness of the opening hand does that. Try to stick too many high-casting cost spells or too few lands in your deck to get an advantage, and you will find that you have to send more hands back than an opponent who has a more balanced mix.

While that part of mulligans is good, the part where some number of games of Magic play out where one player has very little agency due to a bad draw is not. A goal for us when looking at improving the mulligan rule is to keep some amount of the reward for having a good deck mix, while reducing the number of games where one player does basically nothing. The player who takes a mulligan should be at a disadvantage, but it is all about finding the correct amount.

A second goal that you might not think about is to keep tournaments going at a reasonable pace. Shuffling, especially pregame shuffling, takes a long time, and constantly adding those actions to games just increases the number of games that go to time. One of the reasons we changed the rule from "one player makes all mulligan decisions before the other" was to speed up the total pace of play and reduce the number of games that go to time and delay tournaments. While that rule was somewhat contentious when it was announced, I believe that—looking back on it—it seems both obvious and wholly an upgrade.

The most obvious change to the current mulligan rule, and the one we hear suggested the most often, is "shuffle any number of cards back once, draw that many cards." While this would probably greatly reduce the number of games that come down to one player stumbling, it would do so at the cost of warping formats and how Magic decks are built. While I believe this would greatly improve the experience for many kitchen table players, it would quickly be broken at the top levels of competition.

Serum Powder | Art by Matt Thompson

Being able to throw cards back is much stronger in a combo deck, for example, where you may not want to draw two Show and Tells. Just keep the part of the combo you have and put the rest back. Decks would change their land counts depending on what they were trying to accomplish. Hyper-aggressive decks could throw back every land above one, and control decks could throw back all of their late game cards for the opportunity for more land or early removal spells.

The other version I have heard bandied around now and then is one that cuts out the middle man. It lets players either put a land on top of their library, or exile some number of cards for a land. The problem with this is that you end up with "two-land decks" that are built around using this rule to get their mana, and then draw a spell every turn.

Our goal with these changes was not to totally change how Magic works. Instead, it was to make some minor changes to how we handle mulligans to improve the overall game play and reduce the number of non-games. This doesn't mean that player shouldn't have a penalty for mulliganing. Players should have the opportunity to recover from a mulligan over the course of the game with good play, rather than automatically losing due to not having a reasonable mix of lands and spells.

New Mulligan Rules Goals:

  1. Reduce the number of non-games resulting from mulligans.
  2. Don't dramatically increase the frequency of player's mulligans.
  3. Keep the basics of Magic deck building and game play the same.
  4. Have the same mulligan rule for Limited and Constructed.

I do want to hit on #2 and #4 very quickly. For #2, this goal is to make sure that whatever rule we put into place won't be so much stronger than the current one that players mulliganed more frequently, or that it was often better to mulligan than not. In the "mulligan any number once" example, it would be pretty rare that somebody wouldn't mulligan at least a few cards. That definitely violates #2 and #3.

For #4, I believe the greater total number of non-games happens in Limited. And while it might be possible to have different mulligan rules for each format, it isn't something we're particularly interested in at this time. While I won't say it's a total non-starter forever, the fact that mulligans work the same in all of our high-level competitive formats is important to us.

With that out of the way, I want to talk about a few of the experiments we tried when coming up with potential new mulligan rules.


7-7-Stop

This came from looking at how frequently players mulliganed to five, and their win percentages. The number of games a player mulliganed to five was low, but so was the win percentage of players having to do so. These also made up a lot of the non-games. One possibility was to just improve the win percentages of people who mulliganed once (by giving them the extra card) and hoping that would make up for the number of times someone just loses if they have a second bad hand.

What we liked: This dramatically reduced the amount that a mulligan hurt players. It wasn't zero, since a player who keeps their seven will have—on average—a better seven than someone else who doesn't have a choice in the matter. But it was a big improvement in games where either player mulliganed exactly once.

What we didn't like: One of the goals was to reduce the number of non-games. And while this went a long way to improving the games where one player mulligans once, the non-games would be much more punishing than ones under the current system. We don't want to force players into just passing their first turn without any ability to change that. As a whole, people had a pretty negative reaction to this system, even though it was pretty powerful. This felt like an interesting experiment, but not one that would make the player base happy as a whole.


7-6-6-5-5

The start of this mulligan rule was the idea that going to six was a disadvantage, but not nearly as bad as going to five. By giving players a second six-card hand, we still give people a penalty for mulliganing, but it's not nearly as bad. Also, a number of the six-card hands that people lose with are only kept because they are bad, but better on average than a five. One land and several two-drops? Well, probably better than a five. This let players return hands like that in the hopes of getting six-card hands that were actually competitive.

What we liked: This greatly reduced the number of non-games by giving players who mulliganed a much higher chance of being competitive. There was still a disadvantage to mulliganing, but it was less likely that a player would lose quickly to mana problems.

What we didn't like: This rule encouraged players to mulligan very aggressively in Limited games, and even more so in Constructed. What we found was that people in Limited would be more likely to mulligan mediocre sevens with poor curves, just because they got so many additional shots at having a good six. In Constructed, it meant that the range of hands people would mulligan would go up very quickly, especially in situations where you already had a dead card in your hand. Drew a Languish against a creatureless control deck? You might as well take your two shots at six. Looking at older formats, it also highly incentivized people looking for sideboard answers. Cards like the Leylines got much stronger when you were punished much less for mulliganing.


7-7-7 (Shuffle x Back)

This was attempting to do something similar to the scry mulligan we used at Pro Tour Magic Origins, but in a way that was more powerful. Basically, each time you mulligan you draw up to seven, but get rid of cards that you can't use (down to the appropriate smaller hand size per mulligan), thereby increasing the chances that you will have a reasonable hand.

What we liked: This seemed pretty close to right power level for Limited, but had some problems. You generally ended up shuffling your most expensive card back, but if it wasn't obvious, then the decision on which to shuffle back was pretty hard—and made this take a lot longer than a regular mulligan.

What we didn't like: This mulligan was way too strong in Constructed, and encouraged big changes in deck building. Perhaps the most notable thing was in Modern and Eternal formats, where sideboard hate got a lot stronger since you could shuffle extra copies back into your decks. Similarly, combo decks got a huge advantage since they could mulligan away possibly useless cards. In one of our biggest rules violations for changing the mulligan rule, it clearly changed the parameters for deck building, and would have a profound impact on how older formats played out.


Scry Mulligan

It was around the time that we were looking at these mulligans that Pat Chapin happened to suggest the "scry mulligan." He, of course, had no idea that we were planning on making scry an evergreen ability, but it lined up well with that change.

What we liked: The biggest advantage of this mulligan was that it was meaningful but not too strong. It does a good job, overall, of increasing the chances that you will hit your early land drops, but without being too easy to abuse. A high number of the non-games result from one player stumbling on lands for a turn or two, and this should help greatly with that problem.

Concerns: While we like this mulligan a lot, we haven't yet (at the time of writing) had time to look at how the Pro Tour played out to see if there were any problems. Part of the idea behind using it at a Pro Tour first is that if there is something we missed, or something that is prone to abuse, it will show up there. We will be looking at both how the Pro Tour played out, as well as the reactions from the players regarding the rule. I don't know when we will have a decision based on this, but we hope to have new information fairly soon.


That's it for this week. Next week, I'll be back with some of our Magic Origins Future Future League decks, the decks R&D thought would be a part of Standard.

Until next time,

Sam

(@samstod)

Latest Latest Developments Articles

LATEST DEVELOPMENTS

Changes by, Sam Stoddard

Hello and welcome to another edition of Latest Developments! Today I'm going to talk about several kinds of changes within R&D and how we deal with those. Card Changes From the day ...

Learn More

Latest Developments

Things I've Learned by, Sam Stoddard

Hello, and welcome to another edition of Latest Developments! This week is the five-year anniversary of me joining Wizards of the Coast as a contractor on the development team. My officia...

Learn More

Articles

Articles

Latest Developments Archive

Consult the archives for more articles!

See All

We use cookies on this site to enhance your user experience. By clicking any link on this page or by clicking Yes, you are giving your consent for us to set cookies. (Learn more about cookies)

No, I want to find out more