The New Mulligan Rule

If players couldn’t mulligan, a large percentage of all games would end before both players had a chance to meaningfully develop their game plan. Some games ending this way is fine, since guaranteeing that everyone gets to deploy their cards on their own terms creates its own issues and erodes the tension in Magic between power and consistency. Mulligans attempt to relegate these games to a corner of the tournament Magic experience, rather than the center of it.

The first time I learned of a mulligan rule in Magic, it was explained to me as something like, “If your hand has no land or all land, you can show it to your opponent to confirm, then toss it back, shuffle, and get a new hand of 7 cards.” At some point, this became 0, 1, or 7 lands instead of just 0 or 7. The few games I managed to win back then were likely the result of my opponent not being able to mulligan unplayable hands that didn’t conform to 0, 1, or 7.

One day around 1997, I found out that the local “big” tournament center in Southern California, The Costa Mesa Women’s Club, would be holding a tournament with cash prizes intended to test a new mulligan rule. This new rule was the mulligan rule you’re now familiar with, the Paris mulligan. The Paris mulligan got its Pro Tour debut in the Los Angeles area as well, aboard the Queen Mary, a tournament venue that was the location of so many great Pro Tour moments and crazy stories.

More about the history of the Paris mulligan can be found here.

The Paris mulligan rule was a great step toward reducing the number of “non-games” of Magic we’re forced to play. Mulliganing under the Paris mulligan rule is still costly, but not nearly as costly as it was under prior rules. The next evolution of the rule pushes that concept a little bit further.

Testing is Underway for a New Mulligan Rule

The newly proposed rule is that after both players are done (Paris) mulliganing as usual, any player who did choose to mulligan will get to scry 1 before the game begins. The following new language has been added to the end of the mulligan rule if you’re curious about the actual rules text: “Then, beginning with the starting player and proceeding in turn order, any player whose opening hand has fewer cards than his or her starting hand size may scry 1.”

For now, this rule is only in effect for Pro Tour Magic Origins, but I suspect this rule or something like it will become the default soon enough.

One of the things I hate about Hearthstone is that, unlike Magic, so many of the games feel exactly the same. I used to have to talk in abstract terms when explaining what I liked about Magic’s variance and its mulligans, but now I can point to this concrete example of how smoothing out the variance with a lax mulligan rule moves too many games into safe, familiar territory. Magic: The Gathering is for the bold adventurer who wants to encounter the unfamiliar, to occasionally have his or her back against the wall with things not developing as planned. Checkers and Hearthstone are out there if you need some comfort.

The new scry 1 rule preserves this key feature of competitive Magic while smoothing things out just a bit to allow for those “back against the wall” games to be a little closer and for those come-from-behind games to be a little more frequent.

What Was Wrong with the Old Rule?

The old Paris mulligan rules (the current rules for now) were not horribly broken, but players mulliganing to 5 probably won something like 30% of the time and even a mulligan to 6 might drop your chances of win to around 38%. These figures would of course vary by format and by mulligan skill level, but the size of the effects were large everywhere. Mulliganing is very costly.

I will miss this aspect of the mulligan procedure because I felt like I was pretty well calibrated to the cost of mulliganing and made it a point of emphasis in my preparation to learn to mulligan enough but not too much. Often this left me keeping hands others would mulligan. I read Keep or Mulligan and agree with well under 50% of the decisions PV presents (he mostly uses borderline cases which exaggerate the effect, I’m not saying in general we would disagree more than half the time).

Now, mulliganing will be correct more often than it was, as your downside scenarios are less likely with a scry 1 than they were without it. I feel the change is bad specifically for my win rate, but it’s good for the game and I like my chances at figuring out how to adjust. Mulliganing will still be very difficult and even what to do with that scry 1 might be somewhat tricky half the time and extremely challenging ~20% of the time, if I had to guess.

Did this Go Too Far?

I’m sure some people will claim this goes too far, with the risk being helping combo decks and helping good players (who might need to get mana screwed to lose the occasional match to a newcomer). This does seem to be the direction of the changes, but in terms of the size of the change I don’t think it’s large enough to be a significant drag. 7 remains better than 6, and it’s not like scry 1 is a Demonic Tutor for the combo decks. The midrange decks that mulliganed well and were very redundant take a small hit, but their dominance in recent Standards can afford to be softened ever so slightly.

Specific scenarios I’ve seen mentioned include Delver decks on the play getting extra help at flipping Delver. This represents a total of 2-3 extra damage (1-turn early flip) in cases where the Delver player is on the play with a Delver and a fetchland was not needed to cast it, and the top card was a non-spell you didn’t want. It’s just not a big deal. Fetchlands in general get a little bit worse as any time you’re on the play you could lose your scry advantage or be forced to hold the fetchland. Just like redundant midrange, if anything can afford to lose a little bit of edge it’s the fetchlands.

This change is at least worth experimenting with at PT Magic Origins. I think the PT is the perfect stage to test it—high stakes, strategic thinkers, and a rule that we know isn’t a large enough change to be a threat to ruin the event. I think the experiment will be a success.

A Couple Things that Would Help this Process

• Pairings boards should state who “won the die roll” by simply noting which player gets to choose play or draw. If the reporting software can’t fairly randomize we’ve got bigger issues, and I’m sick of trying to convince players to do “odd/even” or to roll the same dice I rolled for high roll. What is the point?

• Match slips should ask the players to report who went first and how many mulligans each player took for each game played. This is info Pro Tour players should already be tracking, would be fun to have for people interested in the statistical reports WotC could create, and will help them quantify the impact of the evolving mulligan rules.


Scroll to Top