Breaking Through – Unorthodox Mulliganing


I understand right now that this article will stir up some negative feedback and/or controversy. I also understand that my views on mulligans are not going to be the same as the majority of people out there and because of this I have a lot of work ahead of me. If nothing else, I urge you to follow along here with an open mind, as even if you eventually disagree with my methodology, information is still valuable, even if this teaches you what not to do. Also, due to the fact that I will be fighting off of my back foot here, if something seems unclear, please note it in the comments and I will try my best to elaborate, but hopefully that won’t be necessary.

When Paulo posted an article (or like 10) about mulliganing here, I offered to help out when asked of course. The feedback towards his article was very positive but quite a few people wanted to take a closer look at my mulligan philosophy in particular, so that is what I plan to lay forth today. Mulliganing, like everything in Magic, is a vital skill that can directly lead to wins and losses. That said, I feel the popular analysis of mulligans has gone a bit overboard and actually begun to work against making correct decisions. I hold a fairly unique view that is due in large part to my “rogue” nature of creating my own set of rules and restrictions. It is not that I do not respect the popular views, but I do not always feel they are correct, at least anymore.

Prior to the M10 rules changes actual took effect but were still made public, everyone was up in arms about how the game was being dumbed down. In essence, the argument was that taking away damage on the stack led to less decisions and thus less skill towards the game as a whole. Now that we have had time to play with the new rules, many people think that taking away damage on the stack ADDED to the overall skill and decision process of the game. You see, the old rules, while appearing to add more decision trees, ultimately resulted in the same line of play over and over again. While learning to stack damage may have been a vital skill when it first was released, after years of doing just that, the act turned from skill to habit, and games reflected this. Players would know how to snap-call damage on the stack and also to how to play around it from the opposite side of the table. The game had become simpler once a small rule was nearly universally followed.

Transition over to mulligans and I believe we are in the same boat. People understand that mulliganing is a vital skill and that most professional players claim it is an area where many people need to improve. Because of this however, some blanket rules have taken a hold of mulligans that lead players to throw the hand away too quickly. Most one-landers get shipped back without a second thought for example, primarily due to the fact that the general consensus tells us that one-landers are bad. But if this has become the rule with few exceptions, are we not walking ourselves into a corner where damage still stacks? Have we not taken a skill intensive technique and removed skill from it by blanketing hands that are mulligans or not?

LSV’s ‘loose’ Keeps

I hold some unique views on mulligans, but they are not so unique that I never see others use them. LSV for example, keeps many of the same hands I would during his draft segments, with things like one-landers on the draw. Here is an actual hand kept by LSV during a recent Saga draft on the draw, game 2:


It is easy to look back in hindsight and claim that you would have kept this, but for many or even most players, this is simply not the case. Yet in my eyes, this hand is not even close to a mulligan despite having the initial appearance of sketchiness. Players have stopped analyzing their hands in a full context and have instead started to keep hands analyzed in a vacuum. But because we are told that good players mulligan, and that mulliganing is a skill people lack, we feel justified when we OVER-MULLIGAN. Yes I said it, and even all capitalized and stuff. People over-mulligan hands due to the stigma telling people that good players mulligan and therefore justifying every mulligan someone may make, but this is clearly not the case.

While it is true that players should mulligan speculative hands, the definition of a speculative hand is so misunderstood which leads players down a path where they follow a set of rules. Anytime you have specific parameters in Magic, you are bound to be in trouble, as there are too many unique situations and scenarios to accommodate them all correctly. Just as you cannot say that you should always block an opponent’s bear with your bear, you cannot claim that mulliganing every 1 land hand on the play is correct. Every mulligan decision should be taken in full context, as without assessing everything, you may reach an ill-informed decision.

Bant Keeps

Whether I am on the play or draw is probably the biggest factor for me when I decide to keep a hand. Hands on the draw are typically much more forgiving due to the entire extra turn you are granted by the top of your deck. Most of the “questionable” hands as documented by others toward me have been on the draw where I don’t think they are that big of a deal. In Oakland during the top 8, I was up against Zoo, in game 3, on the draw, and had already mulliganed to 6. I debated it for a while, but ultimately kept this hand:


On the play I would never keep such a hand, but on the draw it provided me with all of the tools I needed to beat Zoo, with the exception of any one land, of which I turned to the top of my deck. I of course mised a Temple Garden for my first turn and blew my opponent out as he didn’t even have a one drop (This is something I will talk about later). There is a ton of contextual clues that led me to keep this hand, but let’s begin with being on the draw. (And no, I am not being results-oriented here, but an example was needed.)

On the draw I have 22 lands that are live, as well as 3 more Path to Exiles that I would be fine drawing. That amounts to 25 good draws out of a remaining 53. Now while I am not the type to start breaking everything down into numbers, having some semblance of what your projected outs are is important to help make these decisions. For convenience sake, we will assume I have a 50% chance to draw a live card in 1 draw step and therefore (roughly) a 75% chance at drawing a live card in 2 draw steps. It is at this point however, where one must consider that if they do draw a live card, the chances of victory are very high. Due to the Path to Exile, even missing on the live card until draw step 3 grants you a very good chance at winning, since you have 2 additional spells to work with at this point as well. Given this, you now must consider the option of mulliganing to 5.

First of all, let us not kid ourselves in assuming that every 5 card hand is going to be good, as there are certainly chances that our 5 card hand is also a mulligan, or also in the same situation as our 6 card hand but with 1 card less. Obviously if every time you mulliganed to 5 you were guaranteed a solid hand, mulliganing would be correct more often. But since that is not the case, and there is some chance we just have a bad hand, that must be taken into consideration. Even if we do have a keepable hand, are the chances of victory for that hand approaching 87.5% like our other hand likely does?

Now we must shift our context from simply being on the draw to looking at the matchup at hand. You are playing Bant against Zoo and are on the draw. In this scenario, you are looking to 1 for 1 your opponent until more powerful cards, like Rhox War Monk, Umezawa’s Jitte, or Jace, the Mind Sculptor come online. Because of this, going down to 5 means one fewer 1 for 1 even if you do have a decent hand. Another thing to take into account is that in our hand is one of the few 2 for 1’s available to us in the early game in Engineered Explosives. While we need another land to set it off, we can still get value out of it by giving ourselves time if our opponent plays multiple 1-drops. You cannot allow a deck which intends to win the damage front also take a lead in the card advantage front or else winning becomes exponentially harder.

I will not act as if there is no risk involved in keeping hands like the above, but there is also risk in mulliganing. It is up to the player to weigh the pros and cons and decide for himself, but there should be no hard and fast rules that dictate that decision. A similar situation came up in Paulo’s mulligan article where players had to decide to keep or throw back this hand on the draw against Zoo.


Nearly everyone else claimed it was an easy mulligan but I just can’t agree. What more can you ask of a hand than a almost sure shot of 87.5% to win against Zoo? Remember, if you aren’t hitting your lands for the first few turns, you are drawing spells, which means once you do hit your land, you have 7 cards to work with. Obviously drawing a land on turn 1 is ideal, but this hand can literally miss for 2 turns and still absolutely destroy Zoo.

Their first two threats are irrelevant, as you are not ramping Zoo into something scary with your Path to Exiles. Afterward a Goyf plays cleanup with a Bant Charm backup in the not too distant future. Sure, there is some chance that your going to draw 3 straight spells and lose the game. This is probably around 10% since the remaining 2.5% can be mitigated by the opponent having a loose draw as well and your Path’s clearing a.. well.. path while you dig some more. But if we mulligan, can we assume a better than 87.5% chance at victory? Do we not have a bigger chance at losing than 10% considering the variance in the opening 6 combined with the variance in cards off the top of our deck?

Just because a hand is not ideal does not mean it is worthy of a mulligan. When you keep a hand like the one above, but with a second land over Vendilion Clique, you are still needing some certain cards off the top of your deck. Assuming the probability is in your favor, why not keep a hand that will 100% be better than any 6 card hand so long as you draw a land in the next 2 draw steps and will likely be better if you draw your land in the 3rd draw step. Obviously you can mulligan here and open up the nuts and “prove” me wrong, but you are just as likely to end up with a no lander and be forced to mulligan once again.

Obviously we are not looking to keep blatantly bad hands, like no landers or 1 land and a bunch of 4-drops, but the math supports our mulliganing here. In the above scenario though, your hand is not bad by any means, it just needs some help, and that help is more likely than not, by a significant margin. Without doing the proper analysis, you end up relying on probability once again with a new hand.

One thing I avoided mentioning during the first example hand was that my opponent tanked on the play before keeping his 7. Now while I don’t suggest you rely on “reads” of body language or anything like that, if you are torn between decisions, take advantage of the fact that your opponent is human. This was not Magic Online and I watched my opponent’s face as he decided to keep. From this, the only thing I could tell was that his hand was marginal. I extrapolated that he either had no 1-drop, or it was a wimpy 2/3 and not a Wild Nacatl, or he was stuck on 1 land as well. Obviously the latter is worse for me if I have to Path, but when I opened my 6 card hand up, I could not just forget about the information my opponent handed over to me. In true results oriented fashion (in before the forum posters) my opponent whiffed on his one-drop and my hand became amazing.

You are allowed to use more information than just what you hear everyone talk about constantly. There is a human interaction to this game and to always ignore it is a mistake. Along this same line of reasoning, while it did not occur in my example, if an opponent mulligans on the play, you can use that to aid your decision making process as well.

Obviously some hands increase in value slightly if an opponent mulligans. The general trend is to mulligan a hand that you would otherwise mulligan regardless of an opponent’s mulligan behavior, but some specific cards change this. Thoughtseize effects for example, become much more potent with a smaller hand size from the opponent. A hand of 5 lands, Dark Confidant and Thoughtseize may seem weak on the draw, but if an opponent has already mulliganed, you should at least take the time to reevaluate your options. Again, we are primarily concerned with breaking the stereotypical mold of mulligans, so it is not actually about whether you end up mulliganing, but what you do en route to do so.

If you run through all of the data and still feel it is right to mulligan, by all means, go right ahead. But do not rely on too many shortcuts to get you there. Falling into predictable patterns only harms us when we apply them to such a fluid game. You may be criticized initially for some “different” techniques and thoughts, but ultimately, if those behaviors lead to more wins and a better understanding of the game, who cares about the appearance of them on the surface? Mulliganing is important, but this does not translate into mulliganing more. Mulligan smart and you will begin to see a difference between the two. Thanks for reading.

Conley Woods


Scroll to Top