How to Fix Magic

Ryan Normandin
May 29, 2020

***Disclaimer: I love Magic, have a lot of respect for Wizards R&D, and understand that building formats is tougher than I can imagine. This article certainly contains criticism, but I hope it is all constructive.***

Relative to Magic's full history, the last five years have stood out as being particularly unusual. After the late 90's, there were zero bans in Standard until Skullclamp got the boot in 2004. The article in which the banning of Skullclamp is explained is worth a read; for all the improvements in internal processes that Wizards of the Coast has made in the sixteen years since, the problems sound remarkably familiar.

“In hindsight it should have been obvious. Part of the problem was that our views of the card were tainted by its earlier incarnations, which were very bad. The other part of the problem was that, even without old bias, the card's power is not that easy to parse at first glance.”

“By no means am I saying that Skullclamp's hidden power is an excuse for us missing it. We're paid professionals—we're supposed to find and fix stuff like this. But we were entrenched in our own internal metagame and this card slipped through the cracks.”

“…Skullclamp was a non-entity to us.”

Skullclamp was a perfect storm. It was a new card type that Wizards presumably struggled to evaluate early on and that began in the file as a truly horrendous card that hadn't even seen play in limited. Because of their own bias and perhaps because of the newness of the card type, they missed it.


In 2005, Wizards banned Ravager Affinity: throwing all six artifact lands plus Disciple of the Vault and Arcbound Ravager on the chopping block. They discussed the banning here. There is foreshadowing of their present-day ban philosophy, along with some interesting lines.

“But in the past three months R&D and the DCI have been reminded that Magic is not a series of balanced equations, spreadsheets of Top 8 results and data of card frequencies. Magic is a game played by human beings that want to have fun. One of the most damning statements that can be made about a game is that it is not fun, and that's exactly what we've been hearing lately about Standard.”

With the banning of Skullclamp still fresh in the player base's mind, Aaron Forsythe expressed that they were afraid that frequent bannings would lead to a loss of player confidence.

“We like to avoid having to solve problems by banning cards, as that leads to a culture of fear. We certainly don't want people to start believing that all the good cards they own are in the crosshairs of the DCI.”

Forsythe goes on to discuss a meeting where they planned to ban only the Artifact Lands, leaving Ravager free. 

“…some vigilant playtesting quickly showed that such a move would have been a mistake… The worst thing that could happen, in our eyes, would be for people to come back to Standard, full of hope and under the impression that Affinity was dead, only to lose to a weaker-but-still-potent Affinity deck in Round 1 of Regionals.”

The article then goes on to discuss the restriction of Trinisphere in Vintage, including this wisdom on format development:

“Vintage, like the other formats with large card pools, always runs the risk of becoming non-interactive, meaning that the games are little more than both players 'goldfishing' to see who can win first. Trinisphere adds to that problem by literally preventing the opponent from playing spells. We don't want Magic to be about that, especially not that easily.”

Can anyone think of any other three-mana permanents that prevent opponents from playing spells?

The article concludes with the hope that “nothing” will be banned for the foreseeable future, which was borne out. I began playing Magic around 2005, and as I met other players who had been around longer, I often heard stories of the days when mass bannings had taken place to save the game. But, luckily, it seemed that Wizards had fine-tuned its internal processes sufficiently to generate fun formats at a reasonable power level. We had Ravnica, Innistrad, and Zendikar, certainly high points in Magic design. Of course, the Zendikar block brought Jace, the Mind Sculptor and Stoneforge Mystic, which, after the rotation of Bloodbraid Elf, were banned.


Aaron Forsythe, still Director of R&D today, again writes the article explaining the decision, starting off with a line that, for those of us living in 2020, is sad to read.

“I would have preferred to go through my entire career as Director of Magic R&D without ever having to ban a card in Standard. Sadly, that was not to be.”

The Jace/Stoneforge ban was interesting because, despite an enormous amount of homogeneity in the metagame, the article references arguments that nothing needed to be banned because “the format is very interactive and skill-testing right now.” Nothing inherently broken was happening; players weren't dying on Turn 3 to Tolarian Academy or Turn 4 to Arcbound Ravager. And yet, the article goes on to reveal that there had been drops in attendance at PTQ's, Game Day, and even Friday Night Magic. Again, the article points out that banning cards is something they hate to do because it is an “…admission of grievous mistakes on our part, it hurts consumer confidence in our product.”

Then, the piece references an article written in 2003 by Randy Buehler, who took the devil's advocate position, arguing that banning cards was actually good and meant that R&D was doing its job. After all, “…Magic is more fun when there are powerful cards floating around so it's a mistake to constantly weaken cards just to avoid the threat of sometimes having to ban one.” The argument goes on to outline that innovation is vital to the game, and the more innovative a mechanic is, the more likely it is to be misunderstood and accidentally broken.

This is an argument that has been made often in today's world, with Mark Rosewater even posting a poll on his Twitter account asking players if they would rather have balance or innovation. Randy Buehler goes on to reject this argument, with Forsythe echoing the sentiment that “all those goals can be met without resulting in bans—that's our job, after all…”

The article goes on to explain that Jace's power was missed because Planeswalkers were a newer card type at the time (Jace was the fourteenth), and they had deliberately printed answers for it: Phyrexian Revoker, Despise, Hero of Oxid Ridge, Thrun, the Last Troll, and Hex Parasite. Forsythe admits that the answers were too weak and narrow to deal with the threat. He then praises the need for catch-all answers such as Oblivion Ring and Pithing Needle. Stoneforge Mystic, on the other hand, was R&D still struggling to balance Equipment properly, which is why, post-2011, Equipment are largely not Constructed-playable, though the recent printing of Shadowspear is a triumph of creating a reasonable power level for an Equipment.

Forsythe concludes by acknowledging that the world has changed; the rise of the internet and the explosion of Magic's popularity have led to metagames being solved faster. But, he also says that one of the great things about Magic is that “…even if one expression of Magic begins to fatigue… there are still many other fun ways to engage the game, whether through Limited play, other Constructed formats like Legacy…” Based on this, it would seem that special care should be given not to break all the formats simultaneously.

Magic again settled into a stable stretch, with no Standard bannings until 2017. And then… the floodgates opened.

January 2017 – Emrakul, the Promised End; Smuggler's Copter; Reflector Mage


April 2017 – Felidar Guardian

June 2017 – Aetherworks Marvel

January 2018 – Attune with Aether; Rogue Refiner; Ramunap Ruins; Rampaging Ferocidon


October 2019 – Field of the Dead

November 2019 – Oko, Thief of Crowns; Once Upon a Time; Veil of Summer


In the span of two years, Wizards banned more cards in Standard than had been banned in the previous twenty. If banning cards is “an admission of grievous mistakes on our part,” then what is banning thirteen cards spread across eight sets in three years? And that's only Standard! While Wizards has openly admitted they do not test for older formats due to a lack of resources, which makes perfect sense, cards like Hogaak, Arisen Necropolis and Arcum's Astrolabe are difficult to explain away.


In addition to bannings, the word “unfun” has been thrown around a lot to describe Standard over the last few years. All this begs the question: what is going on? Is power level truly linked to fun? Are increased bannings an inevitable consequence of increased innovation? I believe that Standard is actually suffering from a blend of problems, which makes it difficult for people to pinpoint exactly why everything just kind of stinks. All these problems stem from Wizards repeating old mistakes.

1) Don't Cheat on Mana

In this 2011 article, Mark Rosewater shares what he believes to be the fundamental reasons why Magic is such a good game. The mana system is one of these things. It is easy to breeze by the mana system, as it seems so obvious; how could it work any other way? Most games with a strategy component, whether they be card games like Hearthstone or board games like Monopoly or Risk, allow players to gradually build up resources. They can then use those resources to make more powerful plays. As Rosewater states, mana controls the flow of the game.

“Imagine a world in which it doesn't exist… The end result of this is that the game would probably end on the first player's first turn. Maybe the opponent has enough reactive cards to stop the first player, but the abuse of this system leads to games that end very quickly.”

Rosewater also argues that it helps the game to become more dramatic over time. He uses a metaphor describing a time he learned how to do stage combat. What makes for exciting fight scenes in storytelling? I would take this one step further and ask what makes for exciting storytelling? The answer is that “A fight needs to start small and build… Each section needs to one-up the one before it.” First, it's hand-to-hand combat, then it escalates to a weapon, and then to a better weapon. Additionally, the scenery changes; the fight might begin on the ground but ends in the sky or in a volcano. Finally, there's a climactic battle where viewers see something new, where one fighter does something awesome to best another. Also of note, during the entire fight, the person winning goes back and forth. When one person pulls ahead, there exist comeback mechanics for the other player to catch back up. Escalation is exciting, and it would be impossible without the mana system.

Finally, the mana system controls the players' options. Earlier in the game, players typically cast one spell per turn. Later, they can cast multiple spells per turn. This further manages game progression and leads to the exciting escalation.

There is a reason that every strategy game and most stories in existence start small and escalate; it's because it is fun. It's what players like. It makes for good gameplay and great stories. Think back to your favorite Magic moments; are they games where you absolutely annihilated your opponent, or were they games with lots of back and forth, with a gradual escalation that allowed you to defeat your opponent because of good resource management, clever plays, and a little bit of luck?

The biggest problem with modern Magic is that R&D seems to have abandoned this fundamental lesson. Imagine a world where all the cards that violated this rule had never been printed.

Emrakul, the Promised End could be played as early as Turn 6. It rendered all decisions made before it entirely irrelevant, and the game devolved into a grueling Mindslaver battle. They were not fun; they were repetitive, a slog, and they rendered nearly every non-Emrakul card in the deck irrelevant.

Aetheworks Marvel was hated by players not because its win rate was too high, but because it was utterly unfun. On Turn 4, a player could cast Ulamog, the Ceaseless Hunger, a ten-mana spell six turns early. This is an absurd violation of the mana system.

Energy (which will be coming up again) is free. Players make actions for zero mana investment. Bristling Hydra had no play to it; there was no need to hold up mana to activate the Hexproof ability. Instead, if an Energy mirror devolved into a Hydra subgame, it often involved a stupid chain of six Harnessed Lightnings on the stack. Longtusk Cub could similarly pump itself at instant-speed for no mana.


Field of the Dead generates substantial bodies on a difficult-to-interact-with permanent type for zero mana investment, and it isn't even legendary.

Once Upon a Time is a free card selection spell. Ponder and Preordain are banned in Modern, but Once Upon a Time was printed for Standard?

Hogaak, Dig Through Time, Treasure Cruise, and Dredge are non-Standard examples of this same lesson. Dredge has suffered numerous bannings to control its power because it is fundamentally violating the rules of Magic. The Delve spells are no different.


But the problem of cheating on mana is not only omnipresent in the list of banned cards, it's fundamental to today's Standard metagame. Fires of Invention provides ten mana on Turn 5, and then scales up. Wilderness Reclamation does the same thing. Nissa, Who Shakes the World can generate absurd amounts of mana with a single untap. Growth Spiral and Uro speed up your mana production without even costing a card. I'm hard-pressed to find a Standard format where cheating on mana was so aggressively pushed.


Blue decks in today's Standard format shrink the early and mid-game dramatically, instead making the majority of play time the late game. Why? This is not fun; it's a bug, not a feature. Would anyone want to purchase a four hundred-page novel and begin reading on page 150? Magic's broad structure of early game/mid-game/late game is one of its greatest strengths, and yet R&D has recently been treating it like a weakness. Players begin in the midgame and then skip straight to stealing six permanents and drawing nine cards a turn.

If Wizards wants to increase the average power level of Standard, they can go right ahead. But where they focus that power is vital to the game's health. It cannot be concentrated in cheating on mana while also staying healthy. 

Michael Majors, a professional player who joined Wizards' Play Design team, recently left and did an interview with the Arena Decklists podcast. In that interview (around 18:00), Gerry Thompson asks him about Standard's mana issues.

“A lot of the best decks in Standard cheat on mana. Is that a bug or a feature?”

“I think that it is a feature as long as it is not true that that is what defines the format, and I know one could argue that it is a defining part of the format now between Fires of Invention, Wilderness Reclamation, just generic ramp being super strong. I think the easy answer, as with most things, is that moderation is key. Maybe we are a little too heavy on those elements being some of the strongest things in the format.”

Majors goes on to share that, as one would expect, R&D is forced to experiment, gather data points, and then adjust future sets based on the outcome of the experiments. But where I disagree with Majors is that breaking the mana system is not a feature and should never be a feature. The problem is not with R&D running experiments to test boundaries and innovate; the problem is that they run experiments we already have the results of. Based on the banlist from the last three years (not to mention Affinity, Delve, and Dredge), we know cheating on mana in a format that can support the cheating requirements doesn't work out, and yet this Standard pushes it farther than it has ever been pushed before. When the cheating requirement is just “tap four lands and play instants” or “tap four lands and don't play instants,” well… those aren't exactly prohibitive in Standard.

Rosewater has spoken in the past of the truth that players don't always know what they want. Players might think they want to cast big flashy threats ahead of schedule, but they don't. It's always great to listen to what players want, but if satisfying that desire conflicts with a cornerstone of the game, then it needs to be ignored.

2) Let Players Interact

Aaron Forsythe spoke to the dangers of “goldfishing,” specifically in the context of Trinisphere preventing players from casting spells. Wizards is so cognizant of this that they avoid printing pushed counterspells and hand disruption, especially in products that are targeted at newer players. This is a good thing. For a while, Wizards did a really good job with this. Outside of Hexproof, the worst mechanic in the game for primarily this reason (when it's good, it's unfun; when it's bad, it's irrelevant), they stopped the printing of the aforementioned spells alongside cheap land destruction, Ensnaring Bridge effects, Blood Moon effects, and other things that either explicitly or implicitly state: “opponents can't.


And then War of the Spark happened and, once again, it felt like Wizards threw away all the lessons they'd learned.

If it were possible to quantify how much “fun” a card adds to the game, with either positive or negative values, I strongly believe that Teferi, Time Raveler has one of the largest negative values of any card in recent memory. There are other examples of cards like this, which I will get to, but Teferi is truly the poster child.

As I discussed previously, a good game of Magic shares a lot of elements with a good story, one of the most important being escalation. Another vital piece of this is the ability of all players to interact. Imagine a fight scene in a movie that was turn-based. One fighter pummels the other one. The second, if they can stand after, then takes a turn pummeling the first. This is unnatural, boring, and uninteresting. Fights, and stories, are made interesting because of counterplay. On a small scale, a punch can be deflected, or caught and redirected. On a larger scale, heroes and villains strive toward conflicting goals, forcing confrontation. In Magic, whenever a player takes an action, they need to be cognizant of the responses their opponents could have. Similarly, when an opponent takes an action, the other player has to weigh their response. Respond now, on the stack? Respond later, on the battlefield? Move to my turn and try to win combat?

Teferi robs games of interaction, but he's not alone. Narset, Parter of Veils shuts off decks that need to draw extra cards to play the game. Printing Narset as a follow-up to Arclight Phoenix, an archetype that was fun, fair, and interesting, is a confusing decision at best. Dovin's Veto and Veil of Summer are two more cards which, though obviously very different in terms of power level, add negative fun value to the game because they prohibit interaction. Karn, the Great Creator and Tamiyo, Collector of Tales are no different; they're simply less egregious because the axis on which they interact is a less common one. This is yet another area that power should not be concentrated in.


Contrast, for example, a card like Teferi with a card like Eidolon of Rhetoric. Both prevent players from doing something they may want to do. Eidolon is a three-mana 1/4 that, because of its lack of stats, is essentially an enchantment that can be killed by creature removal. Teferi's static comes on one of the most difficult card types to answer, especially early in the game, two other relevant abilities, both of which make it even more difficult to answer, and a free card. This is not what players wanted when they asked for a higher power level in Standard.

One of the places where this change in gameplay was most apparent was in the UW Control mirror in Modern, immediately before and after the printing of Teferi, Time Raveler. Before, the UW Control mirror had all the elements of desirable gameplay. There was an early game, marked by cheap card selection focused around hitting land drops. There was a mid-game where, if one player missed land drops, action was forced. Decisions were skillfully made around when spells should be cast, which spells were worth countering and which were elaborate baits. There was a late game involving chains of counterspells, attacking and protecting threats, and desperately trying to accumulate card advantage. While the axis of interaction (the stack) was one that not all players might enjoy (and those players are probably not playing UW Control), the rhythms of the gameplay hit all the boxes. A slow escalation of spells as players wrestled for control. An exciting build-up. A flurry of powerful spells all cast at the same time. Comebacks that could be orchestrated because of a sloppily-timed counterspell or even the mistimed cracking of a fetchland.

Post-Teferi, UW Control mirrors bore zero resemblance to the ideal gameplay that its predecessors had shown. The game became all about Teferi. If a player stuck Teferi, there was very little the other could do, in an absolute sense: they couldn't deploy threats and they couldn't answer threats. Narset shut off another axis of the game, by limiting players' abilities to stockpile cards to build to a big, exciting lategame. While Modern UW Control mirrors are an extreme, this is the kind of impact that these cards have had across all of Magic.

Emrakul, the Promised End (indirectly), Oko, Thief of Crowns and Reflector Mage fit into this category. Emrakul rendered all decisions and interaction points previously set up in the game moot. Oko shut off or stole anything relevant the opponent could do. Reflector Mage built an enormous amount of tempo by combining Man O' War with a temporary Nevermore. Energy was another indirect culprit, as the total inability to interact with the resource exacerbated its already-present flaw of circumventing the mana system.

Magic is a game, a story, a dance that requires two players. Any card which comes down and says that one of those players cannot play should give R&D serious second thoughts.

3) Variance is Vital

Only a few months before Wizards of the Coast printed Companions, Mark Rosewater wrote a two-part series on why variance is so important to Magic. He breaks it down in far more detail than I have space to here, constructing a two-dimensional coordinate system and then generating a list of ten lessons on variance. I will largely be focusing on the idea that gameplay is more interesting when games play out differently.

Suppose you read the Harry Potter series and enjoy it. You then read another book and are surprised to find that it also stars Harry Potter as the protagonist. In fact… every book you can find has Harry Potter as the protagonist. The stories themselves may be interesting. There may be a great deal of innovations and twists and surprises, but at some point, you become rather sick of Harry Potter, and the fact that every story is going to operate in the framework implicitly set just by starring Harry Potter. For example, you will never have a protagonist who is not a straight, white male. You will never get to examine the story of a non-magical individual. You will never get to experience any story that falls outside the invisible bounds created by setting Harry Potter as your main character.

This is what happened in the Jace/Stoneforge metagame, and the same thing is happening today.

Until action is taken by Wizards of the Coast, a nontrivial number (that's generous; almost assuredly a majority) of competitive games played are going to involve Companions, primarily Lurrus and Yorion. The games may be interesting and the gameplay may involve all the twists and turns and even the escalation that is so vital to Magic, and yet… at some point, you get sick of knowing that your opponent is going to cast a Lurrus or a Yorion and engage in the patterns of play each entails. It means that any permanent in the game that costs more than two mana or doesn't have an enter-the-battlefield trigger is going to be at an eternal disadvantage. 


The problem with Companions is the same problem that Splinter Twin had; for as long as Splinter Twin existed, you were, in practice, prohibited from playing any Blue deck without it in a competitive setting. This meant that any number of strategies could not be played, and all Blue decks in Modern had homogenized into Twin decks. We are already seeing the same thing in Constructed Magic today, across formats in a way we've never seen before. This is problematic because it is unfun; players do not enjoy being told that their deck-building must begin by selecting a Companion. Smuggler's Copter was banned because, in Standard, every deck was 56 cards plus four Smuggler's Copter. Homogeneity is not enjoyable in Magic, particularly for a game that so often flaunts that one of its strengths is customizability, its ability to allow players to play whatever kind of strategy they want.

4) The End Game Should Not Be the Longest Part of the Game

To again pull from Mark Rosewater, games should leave players wanting more. This means that the length of a game is important. If a game ends too early or too late, players are not going to be clamoring to play again; there is a sweet spot that should be hit. So now I'm going to focus on how games of Magic end.

Historically, some of the decks people have least enjoyed playing against include decks like Elixir of Immortality UW Control or Teferi, Hero of Dominaria UW Control. This is because, when one player has effectively won (> 99%), there are still many, many turns to go if the other player does not concede. Contrast this with UR Torrential Gearhulk control; when UR Control had stabilized, they killed in only a couple of attacks. 


The problem of how to end games is not, however, unique to control decks. Consider the current Jeskai Lukka decks in Standard, which win off the back of repeatedly triggering Agent of Treachery. In the same way that mirrors involving Emrakul, the Promised End devolved into each player repeatedly taking the other player's turn, Lukka mirrors involve stealing multiple permanents each turn. Once Emrakul is cast or Agent is on-board, the end-game has been established. Players have their resources and are doing the biggest, flashiest, most powerful thing their decks are supposed to do.

And then they play another ten turns.

This does not leave most players wanting more, as it undercuts the escalation element of a game of Magic. While skipping the early game and cheating on mana, as Emrakul and Agent do, exacerbate the problem, the end game in these scenarios becomes entirely focused around one game action, and whichever player is able to do that one game action more times or more effectively is going to win, though both players will do that game action many, many times over the course of many turns. Imagine watching a movie where the scenes in which the heroes win is an hour and a half long, with only thirty minutes of buildup. Nobody would want to watch a movie like that ever again. And yet the climax of this and other recent Standard formats revolves around one game action and lasts for the majority of the game.  It breaks the narrative flow and natural escalation and does not leave players wanting more. Instead of Agent of Treachery and Emrakul, which drag out the end game, Wizards should print more cards like End-Raze Forerunners and Torrential Gearhulk, which actually end games.


If I were in charge of Wizards R&D, I would try to do the following.

1) Fix current mistakes.

a) Ban Companions. This may seem extreme, but many of them will probably     have to go eventually anyways, and they make for less diversity of gameplay.

b) Ban Teferi, Time Raveler, Wilderness Reclamation, and Fires of Invention in Standard, unless it is decided that it's higher EV to just let this Standard format be pretty disappointing until rotation.

c) Ban Teferi, Time Raveler (and maybe Narset, Parter of Veils) in all formats.

2) Stop making mistakes you've already made.

a) Do not print cards that cheat on mana (free color fixing *cough* Arcum's Astrolabe *cough* is included here). An n-drop should usually be cast no earlier than Turn n, and resources should not be generated without an investment of mana. No one-mana dorks should be printed in Standard, as they exacerbate the already large play/draw differential and heavily restrict design of three-mana cards.

b) Be extremely careful about printing cards that say “players/opponents can't” or otherwise restrict opponents' game actions.

b2) As a follow-up to (b), include powerful, catch-all answers to problematic permanents, especially planeswalkers and colorless cards.

c) Instead of endgame threats that encourage you to do [game action] over and over again (Nexus of Fate, Agent of Treachery, Elixir of Immortality), print end-game threats that actually end the game if unanswered.

d) Do not print cards, or classes of cards, other than basic lands that players must include in their decks lest they operate at a substantial disadvantage (Companions or pushed, cheap colorless cards).

e) Internal timelines need to be reworked to allow developers time to test cards after they've been iterated on. Or put new eyes on the cards after multiple iterations. Too many mistakes come from cards being missed because the prior versions were weak.

f) No two-mana planeswalkers in Standard, and three-mana ones better be really well-tested. If there is any uncertainty, don't print it. The card type is simply too difficult to interact with early in the game before a board presence has been established.

g) If you choose to raise the power level while also ignoring (a) through (f), then at a minimum, do not put that power into places that undercut the key axioms of the game ((a) through (f)). 

Today, Wizards looks back at Planar Chaos as a huge mistake. It violated one of the fundamental axioms of the game: the color pie. At the time, it was exciting and surprising because it violated that axiom. In the short term, it was cool, but in the long term, it endangered the very fabric of the game.

Flashiness and excitement are great, but they cannot come from breaking the fundamentals of the game: primarily the mana system and the color pie, but also the ability to interact and the balance of variance with skill and agency. I'm optimistic that the hard-working people at Wizards of the Coast will learn from all of this and turn things around.

Ryan Normandin is a grinder from Boston who has lost at the Pro Tour, in GP & SCG Top 8's, and to 7-year-olds at FNM. Despite being described as "not funny" by his best friend and "the worst Magic player ever" by Twitch chat, he cheerfully decided to blend his lack of talents together to write funny articles about Magic.