Reassessing the Modern Banlist – A Data-Driven Proposal

The Modern banlist contains a number of the worst mistakes in Magic history – cards like Skullclamp, Treasure Cruise and Mental Misstep. Each of these cards is banned in Legacy for its raw power level.

Skullclamp is my personal pick for the most absurd Limited card ever printed – beating Umezawa’s Jitte, Pack Rat, Sol Ring, Parallax Wave, Vedalken Shackles, Attrition and Citadel Siege. Unlike the last two cards, it’s absurd in Constructed too, often drawing 6 cards or more and totally taking over games for just one mana.

The list also contains cards that seem perfectly reasonable at first glance, but that enable combos that are too fast or too consistent to be legal in the format.

Summer Bloom, Hypergenesis, Blazing Shoal, Eye of Ugin, Golgari Grave-Troll – each of these cards is innocuous on its own, and none of them save the last were good during their Standard tenures.

But each of them enables combos that are too powerful for the format. They either lead to overwhelming board states – sometimes even outright kills – on turn 2 too often for the format to handle, or alternately they allow fairly consistent turn 3 kills. From day 1 of Modern until today, every time a combo deck has hit a 2% turn 2 win rate or a 20% turn 3 win rate, the banhammer has always come for it.

Hypergenesis was outright bad in Standard and at the time a bulk rare, but with the Cascade cards introduced in Alara Reborn, it lets you vomit out an Emrakul, the Aeons Torn on turn 2 around 11% of the time – in a format with fundamental turn 4. This becomes as high as 25% of the time if you are willing to settle for ‘lesser’ threats like Griselbrand or Ulamog, the Infinite Gyre.

The list also contains a number of cards that are deemed simply too strong for the format, cards that dominate games in which they are used and that simply cost less mana than their effect should.

Deathrite Shaman, Bloodbraid Elf, Stoneforge Mystic, Umezawa’s Jitte, Jace the Mind Sculptor, Birthing Pod, Splinter Twin, and the artifact lands – these cards all allow counterplay and sometimes are poor draws, but each of them have, at some point in the format’s history, earned the banhammer’s caress through being too powerful.

Some of these bans I agree with, others I do not. But I’m not going to make the case for my opinion on which cards should remain on the banlist and which should earn a reprieve. I could easily be wrong, and even if I am not you have no compelling reason to agree with me. Instead, I have a proposal for how to methodically test the waters as to which (if any) cards would improve the format if they were unbanned.

But first let’s look at some history.


The Golgari Grave Troll incident

This card breaks my first fundamental rule of Magic: cards cannot impact the gamestate unless they have been drawn.

An aside.

GGT was unbanned a couple of years ago. It found its way back onto the banlist after Shadows over Innistrad and Kaladesh provided the last pieces for a broken deck based around the Dredge mechanic.

If you aren’t familiar with the Modern dredge deck, it used GGT and Stinkweed Imp alongside card filtering spells (Cathartic Reunion, Faithless Looting) to fill its graveyard, and to then vomit out Bloodghasts and Prized Amalgams from the graveyard without ever drawing them, much less spending mana on them.

It was an aggro-combo hybrid, much like the Affinity decks still legal in the format – but much stronger than Affinity has ever been.

Everyone knew GGT was dangerous both before its unbanning, and in the period that it was legal. Stinkweed Imp was already legal, and allowing decks to play a larger critical mass of high dredge number cards posed a risk.

WotC went ahead and made the perfectly reasonable decision to unban the Troll at a point where Amalgam and Reunion did not exist. Then, they made the perfectly reasonable decision, when the card proved itself too good for Modern, to ban it again. (I advocated for banning Amalgam and Narcomoeba instead at the time, but the result would have been the same in either case – Modern Dredge as a deck would have been removed almost entirely from competititve play).

The experience of GGT’s unbanning and rebanning sent shockwaves through the Magic secondary market, however. Both GGT and its supporting cast became extremely expensive for a time, then cratered in price again after the rebanning. A number of competitive players lost money as a result, and this soured the waters against ‘risky’ unbannings.

The end result of the unban/reban was that we got data stating conclusively that GGT was not safe to be in Modern alongside Amalgam et al, but the cost for this data was high.

There is a better way.

Isolation Testing on MTGO

Magic Online has a thriving Modern community, with thousands of players entering competitive leagues.

For those unfamiliar with MTGO’s leagues, competitive leagues run on demand, allow players to play 5 rounds, cost $12 to enter, and award prizes as follows:

5-0: 16 Treasure Chest, 180 play points (current secondary market value $57.36)

4-1: 8 Treasure Chest, 180 play points ($37.68)

3-2: 1 Treasure Chest, 120 play points ($14.46)

2-3 or worse: No prizes.

Play points are 10c coupons used toward paying for future events (for instance 120 of them can be used to enter the Competitive League) and are unable to be traded, treasure chests are ‘lucky dip’ bags of cards from Magic’s history and at the time of writing can be sold to dealers for $2.46, although this is a price that fluctuates and is typically around $2.20. 

MTGO’s competitive leagues provide an enormous incentive to ‘break’ formats.

As an extreme example, imagine a hypothetical unbeatable deck – each time you enter a league, you play your 5 matches, then get rewarded with $45 in ‘profit’ for your time. You then enter the league again, and you can build the deck on multiple accounts and enter more than one event at once.

If your deck remained unbeatable (except when you play against your own alternate accounts), and you play fast, within a day you would have made thousands of dollars. Within a week, you’d have made more than a GP winner, and within a month, you’d likely make more than a PT winner.

This dynamic – caused by the ability to immediately enter a new tournament with prizes each time you finish one, makes MTGO exceptionally good for deck innovators. In paper, if you innovate and build a format-breaking deck you may only get to enter 4 tournaments with prizes in any given week – online, you can enter four leagues in an hour and a half.

While no deck is completely undefeatable, sometimes formats are broken and a deck exists that has a 70% or even 80% win rate against the field. The various aggressive Eldrazi strategies at Pro Tour: Oath of the Gatewatch are an example, particularly the Eldrazi Skyspawner versions.

If you discover such a deck and pilot it on MTGO, you will very quickly win a lot of prizes. You will also attract attention – MTGO publicly reports decklists of a minority of players that reach 5-0 finishes, and also publishes the screen names of players that perform well on a consistent basis. If you achieve 5 5-0 results in a single league, your screen name will attract attention and people will attempt to find what deck(s) you play. Get 15 trophies (5-0 results) and even if your list is not published, people will pay attention when you are paired against streamers and will quickly determine the deck you are playing.

This means that if a player breaks a format, MTGO will adapt to them very fast. Some will build the new deck, others will play an established deck but will attempt to answer the new deck with sideboard tech. The metagame moves far faster than in paper – which is ideal for my proposed experiment.

The Modified Banlist League Proposal

I propose that an MTGO Competitive league be set up with an alternate banlist, after the February 12th B&R announcement comes into effect. This would run at the same time as the ‘real’ Modern Competitive and Friendly leagues.

This league would run for 6 weeks, use the usual MTGO competitive league prize model, and would contain the then current Modern banlist with the following exceptions:

– Bloodbraid Elf is unbanned

– Jace, the Mind Sculptor is unbanned

– Stoneforge Mystic is unbanned

– Punishing Fire is unbanned

– Mox Opal is banned. (This will make more sense when you read the next line – I believe Opal is likely to be banned in the next 18 months in any case)

– The five ‘artifact lands’ – Ancient Den; Seat of the Synod; Vault of Whispers; Great Forge and Tree of Tales – are unbanned.

There has been a lot of testing done by individuals as to whether various permutations of these cards would be safe in Modern. But individuals aren’t as good at innovating as the Modern community in general. Individuals, no matter how talented they are as deck builders, miss ideas. It was correct to play Eldrazi Skyspawner at PT:OGW, but this card was only played by about a dozen participants – the rest never thought of it.

On top of this individuals cannot generate statistically significant amounts of data. This proposal would generate hard data, and would track how the metagame changes over time.

It would also test something that can’t be determined by a small playgroup testing alone. Standard prior to the Emrakul banning was in a strange space where the format was fairy balanced, with no tier 0 decks and more than one tier 1 deck, but the format was widely considered not to be fun. It’s entirely possible that unbanning some of these cards might lead Modern in a direction like that, and no single playgroup will have enough data to truly protect against this.

But a whole league – tens of thousands, maybe hundreds of thousands of matches played by people who have a material incentive in winning – this can and will provide very serious stress testing. It would not just be league results providing data, but also league attendance – if three different JtMS decks are widely accepted to be the only Tier 1 decks and the number of people playing the league falls off sharply by week 4, this would provide compelling evidence that JtMS is not safe to unban.

I expect that MTGO could reasonably easily be set up to put four phantom copies of each ‘experimentally unbanned’ card into each player’s account for the duration of the league, too. If this is possible it lowers the barrier for deck brewers looking to experiment with these cards.

If some of these cards turn out very quickly to be proven to be mistakes, they could be ‘banned’ midway through the league. If Seat of the Synod eats a ban on the third day of the league, players that have already enrolled in the league with Seat of the Synod in their deck would be able to complete their five matches using the card, but noone could enter the league again using it.


I believe JtMS and BBE are safe to remove from the Modern banlist, that the artifact lands might be, and that no other cards are. I could very well be wrong and am very open to eating these words.

Assuming you disagree with my thoughts on the ban list, we could argue this point for hours and, in all likeliehood, neither of us will shift. Or worse, I could be wrong but convince you of my opinion, or vice versa. In any case this arguing without serious testing is about as useful as nailing a cat to your door. (Disclaimer: This isn’t useful. Please kids, don’t nail a cat to your door).

The DCI lacks the resources to do this testing, and recent Standard environments show that WotC aren’t doing a great job of it either (RIP Felidar Guardian… how you got to print is beyond me…).

But Magic Online provides the solution. All that we need is for WotC to use it.

Additionally if this league were to be a success, I think there is room to repeat the experiment in Legacy..

12,374 total views, 3 views today

Leave a Reply

Your email address will not be published. Required fields are marked *