Sunday, September 27, 2020

What Broke Standard?

Since Throne of Eldraine, Standard has been a mess outside of, well, the Theros Beyond Death format that was a mess of messes. 


This is despite an entire new department, Play Design, being spun up at Wizards of the Coast just before that point to test Standard and optimize the format due to the last batch of disasters. What the hell went wrong? Was it mistakes in testing? Was it the world drastically changing? How did we get into this mess so people can even think of ways to get out of it.


Don't Blame the Data





Five years ago, we had data as good as today if not better. Magic Online used to let third parties watch every replay of every match played, with four-round Daily Events happening multiple times a day for each format, similar in scope to the current five-round Preliminaries or SCGTour Online Challenges. There were also larger Premier events, two-player queues firing regularly, and all of this data turn-by-turn could be logged by a bot. You could pull up a grid and have a three- or four-digit sample size telling you who was ahead in every Modern matchups. If you had the raw data set you could even correlate who won more when specific cards were cast, or cast by specific turns, or when a sideboard card was involved. If you wanted Limited content, MTG Goldfish scraped the draft queues to see what cards had the highest win rates when cast!

Even if you didn't like raw numbers because you could move them with deck building decisions, you get every 4-0 or 3-1 deck from every one of those Daily events. Sounds similar to the data set we have these days?

Standard didn't break. Modern didn't break.

We didn't have anything close to this data set last year when Field of the Dead and Oko, Thief of Crowns broke everything.

When Temur Reclamation was crushing, I kept a close eye on the mtgmeta.io data set. It kept saying things were bad matchups when I knew how to win them, and the best players kept winning with Temur Reclamation against the field.

I'm not saying people don't look at the data and it doesn't influence decisions, but maybe it's not the data breaking Standard.


Don't Blame the Tech



Twitter has existed for 10 years and it didn't break Standard. Discord is a well polished version of a 2005-era moderated forum, which honestly is one of the greater mediums in Internet history so I have no issues moving back to it. Things aren't that different from Standard's working days.

The one thing that may have changed is the monetization incentives of Twitch. It is likely someone else has a similar idea to you, and it's likely you can accelerate your ideas by watching their game play at double-speed. That's because when they have a good idea, they get viewers, and that pays real dollars (or as the Pro Tour documents for American venues loved to say, sometimes called “greenbacks”) and not just internet clout. Just kidding, I've seen some numbers for smaller streamers and it does pay in amounts that round to clout, but you know maybe you through enough clout farming you become a big timer and make medium or decent money some day.

But uhh, ask Cedric about The Stream Team one day, and that Pro Tour where he rolled up with Jund and played Geralf's Messenger instead of Deathrite Shaman. Streamer data is a starting point, but not the finish line. Twitch is also almost a decade old, even if it wasn't as populated by Magic content before Arena.

Tech changes might have an impact on metagame churn, but I don't think that tells the story at all.


Threat Design


This is really obvious, and there's no better place to point to this than Uro and Omnath.


What don't these cards do? They draw cards, provide life total buffers, provide mana, one of them even casts itself from the graveyard, and whatever they do they keep doing it. Oko, Thief of Crowns existed in a similar space of life gain, permanent value generation, threatening to blank threats, and just raw speed.


 

Back in the day you had to invest in a different threat for a different job. To use and expand on the classic Patrick Chapin role assignments, you had Baneslayer Angels that could rectify game states but died to removal, Mulldrifters that were raw card advantage at the cost of smaller immediate impact, Morphlings that were unkillable and closed the game at cost. Dark Confidants (or Ophidians or Thieving Magpies for the ancients) didn't provide defensive or offensive strength but threatened to accumulate value if given time unanswered. Depending on whether you expected more aggro, midrange, or control in the metagame the optimal choice changed.

Where is the choice these days? Uro just beats aggro, control, and midrange. So does Omnath. So did Oko.

When you don't have to make real choices, you get homogenization. You don't have to choose the scale your deck operates at, you have tools to cover everything.

That's a problem.


Answer Design


We learned the lesson in years past that removal that is too efficient, too broad make it so the cards in the format don't matter much outside of doing something and costing little enough to avoid trading poorly for the answers. My hot take is maybe we deserve Mana Leak and maybe Pioneer and Historic need it, but it certainly would have baggage. Again, the theme of decisions mattering or not mattering shows up.


Answers haven't been quite as bad as people think they were, Temur Reclamation was basically an all answers deck with some good mana and card advantage. But they got really bad after rotation. You can kill basically any creature pretty easily, but as we mentioned the good creatures don't die or cantrip and it's all horrible. Planeswalkers are pretty hard to kill, especially when they are the Theros Beyond Death or Thone of Eldraine ones printed under the War of the Spark era's absurd numbers. The only good way to kill things right now is Bonecrusher Giant, Mystical Dispute, Spikefield Hazard, as evidenced by all the four-color decks deciding none of the other answers are good enough to reliably play.


This might be temporary and answers could come back once were aren't forced to exclude black mana from our decks.

I think the biggest problem might have been what I alluded to with Temur Reclamation. We've entered a world where the best answers often are best utilized in a deck maximizing the best threats. There's an entire extra blog post about this, especially with how it related to Aether Gust being a problem and Mystical Dispute being great, but this all points back to the threats making mana and drawing cards and doing that over and over. If you keep getting cards and mana, the best thing to do is stop your opponent from doing the same thing so that your cards and mana kill them because they don't have cards and mana.

Answer design probably isn't a major issue, but it's a bit weak right now. Whatever, that happens in small Standard, cards like Carnage Tyrant have often dominated formats because there were only like two answers that existed for them before.


Closing the Ramp Gap


 

Growth Spiral was banned in Standard. Explore wasn't. Lotus Cobra when it was printed the first time was powerful, but not broken like this despite having Misty Rainforest in the format


The things you would Explore or Lotus Cobra into those days often cost six or seven, Primeval Titan or Avenger of Zendikar. These days, they cost four, Wilderness Reclamation or Omnath.

When you look at the answers, this is troublesome. A successful strategy against ramp used to be getting out under them and efficiently stopping their payoff. There isn't really an under them when their best hands deploy that threat after you have only had two turns to play. You can't leave up Disdainful Stroke on the draw after playing a one-drop and expect to win. That extra turn or two of setup gives you time to produce enough things to really punish them even if you only buy a single turn against their threat, thing of how the various Pioneer and Modern Spirits decks hit a specific point where they deal much larger amounts of damage because they set up Lords.

The raw number of cards between four and six also matters. When you need six lands to produce a Primeval Titan, that's six actual cards you need at specific points. The old ramp decks often stumbled, and stumbled more in the games where they drew interaction. Think of how Modern Valakut Titan decks operated before they got squeezed out of the metagame, or the Scapeshift quote I've said for years “Scapeshift is a one-card combo that takes eight or nine cards to win the game”.


Ok, Explore into Jace, the Mind Sculptor was pretty dumb. But Jace was uniquely vulnerable to small-ball creature setups and Bloodbraid Elf, so it kinda worked out. Bloodbraid Elf off Explore was also kinda dumb, but that card played pretty bad with Explore left in your deck and often was better a turn later when the cascade hit something reactive against their later must answer play. Not quite the same as Wilderness Reclamation which is just good and hard to answer whenever it hits.


Like, have you read Grave Titan? Aggro decks could race that card despite what it does.

Yet again the problem lies with the threats, but it's not just the text boxes but the costs.


London Mulligan.... and Open Decklists


I had some lightly tested, lightlyspeculative concerns about the London Mulligan a year and a half ago after testing for the namesake Mythic Championship in London itdebuted at. It's been bad in those ways, but somehow it may be worse in a different way that is definitely a big part of ruining Standard.

There just isn't a reason to play a linear strategy, because even pre-sideboarding your opponent can significantly adjust to preempt your plans.

Before the combination of open decklists and the London Mulligan, midrange decks had a huge issue playing against the field. You would have some Doom Blades in your deck and some Duresses, and you would start each game down a Schroedinger's Cat of a draw when you drew Doom Blades against control and when you drew your Duresses against aggro. When your opponent played their first turn, you would learn how bad your decision was or wasn't.

These days, you can peek in the box. If the cat is dead and you have Duress against aggro, you can trade it in for a new box. Unlike before, if you draw a Duress in that next hand it doesn't have to stay there and you can keep the six cards that do matter. You play way more games with a piece or two of interaction, and then I'll just point back to the part about the current threats making cards and mana and using that to stop your opponents from doing that.

I would guess the win rate of Sultai Midrange in late-Core Set 2021 Standard was multiple percentage points lower on high ladder than in events with open decklists. I won a ton of games with four lands, Uro, and Extinction Event or whatever because I knew my opponent was attempting to play creatures and those six cards would do the trick. And if they were playing control I would just keep the Hydroid Krasis instead or keep the first seven with a bunch of Thought Erasure.


The only time we have really seen aggressive decks crack through in the last while is when they exploit a static metagame with threats that completely dodge the expected removal. Think of how Mono-Black Aggro and a Winota deck leaning on Orzhov color threats smashed through the Players' Tour Finals full of Aether Gusts and light on ways to kill Rotting Regisaur or a lot of small dorks with more toughness than power.

Again some of this is the threats making the smaller starting hand sizes not matter as much, but open decklists plus the London mulligan drastically reduces the viable angles of attack and forces people down the road of mirror matches.

I think you can maybe justify one without the other, but together it's a non-starter. I don't think you can move away from open deck lists if Twitch streaming of events is going to be something you want, but maybe with MDFC spell-lands balancing mana screw and mana flood, we can change the other one.


The Arena Economy


This might be a bit of yet another full post, but one of the bigger issues with the Arena economy is that it pushes players to not try things.


True story: Imagine you are me in early 2019. You see a Reddit post about Grixis Dragons at #53 on ladder or whatever. You want to try it, so you spend four rare wild cards on Dragon's Hoard. It is trash, you learn your lesson, and you go to build your next deck.

You can't spend rare wild cards on Tithe Taker because you spent them on Dragon's Hoards, also because you aren't literally a dragon with a hoard of gold wild cards. You can't ever really use those Dragon's Hoard again because you don't plan on playing some casual nonsense. You can't get those Dragon's Hoard, probably a desireable casual card, to someone who wants them in exchange for goods, services or Tithe Takers.

Aside: It is also kinda depressing to realize the most exciting card you can open from a booster on Arena isn't any specific card, but a blank artless wild card of the right rarity. If you want any assurance of the future of paper Maigc, the concept of a chase rare that people crack packs for is dead on Arena and they keep printing those cards anyways.

The story has two happy endings: I realized my Venerated Loxodon deck was better if I put commons that cost one-mana in it instead of rares that cost two-mana, and I just played the best deck more from there on out. I'm no tangrams, but I certainly have had my fair share of success on the SCGTour Online forcing people to beat my Uros instead of not doing it and I owe it partly to Dragon's Hoard being a garbage competitive card.

And that's me, who has seen time and time again at the highest levels of the game that you learn from trying things and has been rewarded for doing so.

Imagine someone just trying to get into the game now. They get beaten by Four-Color Omnath. They feel sad they lose to it. They build it, they learn to Four-Color Omnath people, and the cycle continues. Or they don't and they stop playing.

I also literally have no idea how to get into a match on Arena in an enviroment where Four-Color Omnath isn't a likely deck my opponent is playing. You can't post “No LD No Counters” and find and opponent. You probably can't even wait for a Brawl opponent, probably not even Historic (also good joke entering Historic if you care about wild cards at the recent rate of releases), and is there another format. Maybe there's like a No Omnath Direct Challenge Discord I don't know about.

The end result is the experience of bashing your head into Four-Color Omnath and quitting or playing it being the simplest options on a lot of levels becomes universal. When Kaladesh Standard got bad I just played Modern for a year.

I don't think that things would be all sunshine and daisies if Arena didn't push Standard this way. Oko and Field of the Dead were paper issues back when we could do things like travel to the basement of the Valley Forge Casino and Resort. But if you want to look at things that changed the tone of the discussion, this is a big one.

I don't think a complete overhaul of the Arena economy is without major other considerations, a good idea, or even doable, but breaking the ways it promotes Standard Tier 0 mono-cultures is necessary to the continuing health of the game.


It isn't just one thing that broke Standard. Just like it wasn't just one card that needed to be banned. A lot of systems broke down the line. Some of them easy to fix, some a bit hard, and some you probably just have to live with. It's possible to build a functional format from the disaster we play in right now, but it will take a few steps. Bans of Uro, Omnath, Lucky Clover, Embercleave, Scute Swarm, whatever are just scraping the surface a temporary patch.