Sullivan’s Satchel: Monday’s B&R Announcement, Arcum’s Astrolabe, And Defining Fun

The recent Banned List update filled Sullivan’s Satchel with plenty of questions! Get his thoughts on bannings, fun as a metric, and more.

Arcum’s Astrolabe, illustrated by Igor Kieryluk

Hello, and welcome back to this week’s installment of Sullivan’s Satchel. A multi-format Banned List update, with both bannings and unbannings, sort of crowds out any other conversation about competitive Magic in the short term, so let’s dive in. As always, you can send in your questions to [email protected] or DM me on Twitter (@basicmountain), to maybe have your question answered and even less possibly but still technically possibly have your question selected as the Question of the Week and receive $25 in SCG credit. With that.

From Nick Prince:

What “should” be the criteria for banning cards in your opinion? Is it just win rate, or are there other factors that are important?

A lot has been made of the emphasis on “win rate” in the most recent ban announcement as well as previous ones. I’ve been a critic of it at times. It’s worth unpacking why the number is significant, however.

A 55% win rate doesn’t necessarily sound problematic; winning 55% of your matches doesn’t get you far in most tournaments. But Magic has a ton of randomness and has something approximating skill-based matchmaking on Magic Arena, and 55% is a bigger edge than the house has in the vast majority of casino games. If the win rate is 55% among all players (including ones who are poor, or who are average), it means your skilled players with the deck are winning a ton.

There are a lot of people who either don’t own the cards for the best deck or don’t want to play it for other reasons, and so the population of the deck swirling around your ranked games undersells how ubiquitous it would be if people were playing for stakes. I’ve worked on digital games where our best deck was something like “56% win rate, played by 15% of people at our top Elo band” — that maybe sounds fine, but then we’d have a tournament for real stakes featuring our best players, and it would be something like half the field, played by all the very best players, and six out of the Top 8 decks. In short, “55% win rate” is a lot worse than it probably reads to most people.

When a deck is winning that much, it’s likely to make the game more boring (more boring to play, to think about, to read about, etc.), and boredom is probably the most dangerous reaction a game can engender. So an emphasis on discussing “win rate” isn’t a myopic focus on balance so much as a crude shorthand for balance plus a number of ancillary considerations.

I do think the announcements could stand to speak more to experience rather than raw numbers. I know it shapes the conversation internally along with other factors, but I think it is a tough thing to translate into a public-facing announcement. People disagree about what’s fun, and the Banned List is likely to have some inconsistencies with an ideology expressed in a specific announcement versus what’s actually on there long-term. Numbers are harder to argue with. It’s a tough problem; I don’t know how to optimize for all considerations.

From @bmemike:

Despite the comment that there’s not a statistical problem with Astrolabe in Legacy, do you think it should have been banned and what are you overall feelings on format health in general?

Arcum’s Astrolabe is the type of card that’s likely to fly under the radar in terms of raw win rate because it is so easy to incorporate into decks that are bad, as well as good (I would be surprised if Brainstorm’s win rate ever fluctuated beyond 49-51%, as an example). So, I wouldn’t put much stock into claims of a “statistical problem,” in either direction.

It is more of the same; the format is already defined by cheap cantrips and even if Arcum’s Astrolabe isn’t as good as the others it is still good enough to play. Once you’re there, Ice-Fang Coatl becomes part of the equation (not quite another cheap cantrip, but not the most different thing, either). From there you kind of fill out with the best stuff on rate, and that’s your deck. I don’t think the gameplay is the worst but it’s not the most interesting thing, either. That’s pretty high praise by the standards of Legacy.

I do think Arcum’s Astrolabe violates some key architecture of Legacy — play with a bunch of duals to cast your spells and risk Wasteland, or play one color, or play two with a bunch of basics and maybe not cast your spells sometimes, or some variation of this incentive structure. Astrolabe, at its most powerful, lets you cast everything while avoiding Wasteland entirely. I understand longtime fans of the format finding this obnoxious.

I play a Legacy league or two a week. The format seems fine, about as good as it ever does, but it’s hard to read too much into playing a few Leagues. I don’t know if things would hold up if the format was under more scrutiny, but I think Legacy is at its most fun when it doesn’t have a spotlight on it, anyway. Then again I’m putting Griselbrand onto the battlefield on Turn 1 so maybe I’m not the best arbiter of such things.

From Drew Levin:

What does “good” look like in a format? A lot of the razors laid out (diversity, win rate of best deck, etc) don’t necessarily correlate with “fun”, so what’s the goal here?

Some of this is a function of the format itself. Standard is supposed to be relatively cheap, accessible, and easy to grok. Vintage essentially has no rules, and no one really cares if the format is only enjoyable for a handful of players; the cards create too high a barrier to entry regardless. Once you tease out the extreme edges, you can sense a sliding scale of sorts, where as a format gets older, has a larger card pool, gets more expensive, etc. there becomes a higher tolerance for play patterns that might not be “acceptable” in the smaller formats. Part of that is a function of those older formats appealing to more established players; part of that is the reality of balancing a format with several times more cards.

I do think diversity and win rate speak to fun, maybe not directly, but with how fun the format is to engage with over a long stretch of time. There are a lot of decks that aren’t particularly fun in the abstract that can be novel every so often — I think playing against Belcher one match out a thousand in Legacy increases the net-fun over playing against it zero times, but there is some inflection point where the fun starts to decrease.

The elephant in the room is that bannings have a massive opportunity cost (or should, at least) and that some amount of unfun needs to be tolerated so people can play with their cards. I don’t think anyone would suggest that any particular format is perfect but a lot of things exist in between the lines of “could stand to be better” and “would do something to change it.”

Lastly, the Question of the Week, from Marcus Davanzati:

In keeping with the theme of the ban list update, or well, ban lists in general and to get some insight into your knowledge of game design

When it comes to creating and maintaining a banlist, what do you think is the most important factor (if there is one) when cultivating that list?  Is it win rate? Is it “fun” (as difficult as that can be to quantify)?  Is it something else entirely?

Always happy to read your articles, and hope everything is well!

I’ve changed my philosophy on this over the years. A while ago, I was critical of WotC for creating a Modern Banned List that was, in my opinion, ideologically inconsistent, too hard to pull apart the underlying logic, and too challenging to draw expectations about future bans from. Turns out no one knows anything and can’t pull it apart anyway, so you might as well optimize for fun and forward compatibility.

Forward compatibility is the biggest thing, I think. By that I mean, banning cards that otherwise will necessitate other bans. Usually this means “enablers,” which is itself sort of a nebulous term. As an example, Mishra’s Workshop’s legality in Vintage requires many cards to be restricted around it. Because Vintage has its own incentive structure, and because Mishra’s Workshop is worth so much money, and because even if the gameplay of Vintage was marginally improved with Mishra’s Workshop being restricted it wouldn’t usher in a new wave of players for various reasons, there’s little incentive to do anything about it. But Mishra’s Workshop is an outlier case, and if you have to choose between letting Mox Opal be legal or banning one card a year that works well with Opal but keeping Opal legal, I think the former path is usually more fruitful.

Concepts of “forward compatibility” don’t apply the same way to Standard because the cards rotate so quickly; it naturally corrects over a long enough timeline. But, the format is less likely to digest outliers and fold them into a larger ecosystem because there are so few total cards. I guess this is where “fun” comes in, though it’s largely informed by win rate and other, less ideologically informed perspectives.

So, I guess my answer is “it depends on the goals of the format.” Standard needs to be fun. Ideally, older formats would need fewer and less frequent bans, with the understanding that some number of bans are required as the card pool expands. I think in Standard you should pull out the stuff that sucks, and in older formats remove cards that are extremely powerful and synergize with large parts of the potential or actual card pool.

SCG Advertisement