While playing in Grand Prix Yokohama, I began to feel like most people really don't understand Sealed. I thought about how very few articles are dedicated to it since for the most part it's considered a less interesting format than Draft, so people's understanding of Sealed is left at "It's like Draft but slower."
This understanding is woefully inadequate. Grand Prix Yokohama made me want to try to write a good article about Sealed Deck, but in the end, I couldn't resist writing about my experience with the tournament on my way to a third place finish. I set the issue of writing about Sealed in detail aside. Today, I'd like to return to that.
"Draft but slower" is a huge oversimplification. There is truth to it, but I don't think there has to be theoretically for all sets, and the reasons it's true are somewhat indirect. That statement will take some significant unpacking.
I think the traditional perspective is that Sealed is slower because decks are weaker, but I think that understanding is based on a time when Sealed generally used one pack fewer than it does now and bombs were less important.
How Sealed actually differs from Draft:
Each player opens twice as many packs and plays most of the bombs opened, so there are "twice as many" bombs. In reality, there aren't twice as many. Late in a draft, someone will pass an off-color bomb that will end up in another player's deck, and people will sometimes open bombs in too many colors to play them all in sealed. But on average, a sealed deck will have appreciably more bombs than a draft deck.
This is basically the same principle. People generally try to play as much of the removal they open as possible, so the fact that they're opening twice as many packs means they'll have "twice as much removal" (again, with the same stipulation as "twice as many" bombs).
More (Good) Artifacts
If we're talking about a card that will almost always be played in Draft or Sealed, like Loxodon Warhammer, a group of eight people building sealed decks will have exactly twice as many Loxodon Warhammers in their decks as a table of eight people drafting because they open twice as many packs and all the Warhammers in either group end up in decks.
This means that in general artifact destruction is better in Sealed. I don't think it applies to Gatecrash Sealed where there aren't really any artifacts worth killing that every deck wants to play; artifacts with a color identity, like the Keyrunes, shouldn't appear in decks much more than they do in drafts.
Fewer Explicit Synergies
When there's a trait on a card that a specific player will be drafting for much higher than other players, that player will be able to get far more of that card from the 24 packs opened at the table than they would from the six packs they'd open in Sealed. If you're drafting a tribal set and taking every Elf, you'll be able to draft far more Elves than you could realistically expect to open in Sealed. This only applies to things that other people value low enough that you can expect to see them regularly beyond the second pick (and the more you can value a certain card or kind of card above other people at a table, the more this effect will be apparent).
Fewer Niche Decks
This is an extension of the previous point, but I think it's worth pointing out specifically. "The Spider Spawning Deck" or "The Dampen Thought Deck" (for those of you who remember Kamigawa Limited), will be much more common and more powerful in Draft than in Sealed because people can actually draft the "build around me" cards.
What this means is that sealed decks in formats that revolve around bombs (and individual card strength in general) will tend to be stronger than draft decks in those formats and draft decks will be stronger than sealed decks in formats that really reward synergy.
What does all of that mean about the format being faster or slower? Is faster directly related to more powerful?
The biggest factors for speed are that sealed decks will have more removal, which will slow the format down, and that it's generally very streamlined, synergistic decks that end games fastest, which makes draft games faster.
Note that the greater amount of removal in Sealed is largely a historical trend rather than an actual rule. The reason it happens is that players are actively selecting for removal. In a format like Avacyn Restored, where most of the removal wasn't good enough to prioritize building around it or including it in a deck, sealed decks didn't have more than draft; there, the primary reason Sealed was slower than Draft was the inability to draft the nut W/R Human deck that could end a game extremely quickly on the power of its explicit synergies.
So, if we're in a typical Sealed format where we know to expect it to be slower, what do we do with that information?
I know there is a temptation to think, "Well, if everyone's planning to play a slower format, I'll just build a super-fast deck and punish them." I've tried to do that many times myself. I think people continue to do this because they just don't understand Sealed since people don't talk about it much.
If sealed decks were slower because they were worse and you could build a deck that looked like a strong fast draft deck, it's reasonable to expect in theory that this would just run people over. In reality, as discussed above, they're slower for other reasons, not because they're worse, and even if they were slightly weaker on average, there's a (very good) chance that they're actually stronger against the aggressive deck you're trying to build.
Removal is very good against aggressive decks. You're probably not going to get enough pressure to kill someone when they can easily trade cards with your most threatening creature. The weak creatures that you played to get the opponent low on life will end up getting trumped by larger creatures.
All of this setup is to get to one very simple conclusion:
Don't play Foundry Street Denizen in Sealed.
The time when people most often fall victim to this is when they think their pool doesn't have a good enough "normal" sealed deck and that they can't compete with bombs so they should just try to win as fast as possible. It might work sometimes, but I think it's very rarely worth doing. (When I say it might work, I don't mean as a long-term best strategy for a pool, I mean it will win some individual games when the opponent stumbles.) I think it's always best to build your deck as a sealed deck rather than a draft deck.
What does that mean?
You won't have as many internal synergies—this statement is true on average but might not be true in every sealed pool. If you open an exceptionally synergistic pool, you might think you've found the time to push that strategy. You'll probably be wrong.
In addition to not having synergies, Sealed is hostile to synergies. This goes back to the point about more removal in Sealed. It's like playing a tribal deck against Jund. It just doesn't work.
If you want to be aggressive, you can do it, but you should try as hard as possible to make sure that every card in your deck stands on its own as a powerful card. You want all of your draws to always be high impact. Every blank or invalidated card you draw is a huge liability in Sealed.
Ok, that's just some groundwork that I wanted to get out of the way. Next, I want to look at Gatecrash in a little more detail and try researching it in a way that's similar to how one might research Constructed results.
I wanted to write an article about how great Orzhov is in Sealed, but I felt like my experience was still limited enough that I wanted to be sure. My plan had been to play some more Sealed events to test my theory, but I've found myself unable to log into Magic Online. Luckily, this made me look for alternatives, and I realized that there's data available that's far more useful than the little bit that I could increase my sample size by playing more games.
Let's set this up scientifically. The hypothesis that I want to test is that Orzhov is the best guild in Gatecrash Sealed and that white is the best color, which makes Boros the next best guild. Outside of that, winning depends heavily on having bombs in other colors, and I believe there's a specific trend toward undefeated decks having Clan Defiance.
The way that I'm going to test this theory is by looking at undefeated decks from coverage of Gatecrash Sealed Grand Prix to see if they played out this way.
First up, undefeated decks from GP Yokohama:
Those were the only undefeated decks in the Grand Prix, but we also have the list of Trial winners:
Orzhov (splash blue)
Gruul (splash white)
Orzhov (splash blue)
Simic (splash black)
Gruul (splash blue)
Orzhov (splash Firemane Avenger)
Gruul (with Clan Defiance)
Adding Grand Prix Charlotte, the undefeated decks from the main event were:
Orzhov (splash red)
Simic (splash red)
Gruul (with Clan Defiance)
And the Trial winners:
Simic (splashing Clan Defiance)
Naya (with Clan Defiance)
Orzhov (splashing red)
Orzhov (splash blue)
G/W/B (no idea how this happens)
Orzhov (splashing Sunhome Guildmage)
Gruul (splashing five white cards, three of which are Smite)
Counting it all up, ignoring splashes and counting solid three-color decks separately:
By color, counting splashes as half:
So white is the best color, and both white guilds are the most popular. There's more Orzhov than Boros but more red than black because Gruul is the next most common guild and there's no Dimir. White is the clear best color, black versus red is debatable, green is worse than them, and blue is the worst, and these jumps seem pretty dramatic.
This seems to support my initial hypothesis that you should play white unless there's something really exceptional about your pool, but it is slightly less dramatic than I guessed.
For what it's worth, none of the undefeated decks included the card Foundry Street Denizen if that point needs further evidence.
I think these numbers clearly point to this being an extremely unbalanced Sealed format.
I'm not sure what that says about design, but I think it's worth commenting on. We've seen huge Sealed GPs, and Gatecrash has been an amazing set. But it's generally understood that WotC tries to make balanced Sealed formats, and from that perspective, Gatecrash is clearly a catastrophic failure. However, they're much more interested in selling packs than creating a balanced Sealed format, so in reality it's probably viewed as a success. The real question: is the set selling somehow because of the unbalanced Sealed format, would it sell better if the format was better, or does it actually just not matter?
Honestly, I think the most likely answer is that it just doesn't matter very much. I'm not sure what the takeaway from that should be for R&D.
As for Sealed, I'm not sure what else needs to be said.
You'll probably save time if you just curve out your white playables and build a deck around them, but it can't hurt to check your bombs first.
Don't be afraid to play five-mana creatures in aggressive Boros decks
Don't play Dimir. (All of the "There are only nine guilds" jokes about Dimir's secrecy have an interesting meaning in light of these results.)
I think that's about it: everything you should need to properly prepare for Grand Prix Pittsburgh and beyond.
Thanks for reading,
@samuelhblack on Twitter