3 Critical Mistakes We Made in our Pro Tour Kaladesh Testing

Pro Tour Kaladesh was our team’s worst collective performance since we got together as a team—we led most of the “average per member” metrics because we put everybody into Day 2, but we were only 5th in win percentage with a 55% win rate, which is below average for almost every member of the team. I believe we all slacked off a little bit in testing this time for one reason or another (League of Legends Worlds, I’m looking at you), which accounts for some of that, but the greatest problem we had was that our deck was just incredibly poorly positioned. In today’s article, I’ll focus on identifying the key mistakes our team made during deck selection so that we can try not to make them again in the future. I’ll also talk a bit about our deck and what I’d do to it moving forward.

Here’s the deck we played:

Jund Delirium

Our main deck was mostly identical, and the sideboard was a little different from member to member, but the differences were minimal (To the Slaughter over Ruinous Path, for example).

This deck is very similar to what we played in the last PT, except we lost Tracker, Nissa, and Languish. Instead of having those value cards that are supposed to gain you an edge in the mid-game, we had Grapples and Vessels to hit delirium and to more consistently cast Emrakul, the Promised End. Our deck last PT was a midrange deck that happened to have an Emrakul, the Promised End if it got to the super late game—our deck this time was an Emrakul deck that was just buying time to play it. Instead of an afterthought, Emrakul, the Promised End was the centerpiece of our deck.

To replace Languish, we added 2 copies of Radiant Flames, which we splashed for. The splash was mostly free, since you want to play Pilgrim’s Eye and Traverse anyway, but it did cause us to play a 24th land when we would otherwise only play 23. Some people think Radiant Flames is bad in a world of Smuggler’s Copters, and while it’s true that it does match up poorly against the card, most Smuggler’s Copter decks have to play a lot of small creatures to crew it, and Radiant Flames is excellent against those.

It’s not as good as Languish, especially against Selfless Spirit, but we thought it was important to have a sweeper because cards like Kalitas, Traitor of Ghet and Ishkanah, Grafwidow force them to overextend, and without sweepers you can’t punish that. If you have Radiant Flames, they never know whether to sandbag threats and lose to your midgame creatures, or whether to go all-in and lose to Radiant Flames. It’s also quite a good combo with Kalitas, Traitor of Ghet.

The rest of the deck is pretty straightforward—it’s just a delirium deck that’s mostly control but has the potential to have early aggression with Grim Flayer.

Our deck was good against aggro decks or graveyard decks, and it was also good against dedicated control decks. It was bad against R/G Energy and very bad against Marvel. It was also quite bad in the mirror in game 1 because we were splashing for a card that did nothing and we had Kalitas, Traitor of Ghet instead of Mindwrack Demon (not that either of those are great, but Kalitas, Traitor of Ghet just dies to Grasp and Demon helps with cheaper Emrakul, the Promised Ends and does a better job of killing Liliana, the Last Hopes), but we had more sideboard cards than most people for that matchup. As you may or may not know, Marvel was the most popular deck by a lot, and the mirror was the second-most popular, which did not turn out well for us.

As a group, we actually did well in Constructed on Day 1—we had the second highest win percentage at around 60%, including multiple members beating Marvel decks, which led to the idea that our deck was perhaps not that bad. On Day 2, reality knocked on our doorstep and we all lost to Marvel multiple times, which normalized our win percentage to around 55%.

So, how did we end up playing such a bad deck? How did we misjudge the metagame so much?

I think we fell prey to 3 basic mistakes:

Mistake #1: We overestimated the impact of the SCG Open and MTGO results.

Our biggest mistake in this tournament was that we overestimated the impact the aggressive decks would have on the format. Most SCG decks were aggressive, and most 5-0 MTGO decks were as well. More importantly, our own testing showed that those aggressive decks were very good—W/R and R/B in particular were very fast, powerful, proactive strategies, and the aggressive Prized Amalgam decks were both fast and resilient. We assumed most people would reach a similar conclusion, and either play those strategies or play something that would beat those strategies, such as control decks that relied heavily on cheap removal.

For the aggro decks, we had the Grasp/Flames/Kalitas, Traitor of Ghet/Ishkanah, Grafwidow package, and believed that was sufficient to give us a good edge. For the control decks with a lot of cheap removal, we had a large number of resilient threats, plus Emrakul, the Promised End. Most of the removal now is toughness-based, and some of the control decks don’t even run sweepers, so cards like Ishkanah, Grafwidow, Kalitas, Traitor of Ghet, and Grim Flayer are not actually easy to deal with, and often require multiple removal spells each. Pair that with the recursion from Liliana, the Last Hope and Grapple, and the late-game potential of Emrakul, the Promised End, and we thought we had a very reasonable matchup. On top of that, our sideboard was very good versus control decks in general, allowing us to take out all our removal for disruption and threats that were hard to answer, such as Tracker and Nissa, Vital Force.

In practice, aggro decks weren’t nearly as popular in this tournament, and Marvel decks were a lot more popular than we expected.

Mistake #2: We failed to take into account the “broken deck” factor.

We tested the Marvel deck, but thought it was too inconsistent. It was certainly powerful, but you had a high fail rate—you could either fail to find Marvel, or fail to get to 6 energy, or fail to find something when you activated it. On top of that, you could actually find an Eldrazi but fail to win—some of those aggro decks can go wide enough to kill you even through an Ulamog. That was a lot of failing that didn’t even include your opponent doing anything. The deck had major trouble with Negate and Ceremonious Rejection, and couldn’t beat a Spell Queller to save its life. Our natural conclusion was that people would not play the Marvel decks because they would also identify all those problems, and therefore we wouldn’t need to worry about it. We predicted Marvel decks to be roughly 5% of the metagame.

The problem with that assessment was that we failed to take into account the “broken deck” factor. Most of the time, when you’re playing a PT, you do not have a huge edge in any format—most players are good and will have tested a lot. The one spot you can get a big edge is when you attack the format from an angle that people did not expect or when you’re playing something that is just more powerful than the opposition, such as Caw Blade and Eldrazi. When you get a deck that is broken and that people are not ready for, then it feels like you’re playing a different format than everyone else, and winning becomes much easier.

Having a broken PT deck is every PT player’s dream because if it happens, your win percentage skyrockets. Marvel is a deck that looks like it’s broken. It’s doing something that no one else is doing, and it’s doing it faster than everyone else. Is it inconsistent? Sure, but that doesn’t matter because it could be broken. If you’re wrong, then you’re wrong and you go 8-8—if you’re right, your entire team Top 8s. Tournament Magic is heavily skewed toward spiking one tournament because the rewards are so much greater when you do, so playing an inconsistent deck that could be broken is very appealing because you need to be right way less than half the time for it to be worth it.

This train of thought (or what I can only imagine was this train of thought) led to many teams playing the Marvel deck. The field was prepared for it (many people were splashing Ceremonious Rejection in their aggro decks, for example), so the deck didn’t actually do well, but we were not, and we got crushed by it over and over and over again.

In the end, even though Marvel is an inconsistent deck, none of their problems are actually problems if they are playing versus B/G Delirium. We don’t play Spell Queller or Rejection, and we have no pressure to punish those decks if they fail to execute their game plan in a timely manner. They do not need a turn-4 Marvel because we don’t kill them until turn 10, and they don’t mind fizzling because they have all the time in the world to activate Marvel over and over. I lost matches where my opponent’s first activation was Attune with Aether, and their second was something like Ishkanah, Grafwidow. On top of that, we can’t really beat Ulamog or Emrakul, the Promised End if they ever get either in play, and they often get to hardcast them. This results in B/G/r Delirium having a very bad matchup against Marvel decks, which were almost 25% of the tournament.

Mistake #3: Our testing was skewed because we always knew what we were playing against.

In testing, we always know what we were playing against, not only in terms of decks but also in terms of what was in each. Sometimes this can be replicated at a PT because there’s no diversity in builds. For example, last PT, if someone played Lumbering Falls into Plains and Sylvan Advocate, I’d probably be able to tell you their deck list to within a couple of cards because the format was already well established. And sometimes it cannot be replicated, because people are playing wildly different things. When you don’t know exactly what your opponent is playing, then it becomes very hard to play a reactive deck. We didn’t know what our opponents were playing, and we were playing a reactive deck.

There was a match, for example, in which my opponent played Smuggler’s Copter and Scrapheap Scrounger early on, along with a couple of dual lands. At the end of his turn I activated Vessel and, given the choice between Liliana, the Last Hope and Grim Flayer, chose Grim Flayer. He then proceeded to play an Electrostatic Pummeler and kill me with it the turn after. If I knew he had that card in his deck, then I would have taken Liliana, the Last Hope from the Vessel, but how could I know?

Similar things happened in other matchups as well. My B/W opponent, for example, played a game-1 Declaration in Stone on my Ishkanah, Grafwidow tokens, and then killed my attacker with Gideon’s Reproach. My delirium opponent had a second Emrakul, the Promised End that I did not foresee. None of those things happened in testing.

Once it was time to sideboard, things got even more complicated. I played against many different Marvel decks and all with wildly different plans. My R/G opponent, for example, had both Chandra and Nissa. My Bant Marvel opponent had Trackers, Spell Queller, and Tamiyo. One of my R/U/G opponents had his own counterspells, but no extra threats. Against one player, I managed to Pick the Brain his Marvels away, only for him to kill me with Elder Deep-Fiend and Drowner of Hope.

This puts you in a precarious position because it’s very hard for you to react to everything. If they have Quellers and Trackers, you want Grasp, if they have Elder Deep-Fiend, you want Murder, and if they have planeswalkers, you want Ruinous Path. If they have none of those, then you want no removal. The problem is that if you’re wrong, you often lose the game on the spot, but if you are right, you do not automatically win.

In the end, we should have realized that our win percentage was somewhat skewed by playing a reactive deck with perfect information, and we should have reacted accordingly.

Moving Forward

If you were to ask me whether B/G/r Delirium was a good deck for the tournament, I’d say no, it really wasn’t. Everyone was doing something broken, whether it be playing Emrakul, the Promised End or attacking with a 40/40 Pummeler on turn 4, or playing 3 Metalwork Colossus on turn 5, our payoff, if our deck worked, was 3 Ishkanah, Grafwidow tokens and a cheap Emrakul, the Promised End.

This doesn’t necessarily mean that the deck is bad, however. All it means is that you have to wait for a metagame where getting 3 Ishkanah, Grafwidow tokens and a cheap Emrakul, the Promised End is actually enough payoff. At the PT it was not, but in a field of aggro and control, it usually is, and this is when our deck can be good.

If you look at the Top 8 from the PT, our deck would actually be well positioned there, with matchups that ranged from even to favorable across the board with the exception of the 1 Marvel deck, which we could never beat. Given how badly the Marvel decks did and how vulnerable they are to any blue card (and particularly to Spell Queller), I’d expect them to not be nearly as popular in the future as they were at the PT, which could make our deck a good choice.

The recent GP results seem to reinforce this belief. Marvel has basically disappeared, and U/W Flash and Vehicles are more common, which has led to a spike in popularity of B/G Delirium as well. U/W Flash wasn’t on our radar for the PT, but it’s another deck that has problems with Ishkanah, Grafwidow, so it’s certainly a deck you can beat.

We talked a lot about Radiant Flames post-PT, and the consensus was that it was bad for everybody. Personally, I sided Radiant Flames out in 9 out of my 10 matches, often alongside the Mountain. That’s really bad. I was splashing for a card that I only wanted in my deck for one round. As a result, we mostly agreed that the splash was not worth it, and would move to straight B/G in the future.

If you cut Radiant Flames, there are a couple things you can do with your deck. Namely, cutting a land and a Pilgrim’s Eye. Pilgrim’s Eye is a horrible card if you don’t have a splash color, but I like having an artifact in my deck for Emrakul, the Promised End, and I think it’s the best cheap one you can play. The second Pilgrim’s Eye becomes a Noxious Gearhulk, which is another artifact you love to mill and adds some versatility to your Traverse targets. I always thought 24 lands was excessive in a deck with 4 Traverse and a number of Grapples anyway, so I think you can safely go down to 23. Without the second Pilgrim’s Eye, I think it might be right to move Mindbender to the sideboard, though you can add it back if you expect a field of control and B/G Delirium.

I also think you want some sorceries because you’re cutting Radiant Flames. Two possibilities come to mind: Ruinous Path and Transgress the Mind. I don’t like either in this format very much, but I’m not opposed to them in small numbers, particularly because Gideon, Ally of Zendikar is very popular now. I’d like to try 1 Ruinous Path and perhaps 1 Transgress, but 2 Ruinous Paths is possible.

As for creatures, I rather like Tracker. It was the best card in my sideboard, and I could see maindecking a couple. You have more room for 3s now that you cut a Pilgrim’s Eye and the Flames, and you can also cut a Kalitas, Traitor of Ghet because Zombies is not as popular as we thought it was going to be (but I’d still have Kalitas, Traitor of Ghet in the board for R/B Aggro and graveyard decks). Another possibility is Sylvan Advocate, which is a decent blocker against a field of Thraben Inspectors, Reflector Mages, and Inventor’s Apprentices. I think cutting Grim Flayer is the worst decision you can make, but playing a couple of Sylvan Advocates in addition to 4 Grim Flayers is reasonable.

If I were to play this deck in a tournament tomorrow, I’d play something like this:

G/B Delirium



Scroll to Top