Over my (admittedly short) career, one aspect of my game that has garnered a bunch of attention is that of the midnight brew. These are decks that, if I may be blunt, were designed the day or two before the event and tend to be lightly tested or not tested at all. The midnight brew is something that many players shy away from, and rightfully so. Few players have developed the skill set necessary to be successful in that department and most people prefer to audible to a known archetype in an effort to give themselves the best chance at a winning Hail Mary attempt.
Part of the reason for that preference though, is simply a lack of information about how to brew properly when pressed for time. I remember reading every single article that was published for about 2 years straight and the amount of information from rogue builders was slim and the amount of information on last minute brews essentially nonexistent.
On the one hand, this makes sense. After all, it is a bad habit to develop and assuming everything goes as planned, last minute brewing is something that should never be needed. That being said, life rarely allows your plans to go off without a hitch and try as you might, sometimes you just have to audible, it comes up. Switching to a known deck is something that does not need much coach, but brewing something from scratch is quite different.
Stick With What You Know
On the surface, this might lead you to think I am talking about a specific deck but by default, if we are brewing, there is no specific deck in mind. When I say to stick with what you know, I mean to use your strengths to your advantage. If you feel you play some type of archetype better than others, try your best to fall within those confines as you will be more comfortable and make better decisions as a result. I would never expect Patrick Sullivan to audible to a Blue deck for example. I tend to gravitate toward mid range decks, often with a Bant feel to them, because I feel I can best attack a metagame that way.
Using what you know does not mean you have to have some sort of archetype preference either. Consider the scenario that most often leads to last second audibles; you have been testing for a few weeks and time after time, you find yourself unimpressed with each deck you test. As the tournament approaches, you have yet to find something you really like. Use all of that testing to your advantage! You know the cards you liked in each of those decks, and you know what you didn't like. Take those lessons, broaden them out into principles rather than individual cards, and apply them to your brew.
I will use Orlando as a reference point, since it is recent and best applies here. Once I knew I was going to brew something the night prior to the event, I went deep into my database and started to pull out information from all of my recent testing. I knew I didn't like cards like [card]Wurmcoil Engine[/card] because of the prevalence of [card]Vapor Snag[/card]. Once I isolate that data, I need to extrapolate the principle behind why that is true. Obviously in this case, it is because you have invested 6 mana and a turn only to have it undone for a single mana and you even get pinged for a point. At that point, the pressure is on me to find a card that fits in the slot of [card]Wurmcoil Engine[/card], but does not suffer from the same downside. Obviously [card]Grave Titan[/card] was the card I settled on, but it all came from using recent testing data to my advantage. Going further than that, while testing for Worlds, I had a UB control list and was using [card]Geth's Verdict[/card] as an out to [card]Dungrove Elder[/card] and [card]Thrun, the Last Troll[/card]. Despite not playing UB control in Orlando, and the fact that no one was playing [card]Dungrove Elder[/card], I was still able to realize that [card geist of saint traft]Geist[/card] and [card]Invisible Stalker[/card] were not that different, and then use the testing data from worlds to plug the hole with a card I was impressed with in [card]Geth's Verdict[/card].
A Threat Is A Threat
Another thing to keep in mind when brewing last minute is that aggression pays off. Don't get me wrong, building a true control is entirely possible, and you might even have success with it, but presenting something proactive is going to more consistently produce better finishes. This doesn't mean you can't build a control deck, but maybe lean toward aggro-control or combo-control, as they are also able to produce threats.
What proactive strategies can do for a last minute brew are to help cover up some of the rough edges that are natural to lightly tested decks. Maybe your mana base is not perfect or you don't have enough sideboard cards for a certain match up; Proactive measures do the following that help to cover up those mistakes:
-Shorten the game so that less of the mistakes have a chance to emerge and so the opponent has less of a chance to capitalize on them when they do
-Allow for more dynamic sideboarding. Creatures are never as dead as an actual dead spell ([card]Doom Blade[/card] against Mono Black) and combo decks tend to have less cards able to be sideboarded out, allowing for more powerful sideboard choices.
-Produce easy wins to bail your deck out when it does not work properly. You can have the worst mana in the world, but a flipped [card insectile aberration]Delver[/card] on turn 2 can win plenty of games on its own for example.
Any of amount of proactive cards in your deck helps to keep you from being at the total mercy of your mistakes. Control with some [card]Tarmogoyf[/card]s, [card delver of secrets]Delvers[/card], and [card]Vendilion Clique[/card]s is probably going to be a smarter choice than the “only [card]Control Magic[/card]s” control deck for example. Again, you are just playing numbers here and trying to lean toward the strategies that are more forgiving of small deck building errors, as you are bound to have some.
Committing To the Numbers
One thing you must be willing to live with as a midnight brewer is that your numbers are not going to be perfect. You can tweak and theory craft all you want, but one thing that you cannot replace from playtesting, is narrowing in on the proper numbers for everything. Outside of very obvious choices that have proven themselves in the past, making something a 4-of can be a big risk. Chances are that you have not tested the specific card enough to know that it warrants that. Check out my list from Orlando:
[deck]1 Birds of Paradise
1 Glissa, the Traitor
2 Acidic Slime
4 Solemn Simulacrum
4 Primeval Titan
4 Grave Titan
4 Rampant Growth
4 Sphere of the Suns
2 Green Sun's Zenith
3 Doom Blade
3 Black Sun's Zenith
2 Geth's Verdict
1 Ratchet Bomb
4 Inkmoth Nexus
2 Kessig Wolf Run
1 Ghost Quarter
4 Woodland Cemetery
1 Dragonskull Summit
Compare this to something like Iyanaga's list from Worlds, which was a big block of 4-ofs essentially. I made sure to capitalize on the good cards though. That is, those cards who have been tested and proved themselves in the past, like [card]Primeval Titan[/card] or [card]Rampant Growth[/card]. Once you start getting into some of the more unique choices though, you see a drop in the numbers.
This is to prevent some card that I was high on in theory from flat out failing and leaving me a significant space of bad cards in my deck. [card]Glissa, the Traitor[/card] made for a great one-of because of [card green sun's zenith]Zenith[/card], and she ended up being insane, but imagine if she were lackluster and I had 3 or 4, things might have gone poorly.
Another trick is to include a lot of 3-ofs in the main deck with the 4th copy in the board. This allows for an under performing spell to take up less room in the main deck and therefore minimize the watering down that it is doing, but simultaneously allow good cards to be reinforced out of the board with the 4th, or functionally 4th, copy. This also allows you to silently improve match ups as you have more answers to more things main deck, even if your answers to one particular deck are not as dense.
Sideboarding Without Buddies
And because we are on the topic of numbers, one place where numbers tend to get overlooked though, is in the sideboard. This is a fault for deck building in general, but applies especially so during last minute brewing.
When you build a sideboard for a last minute brew, you need to consider that your numbers can very easily be wrong here as well. This can go in two directions though. On the one hand, you might have a card that is surpassing expectations and performing better than you could have thought and you bring it in way more often than anticipated. While you might discover you want more of that card in the future, not having 3 or 4 of it is something you can easily work around during the tournament with some improvisational boarding schemes. Now imagine the other way numbers can be off though. Imagine you have some 4-of rotting in the board that you never bring in, even for its intended match up, because of how terrible it was the first time you played it. What is going to hurt you more? Having a card exceed expectations and not having access to extras, or to have an 11 card sideboard because some card ended up being much worse in reality than on paper?
Check out my sideboard from Orlando
2 Curse of Death's Hold
2 Nihil Spellbomb
1 Liliana of the Veil
1 Garruk, Primal Hunter
1 Ratchet Bomb
1 Black Sun's Zenith
1 Go for the Throat
1 Ancient Grudge
1 Thrun, the Last Troll
1 Viridian Corrupter
1 Tree of Redemption[/deck]
Some of this transitions nicely from the point above this, with the 4th copy of a bunch of stuff in the board, but note the fact that there is no card that occupies more than a 2 of slot. This way, if a card never comes in, or is bad, as [card]Tree of Redemption[/card] and [card]Batterskull[/card] both proved to be, you still have a lot of sideboard to work with. I never boarded in the second [card]Batterskull[/card] or the [card tree of redemption]Tree[/card], mostly because I did not play against Red, but I still had a fully functional board for all of the major match ups.
The key here is to find cards with a high cross-use but with distinct features that allow them to have some territory to themselves. [card]Black Sun's Zenith[/card], [card]Ratchet Bomb[/card], and [card]Curse of Death's Hold[/card] exemplify this point. Each one serves as a means to control small creature hoards, but they have unique features that make each one good in different match ups. This means that if a card like [card]Ratchet Bomb[/card] is under performing against a match up I thought it would be good in, at least I have 3 other Wrath effects to pick up the slack, and [card]Ratchet Bomb[/card] is still going to be good in different situations.
Last minute brewing is not something you seek out in order to improve at it though. The best last minute brewers have developed a confidence about the craft by being put into that situation many times prior through unfortunate circumstances. Experience during the moment when you realize you need to come up with something can never be replaced by theory. There is too much involving patience and maintaining a sense of calm that you simply cannot teach and must be experienced. That said, you can practice working on your theory and discussing the application of various cards in various metagames. Those types of exercises will get your mind used to thinking the way you need to in order to brew under pressure.
And as always, if you are more comfortable switching to an established deck when you are forced to audible, by all means do so, as you are going to do more winning with whatever makes you comfortable. But for the brewers of the world, those who force their creations upon a tournament despite learning information a day or two before the tournament that spells bad news for their deck, for them, perhaps it is a good idea to try the midnight brew. It can be a ton of fun, promise