Last weekend, I jumped on a couple of planes and flew to Brussels in preparation for Pro Tour Dragons of Tarkir to test with my team. We have been known to be called "Team TCG Player," but I'm currently referring to us as "Team Can't Beat Luck" to refer to our most reliable method of victory: sacking our opponents out.
Flying in a week early accomplishes a few things: It gives us an opportunity to acclimate ourselves to a new time zone so we don't get wrecked by jet lag at the tournament, and it also gives us a chance to get in some quality testing without the distractions of being at home.
The first few Pro Tours I played in I wasn't really on a team. I would test with a few friends, but I didn't exactly have a real testing team and was mostly in the dark when it came to a lot of things. Interestingly enough, my best Constructed record at a Pro Tour came from one of those tournaments, Pro Tour Theros in Dublin, where some testing against CVM and discussion online with Jacob Van Lunen and Gerry Thompson led to an Abzan Midrange deck. I went 7-3, and JVL unfortunately lost a win-and-in to Mono-Blue Devotion, a deck we didn't see coming. That tournament pretty much summed up testing without a team. We did well, but it was impossible to see everything that people came with.
For the last few, however, I have been testing with real teams, and I want to share some of the insight and ideas for how to test with a team and how to test for a big event that I've picked up along the way.
Don't Get Attached
For this Pro Tour, I came into the testing process with a deck that I really liked in theory. Over time, as I was working on it, I kept making realizations about how the deck was supposed to function, and my deck kept becoming more and more tuned in the process. I was winning a lot of games and felt like I was really making headway with the deck.
It was starting to basically become a pet deck of mine. I had a lot of fun playing with the deck, I was investing a lot of time and effort into the deck, and I could see progress with the deck as I was starting to win more and more matches. However, there were a number of flaws with the deck that kept becoming more and more apparent. For a while I ignored them as variance associated with the deck or as an aspect of the deck that could be fixed, but finally I realized that there was simply a fundamental flaw with how the deck was trying to win and I ended up abandoning it.
While the deck was still solid and could probably be good enough for something like a 5-5 finish, I realized that it simply wasn't going to be the deck that I was going to play at the tournament because there wasn't anything exciting about it. At that point, I ripped off the bandaid and just threw the deck away entirely. I stopped testing it and stopped bothering to work on it.
The same theory holds true for individual card choices as well. Sometimes it can be easy to get blinded by how good a card should be and ignore how it is actually playing out in practice. Once it becomes apparent that something isn't going to cut it, there's really no value in wasting more time or effort on it. It's easy to try to change things around enough to make it work, but more often than not, it's just not worth it.
It's really easy to take a card or deck that you love and apply biases toward how you evaluate it. Your mind works to ignore the times where it is bad, and you really remember and gain an emotional attachment to the times where it was great. As a result, it is imperative to take an objective view of a card or deck and try to avoid seeing things through a tainted lens.
The easiest way to gain an objective view is to utilize data, which leads to the next part.
Keep Track of Results
When testing decks together, it's important to actually keep track of how things go. This doesn't mean you have to say exactly "X was 6-4 against Y in preboard games, and 6-8 in postboard games," but it helps to write out how the matchup played out in both preboard and postboard games.
Sometimes decks will feel like they are performing really well, but when you check the data, the opposite is true. Sometimes this feeling is based on biases, and other times it is simply based on the way games play out. Some decks win in big and splashy ways, and it influences how often it feels like the deck wins. Some decks take so long to win games where they are ahead that it just feels like that deck is winning more than it loses because you spend more overall time winning, even if you haven't actually won more games.
Keeping track of results has another hidden benefit that often gets overlooked. Sometimes you will run the gauntlet with a variety of decks to see how they do against the established decks, and when the smoke clears and you go back to look at results, you simply realize that one of the gauntlet decks is consistently beating everything you throw at it.
At that point, it might be worth starting to tune that deck and turn that deck into one of the frontrunner options to play in a tournament. Sometimes the best choice for a tournament is something boring and simple. Nothing goes wrong when you stick with a tried and true strategy.
Figure Out Powerful Cards
I've done a lot of testing with Michael Majors over the past year, and he approaches testing in a wildly different way than I do. I tend to build decks that are more conservative in nature. I don't really go too far off the deep end when I test because it's pretty rare and almost never the case where a "deep end" deck actually becomes good. To me, it often felt like a waste of time.
However, Majors said something once that was really insightful and actually completely changed the way I approach testing. You're testing to find good cards, not necessarily good decks. Maybe the random deep end brew you came up with is really bad, but you discover that some card you have in the deck ended up performing way above expectations.
You can take that information and use it to put this awesome new card into a completely different shell and see if it fits well in that shell. Sometimes, this means putting it into an entirely new deck idea, and other times it can be as simple as just updating an existing list with this new card. Either way, you've found a card that is good, even if the original deck idea isn't the right home for it.
Also, in more established metagames, it's also important to identify cards that are good. It's hard to really provide a good way to figure out the best cards in a metagame, but the way I tend to approach it is to think of extremely powerful cards and effects that aren't currently seeing any play. Then I think about those cards in the context of the current metagame and see if they happen to attack the big decks in a powerful way.
For example, there are a lot of really powerful cards that are currently very underrepresented. Cards like Whisperwood Elemental and Wingmate Roc, for example, aren't seeing a whole lot of play right now. It's possible that they are poorly positioned against the big decks, but it's also really reasonable to think that in a month or two, when the metagame shifts to combat the current heavy-hitters, that these cards might end up being giant powerhouses again.
Revisit Old Decks
Testing processes in general are cyclical things. When big companies run through processes like this, they tend to be done in a way that you run through some number of steps and then when you reach the end of the steps, you start the cycle over again and repeat the process, utilizing the new information you've gained along the way.
Magic testing can be done the same way. You can run through a bunch of testing and learn valuable information about a variety of decks, and then you can rebuild your decks based on the information you've gathered before testing the ideas and matchups again.
This is especially useful in combination with the previous section. Perhaps you ran through some testing and discovered a very powerful card. One of the best things you can do is try to slot that card into existing decks that can utilize it and run through testing again with those decks and see how the new card hurts or improves a variety of matchups.
Sometimes what will happen is that the discovery of some new card can actually cause a solid but unexciting existing gauntlet deck to actually get pushed way over the top to become a really powerful deck. An example of this is the Jeskai Tokens deck that many of us played at the Season One Invitational. Discovering how powerful Dragonlord Ojutai was in sideboarded games completely caused us to reassess the power level of Jeskai Tokens and eventually play it. Jeskai Tokens was a deck with generally great game 1s against a lot of archetypes, but one that could easily struggle game 2 against hateful removal options.
Ojutai allowed the deck to circumvent that and beat opposing sideboard plans against the deck. This, in turn, pushed Jeskai back into being a very competitive choice because it completely redefined the matchups.
Test Sideboard Games
Earlier today, Brad Nelson made a comment that was probably a bit exaggerated (something he loves to do) but that isn't too far off from being true. He essentially said that if a team did nothing but test only sideboard games, and not test any game 1s, that they would come up with a much better deck than the normal testing process.
At minimum, you will play as many sideboarded games in a tournament as you play preboarded, so it's imperative to test with sideboards and figure out how to win sideboarded games against a variety of strategies. It happens quite often that people jam ten game 1s in a matchup, see that one deck is winning seven or eight games, and then write the matchup off without fully understanding the implications of what a sideboard can do.
Sometimes a deck swings really far in sideboard games to the point where the matchup actually favors the deck that loses game 1, even if they were to lose that game 100% of the time. An example of this is something like a deck like B/W Control from last Standard season that will lose a decent amount of the time game 1 to Sphinx's Revelation control decks, but after sideboard when things like Obzedat, Ghost Council, Duress, Underworld Connections, and Sin Collector get into the deck, it becomes extremely hard for the Revelation deck to win anymore.
If all you did was test game 1s, it would be easy to get the wrong impression of a matchup, which can end up being fatal for a tournament.
Write Plans, Not Tragedies
Once you have a deck that you are mostly settled on playing, the last step is to come up with plans for each matchup you expect to face.
The key word here is plans. Not cards. Plans.
Nissa, for example, is a card that is traditionally good against control decks. Nissa, however, is not a plan against control decks. Strip their hand with Thoughtseize and Duress and then smash their face in with Nissa is a plan, but just "have Nissa in my sideboard for control" isn't exactly a plan. That kind of a plan is prone to failure because these control decks also have the plan of "don't get wrecked by Nissa in game 2," and they will be prepared for your card.
Instead, the important thing is to actually come up with a plan for how you intend on winning the game. This plan should take into consideration how they plan on sideboarding against you. Consider what their deck is going to look like against you in sideboarded games and then figure out what the right mix of cards is going to be to combat that configuration.
Sometimes this means you don't bring in Drown in Sorrow against the deck playing a bunch of 2/2 creatures because you know that after sideboard they are going to side into a deck with Dragons and Planeswalkers and your Drown in Sorrows are going to look awful. A great example of this is how I had Back to Nature in the sideboard of my Abzan Aggro deck at the Players' Championships last December.
Back to Nature was in that sideboard specifically for the Whip of Erebos matchup. However, when I played Gerard Fabiano in the top 4 of the tournament, with him on the Whip deck and access to each other's decklists, he knew that Back to Nature was my sideboard plan against him, and he ended up siding out a lot of his enchantments. As a result, it was actually correct for me to not board into Back to Nature against the deck Back to Nature was specifically in my sideboard to beat.
Back to Nature is a card. Sideboarding into Planeswalkers to beat his plan of overloading removal spells was a plan.
Coming up with plans also helps narrow down which cards you should play in your sideboard. Oftentimes people waste numerous sideboard slots because they just throw cards in their sideboard, and they either have too many cards for one matchup, or cards that serve specific purposes that will never actually come up.
When you have a plan for each matchup, you can also figure out which sideboard cards facilitate coming in for the most matchups to benefit that plan, and thus, completely cut out the chaff from the sideboard and fill it instead with some of the gassiest cards this side of the Mississippi.
All things told, figuring out how to optimally test for tournaments is something that we as Magic players and testing teams are still figuring out how to refine. I think we are a long ways off, and I also think that this is probably one area that we could improve the most at. It's very possible that we look back in five years and laugh at how disorganized and chaotic we were and how little we actually accomplished compared to what we could.
While I certainly don't have all the answers, these are at least some starting points for how to approach testing for big events and some things I've learned from the testing that I've done. I'm hoping that in a year or two from now, I will be able to write this same article again with completely fresh and new ideas that blow the water out of these ideas.
Our testing process for how to improve the testing process is still undergoing, but I can only hope we hit the finished product before I get too bad at Magic and fall off the train entirely! It's only a matter time. Let's get to work.