For this Pro Tour, I prepared with Martin Juza, Ivan Floch, Stanislav Cifka, Matej Zatlkaj, Robert Jurkovič, Robin Dolar, Adam Koska, Nikola Vavra, and Petr Sochurek. We tested for a week in a cabin in the woods in the Czech Republic, then most of us flew to the Grand Prix in Boston, and finally we all met up again in Portland for the last days of testing.
All in all, it was a good trip. Robin won the Grand Prix, Ivan won the Pro Tour, and Martin Juza locked up Platinum. As for myself: I finished 9th at the Grand Prix with Ensoul Affinity (sadly missing Top 8 on tiebreakers for the second Grand Prix in a row) and then went 9-7 at the Pro Tour to finish just outside of the money. My Pro Tour finish was unimpressive, but I mainly felt happy that the team as a whole had done well.
The cabin in the woods was much less deadly than the movie with the same name—it was an excellent location for drafting and testing Constructed. We would drive down to a nearby town to do some shopping once in a while, but apart from a few hiking trails, there was not much in terms of civilization nearby. This was great because there were no distractions, which sharpened the focus on the testing.
Today, I’m going to talk about our Standard preparation for the Pro Tour. I’ll share my views on M15 Limited and Ensoul Affinity at a later time.
My 10-step testing approach was as follows:
Step 1: Define a gauntlet of decks to test against
We picked five popular decks that were expected to make up the larger part of the metagame at the Pro Tour. We started with the deck lists that I provided in a previous article for Mono-Black Devotion, Mono-Blue Devotion, UW Control, Jund Monsters, and RW Burn. After playing a few games, we cut some of the speculative cards in there (such as Jace, the Living Guildpact in Mono-Blue or Perilous Vault in Mono-Black, which didn’t really perform as well as I had hoped), but the lists stayed largely the same throughout testing.
Step 2: Brew up new decks with M15 cards
We built a number of decks around Sliver Hivelord, Nissa, Worldwaker, Genesis Hydra, Obelisk of Urd, Raise the Alarm, Necromancer’s Stockpile, Ensoul Artifact, Ulcerate, Aggressive Mining, Sunblade Elf, Goblin Rabblemaster, Stoke the Flames, all of the enemy-color pain lands, and various other cards. It’s important to get a feel for how powerful the new cards are and to understand what kinds of decks they might shine in.
Step 3: Try out the brews and see what works
It is important to only play the brews against the gauntlet stock lists. It’s no use to play the Sliver deck against the Zombie deck at this early stage in the testing—it’s a matchup that you’re almost never going to see at the Pro Tour. Instead, it’s more valuable to figure out whether the new decks can beat Mono-Black Devotion, UW Control, and the other stock lists.
It’s also important to not waste too much time testing Mono-Black Devotion against UW Control or other gauntlet-vs-gauntlet matchups. These matchups are already well-known, and we wouldn’t learn as much. Better to focus on the new brews. We can always pick a stock list of a gauntlet deck as a fallback option.
At this stage, we’re merely trying to figure out whether certain deck ideas are worth pursuing or not, so all testing is pre-sideboard.
Step 4: Tune the most promising decks complete with sideboards
After some preliminary games, we immediately discarded some brews as horrendous. I may write about some of the crazy builds that we had in a later article—my Chasm Skulker into Aggressive Mining deck was definitely a lot of fun—but we quickly threw these decks into the trash.
For the brews that did work, we tuned the lists based on our testing experiences thus far. Moreover, we built (and, after more testing, tweaked) sideboards for these decks as well.
Let me go over the lists that we liked best at this stage to give you an idea of the decks that we seriously considered for the Pro Tour. I’m going to show you the final (as opposed to preliminary) versions of these decks. I’m going to reconstruct some of them from memory, but I won’t be more than a few cards off at worst.
Ulcerate is an extremely cheap removal spell, but the life loss means that it only fits well in an aggro deck. Fortunately, black has tons of aggressive one-drops in addition to good bestow creatures and discard spells that can help you beat Supreme Verdict, and Mogis’s Marauder is silently one of the best cards in Standard.
This is approximately how the deck looked:
Our initial list had Thoughtseize maindeck, but Martin tested it a bunch and was adamant that the card didn’t fit the deck. It’s still good against unique cards like Supreme Verdict, but often it’s difficult to use it to punch a hole in the opponent’s game plan or mana curve, in which case a simple removal spell would be better.
Although I’m going to spoil the end result of my testing, this is the list that Martin Juza and I registered for the Pro Tour:
Pat Cox made Top 8 with a base-Selesnya aggro deck starring Brave the Elements, but we felt happier with this build. Brave is an extremely strong card, but it doesn’t help much against Supreme Verdict, and black devotion players can try to play around it once they’ve seen it with Thoughtseize. Instead, we preferred to have more hard-hitting creatures like Experiment One and Advent of the Wurm.
Martin already covered the card choices and sideboarding strategies in his article and I’m not going to repeat everything. Instead, I’ll just add a few quick thoughts.
The mana base was a close call. I’m pretty sure you cannot afford Mutavault in this deck with so many demanding mana costs. Temple of Plenty is not great because it comes into play tapped, but I wanted to have 17 sources of each color. 24 lands is sometimes a bit much, especially when you’re facing discard spells, but you still need to cast your spells and need to minimize the number of opening hands where your only two lands are Mana Confluence. So, I stuck with 24 lands, but frequently boarded out a land when I was on the draw against Black Devotion.
Regarding some sideboard cards that we didn’t play: Hunt the Hunter was too narrow; Ajani Steadfast was too slow; and Back to Nature is worse than Reclamation Sage unless Naya Hexproof takes off. I was happy with the list, Martin and I won 65% of our matches in Standard, and I wouldn’t change a card.
There were five main M15 additions for Green Devotion decks: Chord of Calling, Genesis Hydra, Nissa, Hornet Queen, and Reclamation Sage. Nissa was vying for the same slots as Garruk, Caller of Beasts, but we found it to be a solid addition to the deck nevertheless. Chord and Hydra don’t fit in the same deck very well, so we tried both a Chord version and a Hydra version for a while.
The Chord version had a number of sweet one-ofs such as Wasteland Viper, which led to some outrageous bluffs during our testing, but we eventually found the X-spell to be too expensive. In most matchups, especially when opponents have Hero’s Downfall or Supreme Verdict, it was difficult to keep enough creatures around for convoke.
Genesis Hydra overperformed. A mini-Hydra for 3 would typically provide a mana-accelerator or Courser of Kruphix, which was a good amount of value in the early game. Splashing red (which is easy to do anyway with Burning-Tree Emissary) meant that a mini-Hydra could hit Domri Rade as well, so that was a nice fit, too. In the late game when you had Nykthos and Voyaging Satyr going, Hydras for X=20 were not unheard of, in which case the Hydra came close to a tutor for one-of creatures like Hornet Queen.
Here’s the list we had in the end:
This list was tuned to have a good mix between mana accelerators and “big” cards. It was one of the decks that several people liked early on and that we tested extensively.
Once I realized that Mutavault is a Goblin and that Goblin Rabblemaster works very well with Foundry Street Denizen, I built a Goblin aggro deck. It worked reasonably well, but beating Master of Waves and Supreme Verdict turned out to be tough. So, I added white for Chained to the Rocks and Boros Charm.
Here’s the list I tested the most games with:
It’s kind of a mix between the Rabble Red deck that Team Revolution brought to the Pro Tour and the R/w aggro deck that several members of Team ChannelFireball played at the Pro Tour.
I favored Frenzied Goblin over Legion Loyalist. Both are Goblins, but I found Frenzied Goblin helped more in pushing through your creatures. As a poor man’s Firefist Striker, it efficiently taps down huge creatures like Desecration Demon or Courser of Kruphix.
I had Boros Charm and Chained to the Rocks in the sideboard rather than the main deck because Chained to the Rocks was poor against UW Control and Boros Charm was poor against decks with lots of spot removal and blockers. I felt that Lightning Strike and Stoke the Flames were the more reliable maindeck cards. Having 4 lands in the board makes this a true Frank Karsten special, but you never really wanted to board much with this deck anyway, and most of the other sideboard options were bad. Besides, by not having too many painlands in the maindeck, the matchup against RW Burn improved.
Speaking of RW Burn, it was the gauntlet list that we liked best. Accordingly, we considered it as an option to play at the Pro Tour as well.
Matt Sperling may be sick of taking 2 damage from his dual lands, but I prefer to play the full set of 4 Sacred Foundry.
The sideboard may still seem like a random assortment of cards, but all of them have their roles. We added Boros Reckoner during testing because it was great against other red decks.
I consider RW Burn to be the most skill-intensive deck in Standard. It’s easy to pick up and burn out your opponents, but it’s really hard to play optimally. There are so many tough sequencing and scrying decisions that you have to make under uncertainty. You also frequently run into situations where you have to plan three turns ahead to figure out whether to burn your opponent or his creatures. In many matchups, it’s also more of a control deck than a combo deck, and it’s one of those decks where realizing whether you are the beatdown or not is essential in crafting your game plan.
Well, this was the deck that eventually won the Pro Tour.
Stanislav Cifka and Ivan Floch tested it quite a bunch, but I knew before the first day of testing even started that I was not going to play UW Control. I didn’t like the style of play, and I was afraid of getting too many draws. I helped Stanislav and Ivan tune the deck by playing gauntlet decks against them, but I don’t really have anything meaningful to say about UW Cleansing myself.
Step 5: Thoroughly test the most promising decks against the gauntlet, both before and after board
Basically: Jam a lot of games and keep track of the results. Emphasize post-board games if possible. Alternate who is on the play and who is on the draw every game.
I realize the issues with small sample sizes and I know that a result like “GW Aggro defeated Mono Blue Devotion 6-4” doesn’t imply that it’s a 60% matchup with absolute certainty, but it still says something. And once you have played a hundred games in total against the various gauntlet decks, the observed percentages will get reasonably close to the true percentages.
Of course, it is also very valuable to keep track of how certain cards are performing, to gain an understanding of which are the key cards in the various matchups, and to continually make tweaks to the deck. The numbers are not everything. Nevertheless, setting up an archive of the results will help in the next step.
Step 6: Calculate win percentages against the expected metagame
Suppose we have played enough before-sideboard and after-sideboard games in a certain matchup and have estimated the before-sideboard win probability (let’s call that B) and the after-sideboard win probability (let’s call that A). Then the probability of winning the match is equal to BA+B(1-A)A+(1-B)AA.
Once you have estimated the match win percentages of every brew against every gauntlet deck, you have to weigh them according to the metagame expectation. To get a good guess of what the metagame is going to be at a tournament, it is wise to ask several people to separately write down the percentages they expect. Letting everyone do this separately avoids groupthink, and the average of each person’s guess will usually lead to a good estimation. If at this point you find that you expect a reasonable amount of a deck that wasn’t in the gauntlet before, then you should consider testing against that deck, too.
All in all, this led to the following spreadsheet:
(Click to embiggen.)
The results should be read as saying that, e.g., we tested a number of games of Mono-Black Aggro against Mono-Blue Devotion before sideboard, of which Mono-Black Aggro won 6 and Mono-Blue Devotion won 12.
- I don’t have the results of UW Cleansing because Floch and Cifka were keeping track of those separately.
- The column “Black Devotion and BW Midrange” includes the results against both archetypes together. We mainly tested against Mono-Black Devotion, but we also had a BW Midrange deck and another Black Devotion deck with a green splash lying around. We could’ve put them in separate columns (which would’ve revealed that, e.g., GW Aggro is worse against BW Midrange than Black Devotion because Blood Baron of Vizkopa is tough to beat) but there’s so much overlap between these decks that we kept them together for simplicity.
- A result of “100-100” implies that the matchup had not been tested and is counted as a 50% matchup.
- The numbers for “Our expectation of the metagame” were equal to the average of Martin’s, Stanislav’s, Ivan’s, and my guesses. The row “At the actual Pro Tour” was obviously added afterwards based on the Day 1 metagame breakdown, but it matches our guesses quite well.
Step 7: Pick a deck two days before the Pro Tour and stick to it
It’s sometimes tempting to postpone the choice for a deck and/or to make last-minute deck switches, but it’s typically more valuable to stick to a deck and to use the last day to gain experience with it, tune the sideboard strategies, and learn all of the card interactions.
As I mentioned, Martin and I went with GW Aggro. Part of the reason was that it had the highest expected match win percentage against the expected field according to our results. In addition to that, we liked the way the deck played, felt comfortable with the sideboard, and were happy with the individual power of the cards and the mana base.
Step 8: Test the mirror match
It’s easy to forget, but if you like a deck, then other people may play it, too. Knowledge of the intricacies of the mirror match can give you a substantial edge if your opponent forget to test it. For example, Martin and I figured out that Archangel of Thune was pretty good in the GW aggro mirror, especially when combined with Soldier of the Pantheon, so we tweaked their numbers after that.
Step 9: Write down your deck list and a sideboard plan for easy reference
Please check that you have 60 cards in your main deck. The sideboard plan is just a guideline that you can deviate from during the games based on what your opponent is doing exactly, but it can act as a memory jogger on which cards were good and bad against stock builds of the most popular decks.
Step 10: Sleeve up your deck and give it your best during the matches
Pretty self-explanatory. It certainly worked for Ivan Floch!
Hope you enjoyed this inside look into my preparation approach. Next weekend, I will be in Utrecht to do video coverage of the Standard Grand Prix. Best of luck to anyone going there or to the team GP in Portland; it is poised to be an exciting weekend.