Introduction

What if I told you that I had a great deck that I piloted to the top four of two tournaments? It suffered only two total losses in games, in these events, and as you can probably tell, those happened to result in it being in the top 4. But wait, there's more! It has a stable mana base and uses a lot of cards that are otherwise overlooked, and it may just break the format wide open. Are you interested in learning about this deck?

When you read Magic articles, you come across a lot of questions like this. They often, however, end end up being an exercise in middle-school problem solving where you need to identify which pieces of information are missing. For this set of questions, what could these missing pieces be? Could it be that there were not many people in the event? How competitive was the event?

I won't provide a link for this to avoid total embarrassment, but in the example I'm thinking of, a player decided to write an article about a deck that he brewed up. The deck's four color mana base was made up mostly of basic lands and used a plethora of cards that seldom see the light of day in Legacy. Now, the deck did make the top four of two events, but these were weekly events which each had under 32, and probably even under 25, players. This creature-based deck also managed not to play against Swords to Plowshares nor Path to Exile in any matches. My question is,, how likely is it that this was a fluke? I consider it to be very likely, and thankfully, it doesn't seem as though too many people took his advice, as I didn't see a discussion of the deck turn up on any Legacy forums, nor did I see copies of it running around Magic Workstation. So, I think it's safe to assume that the attempted trolling that the author hoped to stir Legacy with was largely ignored.

After reading that article, however, I was concerned, thinking that people who were new to Legacy or even competitive Magic were reading this. What if they thought that roughly twenty games of Magic was a suitable base to not only write an article on, but to attempt to storm a fiercely contested event wit?. It then occurred to me that there is very little written on how to approach testing your deck. The few resources available are widely quoted and are generally of assistance, but even less is done with Legacy which is likely one of the most difficult formats to test against, and metagame for competently. What I want to do today is to share some of my insights into this with you.

What Are You Testing For?

The preliminary question to answer before getting down to testing is ‘what is the purpose of testing'. For the purposes of this article, we're primarily going to be looking at testing for a large [Legacy] event, but I will discuss a few of the other goals that you may have. Testing the functionality of a deck isn't going to require much work and can probably be done with a few dozen goldfishing sessions. Testing for a local metagame is going to allow you the privilege of knowing what you're walking into and the relative play skill of those around you. This means that you're going to be able to really refine your testing not to what is strongest in the community at large, but to attack a more identifiable threat, as you're going to have an idea of what the most competent players are going to be able to bring to the table. It's important to set your goals according to the size of the event you're going to be looking at, but one thing is certain, if you don't test then you're going to be at a disadvantage. This is an incredible disadvantage in a large event if you don't test a great deal to prepare.

Who You Should Be Testing With

The single most important thing about testing a new deck is who you are testing with. Firstly, you should be testing with someone who is at least competent with the deck that you want to test against. This should almost go without saying, and I wouldn't think of mentioning it if I didn't see people testing with partners that have no idea what they're doing, and then accrediting that testing. In Legacy, as in Magic, you'll always be able to pick up free wins because the opponent doesn't know how to properly play his deck. You don't do testing in hopes of being paired against someone who is entirely unprepared, you do it to learn how to combat the deck in question.

There was a story that came out of Columbus after the Grand Prix. I can't verify the source, but the legend goes that an Aluren player is playing in day one and resolves an Aluren against his opponent. He then asks him if he wants to concede. His opponent, being unfamiliar with the deck, declines. The Aluren player actually doesn't know how the combo works and ends up losing the game. There is a similar story for the Flash combo and probably at least one for every year a new break-out combo deck appears on the horizon.

When I was at States last month, I was talking with a friend of mine who is a juge. He is generally the default head judge for any event in the area. However, this time around he wasn't the head judge, but rather, just a floor judge. When I asked him why that was, he responded, "I have nothing to learn from head judging this event." That brings me to the second point, which is that not only do you want to find a player that is good at the deck you want to test against, but ideally you want to be testing with a player that is better than yourself, at least at some aspect of the game. This is going to require a bit of humility on your part, but what do you have to learn from playing against people that struggle to compete with you? When you play against someone who can rival or is strictly better than you, you are put into a position where you must stay on your toes, evaluate every option, and perhaps re-evaluate to find better solutions and perhaps gain greater insight.

Before moving along I'd like to talk about goldfishing. For those of you who aren't sure what goldfishing is, it's assuming that your opponent is a ‘goldfish' that never does anything. This is generally done when you're alone and just want to familiarize yourself with a new deck or some recent changes. While it is an acceptable passtime, goldfishing should never be heralded as real testing. The reason for this is that while you will sometimes generate wins from the opponent not being able to do anything relevant, you were going to win those games regardless of if you knew how to pilot your deck against no interference anyway. Goldfishing is really only helpful to familiarize yourselfself with the deck's basic interactions. The problem is that sheltering the deck from the pressure, disruption or permission from opposing decks creates a skewed view of how to play the deck, and results in the lessons gained from goldfishing not only being a moot point, but potentially setting back any actual progress with learning a new deck, as you've become accustomed to unrealistic lines of play.

What Do You Test Against?

The answer to this question is going to change as time goes on, but there will always be resources for you to check. Up until recently, deckcheck.net would be my first resource to check the trends in the format, as it provided a fairly accurate barometer in a lot of circumstances, now those options are limited. Mtgthesource.com is clearly going to be the primary source for Legacy information. There are some issues with the length of some threads and how dated some material is, but overall it is excellent. Mtgsalvation.com has some of the most popular Magic forums out there. The user base tends toward the more casual crowd however, so it generally isn't as strong of a resource. The problem is that neither of these sites have the same aggrigation of data to compare with so until a suitable replacement for Deckcheck is found, we are in the dark. Thankfully, if you're planning to attend a Star City event, then the metagames for those are pretty similar from location to location, as a great deal of hiveminding appears to be happening. Trends like this are the source for my skepticism that the Star City events represent the format's overall health. Right now we're seeing a lot of Survival decks in Star City events, previously we were seeing an odd amount of Belcher and Counter-Top Thopter decks, neither of which have fared very well outside of that series.

For now, I can share with you what I think the metagame's most important players are, although who knows how long this information will be good for.

Top Tier of Importance

• Merfolk
• U/G Survival Madness
• T.E.S.
• Zoo

Second Tier of Importance

• R/b Goblins
• New Horizons
• Bant Counter-Top
• Affinity
• B/g/w midrange decks

Any third tier decks are really going to just need to be decks that you're concerned about your match-up with.

What To Prepare For

In most Legacy decks, there is a lot of room for flexibility and variance among decks. This presents a huge challenge when attempting to maximize the time spent testing. For example, let's look at Zoo. There are quite a few variants: Traditional (Naya) Zoo, Big Zoo, Cat Sligh, Goyf Sligh, Dark Zoo, Blue Zoo and Domain Zoo. Blue and Domain Zoo are the least popular, Dark Zoo is a fringe deck as well, but one that I actually anticipate getting played more in a field of Survival decks. Zoo on the whole has been on the decline, and Cat Sligh has fallen a bit more than Big Zoo and the traditional Zoo list, but all three are reasonable to expect in force in a large scale event.

So, with a finite amount of time to test against Zoo, which build should you test against? I think there are two excellent theories on this matter. First, I have always been a fan of testing against the reasonable build of the deck that is most unfavorable to yours. When we were testing out Peacekeeper's effectiveness, I needed to know if it was reasonable to expect any Merfolk players to be running a non-Submerge bounce spell in the sideboard. When the data came back that practically no Merfolk decks were using it, this was excellent news, but because about 5% of them did, I was sure to still test against a Merfolk deck with bounce spells as I needed to know that this plan was effective against even the most hateful builds. What I didn't test against were the builds of Merfolk in my Local metagame. This is because those builds had been heavily metagamed against control, which was highly unreasonable to expect. While the local Merfolk decks still weren't using bounce spells in the board, they had a very aggressive plan against control that you just don't see in other environments.

Similarly, when we tested against Zoo, we had to decidedon a build that would be the most reasonably challenging. I was most concerned with the burn heavy builds that maindecked not only Price of Progress but Fireblast as well. The problems with these decks was the fact that they used creatures for the early game and then ended the game with burn, which gets difficult to counter. When we were testing, my partner was using Sulfuric Vortex in the board which is a result of playing in a control heavy metagame. Vortex proved to be a huge problem, but it was a card that was virtually unheard of in Zoo boards, so losing to it felt somewhat hollow.

Because Zoo has so many configurations and flex slots, another theory on how to approach these testing sessions was devised by a friend. Testing against builds that synthesized all common elements of the decks. Initially I was skeptical of this method, but after trying it and thinking about it, I realize that it does have merits, especially if you're not sure what the suspected weak points of the deck are. This method is actually going to require a lot more games to be played to even get a well-rounded view of the decks, and then you'll probably want to do additional testing with a refined list that contains the elements that you feel you need to test against.

You may have noticed that I didn't mention testing against stock lists or the most popular lists. The reason I don't feel that they are necessarily the best to test against is that by the time those sorts of lists become public, from large events, the metagame is going to begin shifting to beat it, just as you have probably already made a conscious decision to play a deck that is either the most popular archetype or that beats the most popular archetype. When you're prepared for the worst that a deck can throw at you, it should be far easier to win against a less hateful build. This method also gives you the knowledge of what to play around, and when you're winning you should really always be considering the things that you can lose to, and play around them as much as possible. If those things aren't there then you're going to win regardless, but if you can play around potential traps, you've given yourself the maximum chance of winning at what should be no extra cost.

I want to close this section by stressing that this can only really be done with decks that are very modular, such as Zoo, Goblins, or some popularized control decks. This exercise can't really be applied to Storm combo decks or other heavily stream-lined decks. In these instances its best to play against stock builds, as they generally won't vary too much from pilot to pilot, and then you determine what your best routes of play against these decks are.

Testing Format And Procedure

I've been in weekly tournaments with a lot of competent players who are looking to use the tournament as actual testing. There are some merits to this, people are playing to win and they are playing competitive decks. The issue is that the actual format of the tournament provides for very few rounds. You're looking at four or five rounds times two or three games, plus perhaps a top eight: over the course of 5-7 hours you get at most 24 games in, assuming that you play every round and go 2-1 in every round. More realistically it is going to be less than twenty games and maybe as few as four if you do poorly. Furthermore, the caliber of players in the event is going to be questionable up until perhaps the final round of swiss, and you're going to be walking into a metagamed area with decks that may be unrealistic for a large scale event.

In five hours, I think you should be able to cover 50ish games on the low end, that was able to with a control deck against the gauntlet. It is most likely possible to do even more with faster decks. I suggest playing 10-15 games preboard and then 20-40 games post board. Traditionally, people suggest a 10/20 mix and even toward 30 post board games, the reason for this being that you're always going to play a game one and you're always going to have to play a game two as well, and very often you'll be forced to play a game three. Therefore, it makes more sense to familiarize yourself with how the post board games are going to go. The reason I suggest playing even more games is that a lot of decks are really going to shift how they play on game one as opposed to game two. Merfolk vs. Zoo is the perfect example. Game one Merfolk is often stuck with a lot of guys that just die to burn and not much action: however, post board, many Merfolk decks shift play style and begin stealing Zoo's creatures, which stand up much better to burn. Merfolk also tends to bring in Umezawa's Jitte and Blue Elemental Blast. It shifts from an aggro deck to an aggro control deck. If you're only familiar with how to play game one against this deck, things are going to be a lot trickier when they start attacking you with your Tarmogoyf and breaking fetchlands is suddenly a liability to Submerge.

The other reason to play so many post board games is that a lot of decks will have a bye in game one, such as T.E.S. against much of the field or perhaps in reverse a Counter-Top deck vs. any Storm deck. Playing post board games removes the polarity in these games, where T.E.S. has access to Pyroblast to attempt to deal with Counterbalance, but causes the deck to be much less explosive, meaning that the pilot will need to figure out how to deal with the slower cards and other hate the opponent has brought in.

Concerning mulligans during testing, I've been a proponent of having the opponent take free mulligans until they find a hand that is viable, whereas the deck that is being tested honors mulligan rules. The theory behind this is that you will win against opponents that keep bad hands and you will when when they mulligan into oblivion, but those are not things that you can test for. You're looking to play your hands against reasonable starts from the opposing deck, and you'll need to know what five and six card hands can work as well. The purpose of testing isn't to win, but to gain knowledge. Even if this unfavorable testing method results in net losses, the benefits should outweigh the loss of morale that comes from playing against a deck that perhaps doesn't need to play by the rules for mulligans. In Addition, who plays first should always be alternated, and not based on who won the last game, this gives you a good mix of being both ahead and behind on tempo.

It is also very important not to metagame the game-ones that you test. This time when I use the word "metagame," I use it in a different sense, because in this instance I'm talking about playing game one under the assumption that you don't know what your opponent is playing. You have to look for hands that are keepable against a blind opponent, not ones that shine against the deck you're currently playing against, (save that mentality for testing games two and three.) It is going to be ideal if your opponent can keep that mentality as well. While I discussed keeping hands that are playable, that doesn't mean hands that are ideal against what you're playing, nor should either player begin taking aggressive steps against the opposing deck before it can be quantified, in game one.

Quick Notes on Proxies

Very often, you're going to need to proxy up decks to test and sometimes in a hurry. When we were testing for Columbus, we were doing testing non-stop for some time and we have developed some means of proxying up decks that I'm sure are not unique but may be helpful regardless.

Firstly, stock up on white cards and Plains, maybe set them aside, as these are by far the best cards to write on. Writing just a card name should be good enough, but you can go further with it if you need to. Sharpies are the marker of choice, pen and pencil never work well. I advise making the entire deck proxied if you don't have most of the cards and not sleeving the proxied decks, as the marker may smear and ruin sleeves, secondly, you waste a great opportunity to double up on your proxies. As a big fan of recycling, I've proxied as many as four decks on a single set of 75 cards. It's a great tip to designate cards as being sideboard cards by writing "SB – Cardname" on the card.

Non-Traditional Means of Testing

Testing on-line has been the resort of many Magic players that don't have infinite free time, but I feel it's important to discuss the implications of testing without a dedicated testing partner on either Magic Online, a service like Magic Workstation, or the newly blossoming Cockatrice. All of these services cater to casual players for the most part, and so the skill level of people that you come across will often leave something to be desired. This means that the wins generated from testing on each of these holds far less weight than in a real tournament or with any player with a known skill level. More importantly however,

Magic Online nor MWS has a real Legacy metagame.

This means that regardless of how well Aluren was doing in on-line events, it was still coming into the Grand Prix blind, and for good reason. Why is this? Well, for one, Magic Online doesn't have access to all the cards that are available in the real world. Furthermore, the card availability is entirely skewed. For Magic Online, Force of Will is nearly $100/100 tix, whereas a card like Moat is under 5, not to mention that Urza's Destiny to Prophecy still don't exist. The numbers on cards printed before Mirrodin are incredibly low, resulting in another branch of cards that have really skewed values, such as Vindicate. Conversely, every card from Imperial Recruiter, Jace the Mindsculptor and The Tabernacle At Pendrell Vale all the way to Cursecatcher is free. MWS is obviously free, meaning that experimentation with decks is free. There is little incentive to optimize a list, as there are so few tournaments available though Workstation leagues. This applies to Magic Online as well, where there is just the daily event, which doesn't even have that many players and then nothing beyond that worth playing in.

Unless you're testing with a known source over either one of these, it's unlikely that playing with either one of these methods can honestly substitute for real, valuable playtesting. MWS can be a valuable resource for testing from afar, but I wouldn't put too much stock into playing against Pox and Rock decks for an afternoon unless there is a very bizarre shift in Legacy.

Conclusion

While I feel that this week I may have written about a topic that many people are familiar with, I hope that some of my experiences and tips can help others to test more effectively. Clearly, these tips aren't news, but more like reminders that can help to really improve your game and time management. I'm sure many would like to see additional tips and suggestions in the comments for this piece and look forward to gaining greater insights into testing. Next week I want to take a step back and address some aspects of the step previous to testing: deck building. Thanks for reading.

~Christopher Walton

im00pi at gmail for Electronic Mail

Master Shake on The Source