I need to talk about playtesting, the gauntlet, and why we writers so often get it wrong. I’ll also list a half dozen common playtesting mistakes, and how to avoid them. But first, I’ll pass on the most important piece of playtesting advice ever. The Numero Uno. Here it comes. Ready?
That’s it. Playtest. Actually play the game. Make decks, shuffle cards and start playing. Play game after game. Sideboard. Play. You don’t win Regionals by reading articles on playtesting. You win because you playtested – for hour after hour.
And, of course, because you followed my guidelines on playtesting.
Rule number one: if you can, you want to playtest as part of a team. You need to do a lot of research on decks, work on play strategies and sideboarding strategies, learn to play the decks proficiently and keep current on the evolution of the decks. It really helps to be able to divide up the work. Having more faces and opinions in the mix helps as well.
The first part of playtesting is to develop a gauntlet. It isn’t enough to build your own custom deck – you have to test it against other decks. That means you need to build other decks – and those should be good other decks. They should also, as far as possible, be the decks and builds you will face when you actually arrive at the tournament.
Saying you should playtest against an accurate gauntlet is easy. Deciding what is in that gauntlet is not. So here are some”tips” – or, more accurately, the tricks writers and at least some pros use to develop a new gauntlet.
1) What was good before?
Unless critical portions of a deck have rotated out, the decks that were tier one in the last version of the format are probably still good. The last major* tournament is a good place to start. For Darksteel Standard, everyone started by looking at Mirrodin Standard, starting with the State Champion decklists.
2) Of those, what had holes plugged?
Next, look at the solid decks from that period and see whether any had significant shortcomings that the new set fills. For example, Goblin Bidding had a fast start, and a nifty late game trick in Patriarch’s Bidding. What it lacked was solid card drawing to maintain pressure through the mid-game. Skullclamp is an obvious solution to that problem: it plugged the deck’s biggest holes. That was a pretty good indication that Goblin Bidding would be Tier 1in Darksteel Type Two, and it is. (Hey – I’m predicting the present. Ain’t I amazin’?)
3) What lost ground?
If any previously Tier 1 deck loses ground, it is usually because of one of two reasons: a really good – and playable – hoser was printed, or the format changed. For example of the first reason, imagine if Circle of Protection: Red was reprinted as a two-mana artifact cantrip: Goblins would be hosed. We have an example of the second reason: Sarnia Affinity, the counter heavy Broodstar deck, has faded because of a change in the format, namely the format getting faster. It is much easier to slip threats, including Skullclamp, under Sarnia’s counters, so Sarnia is no longer Tier 1. So look for decks that are going to slip, and drop those from the gauntlet.
Note: Tier 1 decks from the old format that don’t lose anything, but don’t gain anything, should be placed on the watch list. They may still be good, but there is an excellent chance that the format will shift out from under them.
4) What broken cards are obvious?
What cards are everyone raving over? Alternatively, what card is everyone calling for to be banned? After eliminating all of those that cost more than four mana**, look at the remainder, and build decks with those card. Then put those decks into the gauntlet.
But I mean it when I say”obvious.” Skullclamp and, to a lesser extent, Arcbound Ravager were obviously very powerful to anyone that had seen or played with them. Since nearly everyone could see them, nearly everyone was building those decks. That makes those decks things you are likely to see, and therefore gauntlet material.
Other cards were/are also amazing in certain decks, but they were/are more subtle. Don’t include those in the gauntlet until they become widely discussed on the Internet. You may want to build those to play yourself, but don’t start playtesting against them, until you know someone besides your playtest buddies might play them. The gauntlet should contain decks you are likely to face.
So, those are the trick to getting a start on the gauntlet. The other half to building a gauntlet is to netdeck like crazy. Look at what others are talking about, and what tech others are sharing. That’s not perfect, but it is a start. Keep an eye on the forums, and check out any threads that get a lot of responses. Whether you like netdecking, or whether you consider netdecking to be a mortal sin, the net should be your starting point for a gauntlet. After all, the net will be the source of a lot of your competitors’ decks.
Does that mean that the gauntlet descriptions on the net are correct? Almost certainly not. We writers, as a group, tend to be backwards focused. We often misstate the metagame – but not always, and often not badly. However, we are less than perfect.
I am going to do something awful to my fellow writers. I am going to compare early metagame predictions with the actual metagame, as it is shaking out now in European Regionals and the Type Two last chance qualifiers at PT Kobe. I’ll do it to myself, too. Here were our predictions of the metagame, from before any serious tournaments took place:
M.W. Gribbin, 1/20/04, Finding the Tinker Deck in Standard
Broodstar Affinity, Astroglide, Goblins, White Weenie, MBC, U/W Control, B/W Control, Land D
Richie Proffitt, 2/16/04 , Ya Gotta Have Blueberries
Tier 1 Decks: Goblin Bidding, Zombies/ Aggro Black Decks, Mono-White Control, Affinity/ Tog Affinity Tier 2: Astral Slide, Colossus Staff, B/W Control. Tier 3: Elf Clamp, R/G LD, G/W Control, WW
You can get similar results from other writers on other sites, but even the premium articles didn’t do any better at predicting the metagame than we did – in fact most other writers got it a lot worse. I singled out StarCity writers because they are better, not worse, than the average.
Here are the breakdowns (assuming Knut can handle converting a table from my clunky word processor to the web.) The”writers” column totals our predictions that these decks would be tier I or strong Tier 2. The”Played” column is next, for easy comparison, and is the total of all T8 decks in French and German Regionals, plus the Kobe Last Chance Qualifiers. For France and Germany, I’m counting all T8 decks – remember that there were several Regionals in each country. The French decklists came from Brainburst, and the German ones from Portal Maroox.com.
Ravager Affinity, no blue
Ravager Affinity, w/blue
R/G (LD, Beasts)
Straight Goblins (no bidding)
Mono-black aggro / zombies
RDW / mono-red LD
So, does this show that we writers are bad? Not really. Everyone had Ravagers in their Broodstar decks early on, but it took time for those decks to evolve enough to realize that Broodstar Affinity needs Ravagers, but Ravagers don’t need Broodstar. Beyond that, the really good decks, like Ravager Affinity, took time to develop. Early predictions were based on good builds of some older decks against bad builds of newer decks. It takes time and effort to perfect decks.
Moreover, some writers don’t give away their tech decks, either because they want to play the decks themselves, or, more commonly, because the tech was created by someone else, and that person was not ready to release it. I know about decks I cannot talk about, and I’m farther out of the loop than most pros. So don’t rely completely on the net; it isn’t perfect.
However imperfect your information, you still need to have a gauntlet. Read, study, discuss this with your group, then put together a list of at least half a dozen decks that you expect to see. Decide on generic decklists. You want typical builds. Here are four rules for the gauntlet:
1) Don’t tech out the gauntlet:
I said”play typical builds,” and I meant it. Don’t add unexpected cards – especially not cards no one else has considered. You might want to maindeck Stifle in Ravager Affinity (which should not have Blue, in any case), but most people won’t. If your gauntlet has Stifles, you will learn to play around Stifle, but you won’t learn how to play the matchups you will actually face at Regionals.
2) No pet decks in the gauntlet
You play your pet decks, at least the one you intend to play at Regionals against the gauntlet, not in the gauntlet. Don’t waste your playtest time playing your secret tech deck against your friend’s deck – the odds are very much against you seeing that matchup at Regionals.
3) Use real cards, or photocopied proxies
I have described Adrian Sullivan playing card method in the past, but I much prefer playing with actual cards, or proxies made with a scanner and color printer. During playtesting, you are learning patterns, and cards to watch out for. If you play with proxies that look like the real thing, you will ingrain subconscious clues, and will later pick up on board positions that you would miss if you just play proxies. (Of course, StarCityGames.com would prefer that you use real cards, so go visit the online store…)
4) Let the Gauntlet evolve
Over time, the typical play lists for gauntlet decks will change. The Internet drives innovation, as do the results of large tournaments. For example, early MWC decks were running Mindslavers and Weathered Wayfarers maindeck. They beat some decks, but those builds of MWC lost to fast Affinity. To survive Affinity, MWC started maindecking Damping Matrix, which shuts down Mindslaver and Wayfarer, so those cards were also replaced. MWC has evolved, so the gauntlet should, as well.
Evaluating new decks:
In addition to watching the web to develop the gauntlet decks, and to monitor their evolution, you also want to watch for breaking deck ideas. You don’t want to get blindsided by combos or strategies, if you can at all help it. However, for every new Trix deck posted, there are a couple dozen bowls of soggy, moldy corn flakes. You need to be able to tell the difference. Once again, I have some rules/questions for doing so, and I’ll illustrate by analyzing the Tangleroot / Skullclamp / Elves / Tendrils combo deck.
1) Does it work?
The first question is whether the combo works, and works well. The first step in determining that is to look at the parts, and double check the rules. If it looks good, then proxy it and start goldfishing, then playtesting.
For the elf deck, the rules check is pretty easy, as is figuring out the deck. It is a blast to goldfish – and it is fast.
2) Is it consistent?
A combo that fires once in a while is like a life vest that floats on occasion – you don’t want to depend on it. You need to playtest (or at least goldfish) a few dozen times to see how consistent the deck really is. Theory and analysis is no substitute for shuffling cards here.
For the elf deck, the consistency is pretty good. It generally goes off turn 4 or 5, and only occasionally stalls.
3) What crushes it, and are those cards commonly played?
Once you decide that the deck or combo really works, the next step is to look at what absolutely destroys the deck. For example, a land destruction deck folds if the opponent gets Sacred Ground out, while straight Goblins, no Bidding hates CoP: Red and Silver Knight. Next, look to see whether decks you expect to see played will have those cards either maindeck or sideboard. If a deck absolutely folds to something like Death Pits of Rath or Cowardice***, things that aren’t in any decks anyone would actually play, who cares? But if the new deck suffers from cards that will see a lot of play, watch out.
For the Elf deck, it completely folds to Damping Matrix, and really suffers if board sweepers, like Pyroclasm or Wrath of God, resolve. It also folds to Goblin Sharpshooter. Those cards are being played.
4) Can you sideboard around the crusher cards?
If you find that your deck folds to certain cards or strategies, you can consider playing around them, or carrying answers. If you can find answers to the problems, great. If not, pass on the deck.
The Elf deck folds. It can sideboard in answers to Damping Matrix, like Oxidize, but that dilutes the combo too much. It has no practical answer to Goblin Sharpshooter. Ergo, the combo elves deck has no chance, so it does not make the gauntlet.
Again, I cannot over-stress the need to actually play the decks. You learn by doing. Do!
Do! Do! Do!
(On rereading the previous, all I can say if that I am truly sorry, deeply embarrassed, and wish to apologize to everyone.) [Your political career is ruined, Pete. – Knut]
Moving on, here are some rules for playtesting:
1) Learn to play the gauntlet decks.
The first step in playtesting is constructing the gauntlet. The second step is to learn to play the gauntlet decks. You need to know how to play the decks, whether you are playtesting a gauntlet deck you are thinking of playing at Regionals, or just providing opposition for a teammate to test against.
I recently played a few fun games against a kid who had proxied up Goblin Bidding. It wasn’t playtesting, because it was clear that the kid didn’t know how to play the deck. For one thing, he didn’t sac goblins to trigger the Sharpshooter before (or at least in response to) Bidding. He passed up a half dozen free damage points. I won – at four life.
That isn’t playtesting. Playing against a bad player doesn’t tell me that my deck can beat Goblin Bidding – it tells me I could beat that kid with a precon. (Hopefully, he’ll improve. I did explain how he screwed up.) The point is, however, that I didn’t learn anything important during those games. And if your team cannot play the gauntlet decks well, what you’re learning may not be useful, either.
If your team is big enough, assign decks to each player. Where possible, make sure those assignments match a player’s style: beatdown to the beatdown player, and control to those who like, and are skilled with, control decks.
2) Play the Gauntlet against itself
Here’s the simplest way to learn the gauntlet decks – play them against each other. Make sure everyone gets some experience with each deck, at least on the receiving end. Once again, you learn to play decks by playing them, and you need to learn the gauntlet before you can learn how your pet decks do against it.
If you assign decks to team members, have a secondary player for each deck. If not, and if, for example, you assign U/W Control and MWC to the same person, how will you learn anything about the U/W Control on MWC matchup? You need to have skilled players playing both sides of every matchup.
General Rules for Playtesting:
1) Don’t play to win, play to learn.
Note: If You Read Only One Section In This Whole Article, Read This One! If you remember only one concept, remember this one: Don’t play to win, play to learn.
You are playtesting to learn how a matchup plays out. You are not playtesting to win playtest games. You often win matches because an opponent screws up. You playtest to learn how to avoid screw-ups, and to learn how the match plays when no one screws up. Therefore, if you are playtesting, allow take-backs and correct mistakes. If you see that your opponent taps mana incorrectly, or makes a bad play, stop the game and fix the problem. You need to have each playtest game played as well as possible. Mistakes don’t teach you what you want to know – perfect games do. Correct mistakes, back up where necessary, and pay attention.
If you are unsure about the right play in any situation, stop the game and ask. Discuss your options, and talk about what you need to play, and play around. I have learned far more when games stop and the whole team discusses options than when the game plays out and I have to guess whether I had a better play.
2) Playtest your deck, too.
You want to playtest the gauntlet against itself, to learn how the decks play. However, the purpose of the gauntlet is to learn how to play your own deck, to tune your deck and determine how your deck can beat the gauntlet. Remember, the goal of playtesting is to learn how your deck can win the big tournament – so focus on that. Once you have decided on your deck, you only want to be playing gauntlet decks to provide practice for your teammates.
Just remember to playtest your deck against the entire gauntlet. You want to know the matchups against the most common decks inside and out – but you also need to know how to beat the more random decks as well. You may well face something random, even tier three, in the first round, and in round two be facing a random deck that met an unprepared deck round one and got lucky. After that point, you should be facing deck you have prepared for – but only if you can win the first two rounds.
3) Playtest with Sideboards
This is one of the most important part of playtesting, but one that is all too skipped. In fact, it is so important that I’ll come up with some subrules, just for this.
3a). Use Sideboard A and Sideboard B cards – and 30+ card sideboards
It is a pain to pull cards for all the decks. It is more of a pain to pull sideboard cards. If I am going to cheat and use proxies, sideboards is where I’ll do it. The simplest is a straight, fictitious substitution – for example,”this game, all the Damping Matrixes are now Sacred Grounds.” Then just play the games. That is great when you sideboard equal numbers of cards, but it isn’t perfect – sometimes you pull two of this and one of each of these other cards. For that reason, I use some random card, either cards that would never be played (Chimney Imp, Mudholes, whatever,) or random commons with”Sideboard A” written on four cards,”Sideboard B” written on four more, and so on.
In the early playtesting, I don’t try to limit sideboards to fifteen cards. Instead, in each matchup, I talk with my playtest partners about what I should bring in, and what I should take out. Then I test that change. Sure it may be excessive for Goblins to bring in four Flashfires, Pyrite Spellbombs, and four Stabilizers against MWC, but the first question is whether those cards help or hurt the matchup. Later on you will have to cut the sideboards, but before you get to that point, you need to know what works, and what doesn’t.
3b) Remember that the other deck is sideboarding, too.
It is not enough to determine what you can bring in to wreck the opponent, you also need to know what they will bring in to wreck you. For example, if you are playing MWC against a Red deck, and don’t have a plan or answer to Flashfires, you are in for a long day of side events. You and your teammates should be discussing sideboarding measures and countermeasures.
3c) Refine sideboarding strategies
Once you have a good idea of what is important in each matchup, then you can start to cut down the number of cards. I use the following method. First I chart every matchup, and list all the cards I would want to bring in. Then I list the cards I would take out. Remember, it does no good to list twenty-five cards to sideboard in, if you can only find four cards to sideboard out.
Once you list ins and outs for all the major decks, you will probably have more than fifteen cards. You need to cut cards for the actual tourney. Here are the rules I use: First, if you win the match whether or not you sideboard – in other words, if the sideboard just makes you win more – you don’t need those cards. Second, if you can find a substitute card that lets you combine sideboard cards, use that (Oxidize + Demystify = Naturalize). After that, if you still need to cut cards, cut the ones that appear in only one matchup, or in matchups that are less likely. At this point, things are getting tricky – do the best you can.
3d) Memorize your sideboarding patterns.
This is pretty obvious, and should be second nature, but make sure you remember what goes in – and what goes out – for each matchup. It is simple to do at the kitchen table, with your notes at hand. It gets a little more iffy in the top eight, with the pressure on and after a long day. And double check you patterns with your teammates.
And now we return you to”general rules of playtesting,”
4) Don’t forget the mirror match
The one problem with a gauntlet, at least when I build them, is that I build / proxie one version of each deck. That means that Ingrid and I tend to play all the expected matchups, but not the mirror match. If you are planning on playing a gauntlet deck in a real tourney, practicing the mirror is critical. The more dominant the best deck is, the more likely the mirror match is.
The simple solution is to decide on your own deck – the one you intend to play – early on in testing. Then build your own copy, with real cards as you can get them. Play that against the gauntlet, including the mirror. And don’t neglect sideboarding for the mirror as well.
5) Watch out for good deck – bad player / bad deck – good player distortions.
In any group, some players are better than others, and some are much better when playing certain decks than others. Say, for example, that one member of your playtest group, call him Bill, is a milling deck god. No matter what the format, Bill builds a control deck that wins through milling. He keeps card counts in his head, draws what he needs, and always knows whether to cast something or activate Millstone, or Grindstone, or whatever. Moreover, he is lucky – he topdecks answers, shuffles well and never needs to mulligan.
Bill builds, and plays, killer milling decks. Right now, early in testing, Bill has a Raven Guild Master / Whispersilk Cloak deck that trashes your gauntlet. It is beating everything fairly consistently, to the point that you are sideboarding four Darksteel Colossi into Goblins, but Bill is still Stifling the return to deck triggers and getting wins. So, does Bill have the metagame crushing deck everyone always looks for?
Maybe, maybe not. I said it was early in testing, so everyone (except Bill) is probably learning their decks. The playtest group has probably not learned how to get the best performance out of their decks. Bill, on the other hand, may have the optimum build (for him), and know how to play it perfectly. Once the other decks, and players, are up to speed, the milling deck may lose its luster.
The easiest test is to give Bill’s deck to everyone else, and see whether they can duplicate the results. If everyone can pilot the deck to a winning record, you have something. If not, then you have a gifted player dragging a bad deck along.
Play skill and knowledge of decks can drastically alter the results of what should be closer matches. I recently won a store T2 tourney with my R/G Stompy Stomp deck, mainly because I knew what my deck is supposed to do and played it, (in my humble opinion), flawlessly. My opponents did not always know what their decks needed to do against me, and did not play flawlessly. I lost exactly one game all night. Had I traded decks with each opponent, in turn, I doubt that R/G Stompy would have triumphed. It was simply that my knowledge of the deck, and the matchups, was much better than theirs.
6) Watch out for inbreeding.
Over time, the gauntlet needs to evolve, in parallel to the real metagame. Most serious MWC and Slide decks are running maindeck Damping Matrix at the moment. Slide is often running Spark Spray, to kill Disciple of the Vault and Goblins. These changes have occurred in the last month. Your gauntlet should probably adjust accordingly – and this may impact your estimation of the relative power of MWC verses Ravager Affinity and Goblin Bidding – and decks like infinite life clerics. However, Ravager Affinity decks can also evolve to take those changes into account. Just last week, I wrote that Ravager Affinity had no use for Green. Now I am trying maindeck Oxidizes in Ravager Affinity, because they answer Damping Matrix, and with the inclusion of Matrix in Slide and MWC, every competitive deck has targets for the Oxidize.
That’s an example of inbreeding. If I, or my group, starts including Abunas in the gauntlet MWC – in effect, making changes to counteract other changes we have made, we have drifted away from the purpose of the gauntlet – that being to reflect the decks we expect to play against in the big tourney. Don’t make those specialized changes to gauntlet decks – and don’t make changes to your deck to counteract those changes – until you see evidence, on the net or otherwise, that a lot of people are actually making those changes as well. Make those changes to your personal deck, instead.
7) Play in store tournaments
If your local store runs tournaments, play in them. If you want to keep your tech deck secret, play gauntlet decks. What do you have to lose, rating points? The entry fee? (If you really are at risk of losing those, then play anyway – you need the practice.)
Local store tourneys provide several advantages (besides supporting the store, which is good for Magic as a whole, so you earn good karma for your altruism.) First, tourneys make you play sideboard games, and everyone skimps on practicing with sideboards. Secondly, they give you some insights into the local metagame – what people are playing, and what people have the cards for. Third, they keep you sharp, since playtesting and casual play allow takebacks and promote other bad habits you need to shed for tournament play. Finally, they bring you into contact with the random new players, and the random bad decks. (This is a good thing.)
8) Keep an eye on the random n00bs.
Sure, there are a lot of bad players with bad decks out there, and playing them can be an embarrassingly easy win. However, they can also teach you not to get cocky – and remind you to keep some sideboard space for answers. I watched a Ravager Affinity player playing against what looked like a bad Blue/Black affinity deck last week. Ravager Affinity got a decent start and began beating. Blue/Black dropped Nefashu (look it up, I’ll wait.) Affinity Clamped a guy, dropped a couple Arcbound creatures and an Ornithopther and setting up for a massive alpha strike the next turn. Blue/Black untapped, dropped Cowardice (!) and beat with Nefashu – bouncing all five of the Affinity player’s creatures.
Next turn, Affinity tried to pull out the game by playing out a Disciple, then an Arcbound Worker and a Ravager – to sacrifice enough artifacts to kill the opponent. Blue/Black responded to the Ravager by playing and tapping Tooth of Chiss-Goria to bounce the Disciple, and when the Affinity player tried to replay the Disciple, Blue/Black played another Tooth and bounced the Ravager. In frustration, at that point, the Affinity player tried to Clamp the Worker for cards – but targeting the Worker with the Clamp bounced it back to his hand.
Too funny – but a good lesson. The Affinity player had absolutely no way of dealing with Cowardice, once it hit the table. Moreover, since the Affinity player was not playing Shrapnel Blast (dumb), or Lightning Greaves, he had no way to outrace Nefashu.
You should not lose Constructed games to bad draft cards – but you can if you focus too narrowly on your internal playtesting, and let your gauntlet become inbred.
Okay, that’s it, except to reiterate one last lesson: turn the browser off and start actually playtesting! You learn to play the game by playing the game, not by reading articles.
* Why major tournaments only? Many decks can get lucky and win a couple of rounds easily. Even bad decks. Larger tourneys, with many rounds of Swiss and a top 8, tend to eliminate lucky breaks. If a deck makes it through seven rounds of swiss and wins in the T8, it is probably good, not just lucky.
** People often call powerful but over-costed cards broken. They usually are too expensive to see play. Look at the hundred odd cards on the banned and restricted lists – aside from Dream Halls and Yawgmoth’s Bargain, what costs more than 4 mana? High mana costs are a huge drawback.
*** Are you sure Cowardice won’t see play? How about Death Pits of Rath? Maybe I should have used Mudhole as an example, or Sorrow’s Path. I am quite certain no one will play Sorrow’s Path in this year’s Regionals.
**** Yes, but I am going to gloat over writing that Blue may have no place in Ravager Affinity, getting flamed in the forums about not running Thoughtcast, then seeing two-thirds of the top 8 affinity decks in European Regionals skip Blue completely – and some of those with Blue just ran Mana Leak, but not Thoughtcast. Gloat, gloat.*****
***** If this continues, I might think I actually know what I’m talking about. I don’t, really. I just write like I do.