Metagaming the Draft — An Open Letter to Wizards

Wizards has databases full of wonderful information from thousands of Limited games. In the bottomless reservoir that is MTGO they hold draft pick orders, Sealed pools, maindecks, sideboarded decks, and tournament results. That just makes the little geek inside of me twitch with desire. Okay, he’s not that little… okay, he’s not inside of me… okay, I am that geek, just let me be.

Here is my problem: I’m trying to learn how to draft properly before Pro Tour Prague, and the information that is available online just doesn’t cut it. Sure, Eisel likes to draft Blue in RRG and so does Soh, but how does that help me? What if they don’t write? Are all the other colors so terrible that the only lands that get published are the Islands? I want to learn from others, but it is not easy with a format that is so flexible. There are countless players who draft better than me… what are they doing different?

Limited is my favorite type of Magic. It is less repetitive than setting Deck A versus Deck B thirty times to learn the key cards and how the match up unfolds. We play a game, and all of us started doing it because it was fun. Punking your best friend with a savage top deck that blows the game wide open is a pleasure so great that it should be outlawed. If we know that said friend is playing Owling Mine, then building up a Zoo variant to help deliver those precious twenty points of damage is priceless. The look on his face when you drop critter after critter after body-mangling critter, watching him squirm as the little ones come in for two, and four, and eight, and… Dead, he is dead! I guess you could say I’m Spikeish in my play.

So how does one go about obtaining this important information for a Limited game? I want a way to defeat my neighbor while drafting and earn acclaim all over da ‘hood. How is a draft different than a regular Standard tournament? Constructed Magic can boil down to three elements:

  • Knowing what your opponent is playing. Like the above example, if I know my friend Tom loves his WWr, then I can modify my deck accordingly. I can even choose to play a completely different deck that will drive slivers of pain all across his little creatures… flying lions, legendary dogs, hawks! For Pete’s sake, he attacks with hawks! Eat my Wrath (of God). This is the art of metagaming.
  • Knowing what you are playing. This is simple; if you want to know your decklist, all you have to do is look at your deck. Make sure that it can beat what you expect your opponent to be playing.
  • Playing better than your opponent. This is important. Hours of playtesting usually go in at this point. I would tell you more on how to do this, but it is my weakest area. My Magic playing skills are like my basketball playing skills: real good, except for the fact that I can’t rebound, pass, shoot, play D, run fast, or jump high. I love to play the game, though. I’ll be that guy that has just lost six games in a row and is still smiling. Of course, there are a couple of decapitated bodies lying next to me, but nobody is perfect.

In a Limited environment, there is no way to obtain the information for either what you will be playing or what your opponent is holding. Picture this winning strategy: “For my next Sealed deck I plan to open a Meloku. What’s that you say? Ravnica-Guildpact-Guildpact? Make it a Jitte.” So what kind of information do you want to have before you open those magical packs of Magical cards? Let’s boil down Limited to a few over-simplified elements, for the sake of discussion:

  • Getting good cards. You can draft them or you can open them, but it is a self-evident truism that Glare of the Subdual is better than Zephyr Spirit. By the way, how bad is that card? If I am forced to 15th pick another of these craps-on-a-stick I will drive a nail through my computer screen.
  • Building a good deck. Your 63 card five-colored Sealed contraption might be cute, but it will be dead. This part of the Limited game is very skill-intensive, and a good builder can get by on having superior and consistent decks… that is, until he plays someone that has a clue as to how to play.
  • Playing better than you opponent. I’m just going to copy paste the information above. Not only do I have nothing to add, but this will also increase my word count. If only life was always so easy. “This is important. Hours on end of playtesting usually go in at this point. I would tell you more on how to do this, but it is my weakest area. My magic playing skills are like my basketball playing skills: real good except for the fact that I can’t rebound, pass, shoot, play D, run fast or jump high. I love to play the game, though. I’ll be that guy that has just lost six games in a row and is still smiling. Of course, there are a couple of decapitated bodies lying next to me but nobody is perfect.” Yup, that pushed me above 850. That was a good 100 repeated words. So pretty, so nice.

This whole article will focus on the first two elements of Limited play: how to draft better cards, and how to build a better deck. Now, I can’t tell you how to do either of these since I’m a n00b fanboy, but I can let you know what information I would love to have in order to be able to acquire these skills.

*Click Back-button* What a rip-off! So he will tell us how to do something… only he has no clue as to how to do it?

I have just lost three out of the five readers that got this far. Well, this is for you, my faithful couple of fans. I will keep writing because I know you would like to see the end of the article. By the way, hi mom! That leaves me with just one other reader… you, so please keep reading. It gets lonely to write for one. [I have inserted this editor’s note because I know the good old editor for this here website has to read the entire dribble that flows from my keyboard. Cheers mate. – Pedro]

Wizards has databases full of wonderful information from thousands of Limited games. In the bottomless reservoir that is MTGO they hold draft pick orders, Sealed pools, maindecks, sideboarded decks, and tournament results. That just makes the little geek inside of me twitch with desire. Okay, he’s not that little… okay, he’s not inside of me… okay, I am that geek, just let me be.

If we could get access to all this goodness, what are some questions we might ask? Here are some obvious ones:

1) What is the Average Pick Height for each card, on each booster, for a draft?

If I want to learn how to draft I would like to know the valuations given to each card by the people playing with me. How high are they being picked? At the top of the power curve it is pretty easy to guess. In Kamigawa drafts, Umezawa’s Jitte probably had a Pick Height approaching one. Actually, why not just “one”? Isn’t it always first picked? No, there is always someone that decides to pick the 5/5 guy for nine mana instead of the two-cost artifact. What about Moldervine Cloak in RRG Draft? I’m guessing somewhere around 1.5. It probably gets first picked a large amount of times, but it also gets handed around by people rare-drafting or hoping to send signals. It is in the middle valuations, the slots between 3 and 10, that this list becomes highly relevant. How high are the Karoo lands being picked in RRG? What is the highest picked uncommon? The lowest picked rare?

Why separate the results by booster? Well, in a RRR draft, Glare of the Subdual probably has a lower Average Pick Height in the third booster than in the first. In the first pack it gets taken early, and pushes the player towards playing those colors; in the third one it gets passed on by the dedicated Dimir deck in favor of a Vedalken Entrancer. [Hell, I’d splash it. – Craig.]

This information gets really explosive when you can play around with how you get it. I can see the pick heights, but is this all I want to know? When you do a query to mine information, it is customary to limit the search to a selected range because those results are more interesting than others. “What is the Average Pick Height, on the first pack, for a RRG draft during the first two weeks of April 2006, for players with a rating above 1800 in Limited?” Wow, talk about a learning tool! I just had a little accident from the excitement.

Here are some variables that it would be fun to play with in order to find the Average Pick Height for individual cards:

  • Format: RRG is a different beast than RRR, and card valuations change accordingly.
  • User Rating: What are the big boys picking? Of course, I’m more interested on the card valuations of the experts than the scrubs.
  • Date Range: How has the format evolved? Some cards probably got low-picked at first until people started learning about them. It is a rare thing to see ‘roo lands table anymore, but they went round and round all the time when RRR drafts began.
  • Table Rating: If the table I’m sitting on has five players with a Limited rating below 1600, I pick differently. Same happens with a table full of sharks. Do the valuations change for all the players? How do they respond? I want to know the card valuations of a 1900 player sitting on a table full of 1750s.
  • Queue: Does an 8-4 queue play different than a 4-3-2-2? Are the pick orders that different?

So far, all these queries help us obtain information on how to get a better card through drafting, but that is only a third of the skills you need to become a better Limited player. How we learn about deckbuilding?

2) What is the Average Maindeck Usage percentage for each card?

Having the cards in your pool is only half the battle. Do you utilize them? Do you keep it in the sideboard for use against certain opponents, or do you stick it in to use against everyone? Here we try to learn not only how to get cards… but how to build decks. Again, not all deckbuilders are created equal, and not all situations are the same. Maybe it’s a Sealed deck and the bomb just does not fit into my colors; maybe it’s that $15 rare that has nothing to do with my deck that I picked up in order to pay for the draft… why some cards are available and not used can vary, but it is significant by itself.

What do I want to know about Average Maindeck Usage? What are the questions that go through my head? Why this endless stream of queries that lead nowhere? Must I always ask a question that I will proceed to immediately answer, thus making myself seem like some all-knowing being? Yes.

Anyway, the information for maindeck usage should be sifted through with a fine sieve:

  • Format: Sealed, draft, RRG, RGD, RRR… all these varying situations change the usage of a card. Was it used every time someone opened it? Was that dual in every deck that first-picked it?
  • User Rating: Good players are still good players. I want to see what the top builders are using.
  • Date Range: This is pretty self-explanatory.

3) What is the Average Sideboard Usage percentage for each card?

If the card was stuck on the ‘board, how often was it brought in for games 2 and 3? There may be archetype hosers that get drafted high by good players just in case they face one of these decks. The same conditionals that were explained above (#2) can be used for this stat.

4) What is the Average Game Win percentage for each card?

If the card was in the deck in a game, be it game 1 or games 2 and 3 once sideboarded, how many times did the player win? How much of a “swing” is this card? Here we run into a problem where good players may favor a card, thus giving it a higher win percentage… problem? That is exactly the information we want to look for!

By now you know the drill; I’m now going to list a few variables where it would be interesting to trim down the search spectrum. These all work independently or together.

  • Format: Yada, yada, yada; blah, blah, blah. If you don’t know me by now, you will never never never know me (No you won’t). Simply Red is Simply Fabulous.
  • User Rating: This would work more as a restriction than an actual search field in my head. I want to know how users ranked near me are doing when they play this card; I know great players will win more often, even if for some reason they include One With Nothing in their decks.
  • Opponent Rating: Good beats bad, bad loses like a red headed stepchild to good. Bad sucks, good pwns. I don’t care about 1400 versus 1900 games.
  • Date Range: For my next bit I shall invoke Captain Obvious.

So far, all we have is individual card information, which would help us value certain cards for drafting or know whether to push them for deckbuilding in Sealed… but what about the actual building of a deck? How many lands should we use? Kamigawa ran at a higher land count than Ravnica… but how much higher?

5) What is the Average Land Count for decks?

This would include basic lands as well as nonbasic. It may not seem like it, but of the questions I have asked this is the hardest to answer form a programmer’s point of view. There are many definitions that have to be hammered out, like what constitutes a nonbasic land? Can I group some nonbasic lands together (like the Ravnica Karoo)? How much wood could a woodchuck chuck if a woodchuck could chuck wood?

It is important to keep an eye on the ultimate goal for this stat: If the average winning deck playing RRG draft in March was carrying five islands, six mountains, three Karoos, with a total land count of sixteen, then that is pertinent information.

The same filters that I have been discussing all along apply here, with the inclusion of game win percentage:

  • Format: Since I have written this spot in every aside, I’m just going to use this space to send a shout out to the Costa Rican Magic Players that gave me feedback on my first article. Most of them post on the forums for Sologood. I’d also like to use this space to officially challenge-beg from the O-gaming crowd… couldn’t we just please run a few more drafts? Please? And none of that Mirrodin block crap that Anthson likes… RRG, what real men play. Once we get Dissension sorted I will officially kiss any ass for a draft. Come on guys! Where else can we practice? There are a couple of us qualified for the PT and we need to see the new interactions so we can honorably eat it when we sit in front of Finkel.
  • User Rating: “a mi major amigo Carlos Pal”. He is not really my best friend, but he wanted to be in the next article I wrote. He also looks like Joe Pesci, so it’s almost like being a celebrity.
  • Date Range: I would like to thank my father for getting my mother drunk that night. Without the aid of Jose Cuervo I wouldn’t be here to tell the tale.
  • Game Win Count: A deck that won six games in an 8-man single elimination draft queue is, by definition, the winner of the draft. This would allow us to quickly narrow results to see the decks we are interested in… the winning decks.

We have established the information that is of interest. I’d love to spend endless hours reviewing the win percentages of various cards, and then publish articles that make it seem like I know what I’m talking about. The question remaining is: how do I want it? I want my cake and I want to eat it too. No doubt about it, I want and want and want and I will get what I want.

The optimal solution would be to have a specialized tool built that would only give me access to all these goodies. I could then pwn the drafts and be a stud at night. Since Wizards is not my personal Make-a-Wish Foundation, I can accede to make the info publicly available. Much like Gatherer has made card searching easier, especially since it includes card rulings right there with the card info, this should be an easy-to-use interface. You select the kind of query you want to run, it asks you to set the variables associated with the query, and you get a sortable list that you can then spend endless hours studying until your family calls you and asks you to please join them in the sunlight… but they don’t know! They don’t understand what it takes to master the intricacies of draft.

This article is basically an open letter to Wizards… please, please, please give us access to this info. Let us study Limited the way we study Constructed. Give us the tools to learn from the community. Sure, you can have one of your guys do these searches and publish one article a week, or you can let the little people give it a swing and get countless articles published. Just picture what the obsessive Japanese players would do with this!

Pedro Alvarado
[email protected]
zoropoteame all over the world
expendibleone on MODO