Holiday Magic- Commander Tournament Design

04 Jan



This article is about building a competitive Magic: The Gathering commander tournament that provides both an arena for structured contest and a space for promoting the values of the commander variant of magic- multiplayer, political and social play with a heavy emphasis on individuality, creativity and theme. To sum up the proposal:

  • A tournament format based on four player groups, where winners and losers are paired off in standard swiss style using match results
  • An additional layer of sportsmanship rankings ensuring players remain cordial and play to the spirit of the format
  • A final layer of personalised points awarded by opponents for particular accomplishments within each round- clever political play, creative deckbuilding, absurdly complex combo execution etc.
  • A prizing system using prizes with low monetary value but high specific value to either commander players in general or the specific prizewinner, to avoid ‘value’ players.


It’s been a while, things have been busy and the last major article took a lot out of me to write, but I’m back and this year I plan to be writing on the regular since I’ve moved houses and will now have a good friend and tireless encourager leaning over my shoulder and telling me to stop browsing reddit and start making the world of games theory a better place.

On my way home for Christmas I listened to a recent episode of Monday Night Magic. In it the hosts discuss the experience of an experimental commander tournament.

To briefly cover this for non MTG players, commander (sometimes known as EDH) is a social format of MTG, designed to facilitate multiplayer games rather than 1v1 matches and to encourage creativity and individuality compared with competitive magic where players tend to build their decks based on strict archetypes. It does this by giving players access to more or less the entire list of magic cards ever printed, but requires a hundred cards in a deck (rather than sixty as normal) and limits players to only one instance of each card (rather than four as normal). Consequently, each game is a lot more variable and less predictable, creating interesting and enjoyably chaotic situations, particularly when multiple people get involved.

Inevitably, however, people have tried to repurpose the format for competitions, albeit usually with the goal of a fun, social experience rather than an eye-bleeding test of skill and derring-do. The tournament discussed was one such attempt and sadly it was somewhat ham-handed. I won’t go into detail here, you can listen to the podcast for more info, but suffice to say I thought it would be an interesting experiment to put my money with my mouth is and come up with a proposal for a commander tournament myself, doing so methodically to show some of the aspects of social design that I work with and I think are an area that most designers overlook.

The Basic Problems:

To start with we have to overcome a common social perception- serious competition is not ‘fun’. The two exist on opposite ends of a spectrum. Consequently, if one wishes to make a ‘serious’ tournament, one must do so at the expense of making said tournament optimally entertaining for the average magic player. To some extent this is true- part of managing any ‘serious’ competition involves making sure your rules are established in advance, communicated and kept to rigorously, to avoid the event falling into disrepute. Inevitably some tedious bureaucracy develops from this, but it can be minimized and a charismatic TO (tournament organiser) always helps.

However, a large part of this perception stems from a simple trend in competitive play. In competition, every attempt is generally made to avoid spontaneity, random events and extraneous variables. Interestingly, these are often the things deliberately injected into ‘fun’ events to make them more ‘fun’. From my observations, I suggest that the way they accomplish this is to make the event outcomes less easy to predict, preventing players establishing a pre-conceived model of what’s going to happen (I’ve lost my first three games, I have no chance). The problem is, this kind of addition often makes who goes home in what position a bit too much luck based for the average competitive player, so this is one of our first objectives. How can we balance on the one hand a legitimate contest with on the other hand the kind of spontaneity that keeps players involved and on the ball? 

The second major issue is that for a game like Magic, competitions are largely motivated by prizes. Having a prized event means players will come wanting to win and feel disappointed if they don’t succeed. As was discussed on the podcast, an event whose parameters are not clearly outlined before anyone puts their money down is likely to get a lot of ill will when it comes to prizes, as might an event whose prizing structure is seen to be potentially biased away from skill or too random. In particular, when an event has prizes that have an easily evaluated financial value then the tournament will attract players whose entire tournament experience is predicated on an attempt at financial gain. So how do we create an incentive model that rewards the best players on the day, but does not encourage behaviour out of place at a primarily social event?

Finally, an issue near to my heart. Part of the purpose of normal tournaments is to measure your skill against an unknown field. Your opponent is simply your opponent, rather than Tom or Dick or Harry. Skilled competitive players learn to be able to adopt this mindset even when playing against familiar faces, but this is cross purposes of a commander tournament where social and political interaction is a core of the game. Commander games are exceptionally good at developing a story of alliances and betrayals, vendettas and grudges and so on. Where in conventional competitions a player is encouraged to simply ‘play the game’, optimizing their play based on the probability of what could happen based on the information they have about the cards, rather than their knowledge of the opponent, here we wish to encourage players to interact with their opponents and get to know them, both for social reasons and to raise the level of play in the areas that make commander so compelling. So how do we encourage interpersonal interaction and political play while ensuring players make friends rather than enemies?



Building the tournament structure

Let’s begin with a basic framework. I think the basic shell of the tournament model used by the tournament that sparked all this is quite solid. Players are divided into ‘pods’ of four and play a multiplayer free-for-all in that pod. The first player to be taken out gets a point, the second gets two and so on with the last player standing receiving four points. From there it’s easy to have further rounds structured swiss style, with winners grouping off, losers grouping off and so on. It’s important that these pairings be determined purely by match results so let’s call these basic points MP or matchpoints. Since we may score players in other ways, say for sportsmanship, we don’t want a player with a weak deck but who is really nice being paired off against the terrifying combo-master.

In the tournament discussed, players were additionally given a sheet at the end of each round where they had to mark their favourite opponent and least favourite opponent in the group. These marks would grant either positive or negative points, in an amount potentially greater than the match points.

Let’s pause for a moment here because this is an example of really stupid event design.  By forcing a player to rank their opponents systematically, one forces them to not only pick out someone as better than the others (which isn’t so bad) but to pick out someone who is worse (which is). In most cases, a player’s opponents will all be quite cordial and fun to play with and they’d stand up happy with a fun game, regardless of how the cards turned. If at the end they have to think about which player to give a black mark, they’ll spend time pondering over all the little issues trying to be fair, which is going to make them see otherwise quite normal, decent opponents in a very critical light. Quite the opposite of what is wanted at such an event.

Instead, let’s use a similar system: at the end of each round players get the sheet and instead it has some checkboxes for each opponent:

  • Was this player enjoyable to play with?
  • Was this player’s deck creative and/or thematic?
  • In your opinion did this player keep to the spirit of commander?

For each of these checkboxes a player has three options- they can place a tick, which will give that player a bonus point (I’ll call these SPs to contrast with MPs), they can leave it blank which will be neutral or they can place a cross, which will subtract one point. In the case of both tick and cross, a player can only award one per round per category. These represent the opportunity to reward or punish exceptional behaviour, rather than a necessary choice. It should be emphasised to players before scoring that a blank sheet is the standard case.

So here we have a core points framework-

MPs are based on match results, are public and determine seeding.

SPs are based on creativity and conduct and are kept private.

By itself, this goes a little way towards incentivising good behaviour and creativity, but doesn’t really do much for the other issues we need to resolve. So let’s add some spice.

One of the main tensions that would occur if we purely use this system in the context of a normal tournament with normal prize structures is that people would feel like SPs are just another metric to attack in order to optimize their score. Consequently they would be shy of the political play that makes commander so fun, worried that if they backstab someone that person will mark them down and they’ll lose out on points. We need to counteract that, which luckily isn’t that difficult. The first step is formally acknowledging that such behaviour is acceptable, normal and not to be penalized. The simplest way to do this is by giving that behaviour a recognized place and incentivising it slightly. Let’s say something like this:

In addition to the basic SP checkboxes, a player has the option of awarding an opponent ‘medals’. Medals are worth a point and are from a list of pre-determined titles like

The Nicol Bolas award for most magnificent bastard

The Richard Nixon award for political misplay

And so on. I’m not going to go too deep on this concept here, but suffice to say you can have a list of such awards going anywhere from “The one with nothingness award for most absurd combo execution” to the “Hello Kitty award for best deck presentation” depending on your particular goals for the tournament.

This creates a framework which shows that certain behaviours normally to be frowned on are acceptable within the tournament framework. It’s not entirely a novel concept, I’ve seen this executed both in games (particularly Steve Jackson games like Illuminati and Munchkin) and in tournament frameworks (primarily Warhammer Fantasy Battle tournaments). In both cases, formal acknowledgement of such behaviour forewarns players to be aware of it, lessening the sense of betrayal when they are, for example, betrayed. This also opens up an alternate prizing front where players can be rewarded for receiving the most awards of a certain medal, or a certain category of medals. If aware of this possibility, players may seek alternate ‘win-conditions’ while participating- aiming to be the best sportsman, most conniving manipulator or to have the coolest deck theme.

(To successfully implement the medals system it might be necessary to have some kind of structural restrictions such as a maximum of four medals per round, you get a point for every two medals you give out. This incentivises players to use the system and prevents the tension wherein optimisers don’t want to give their opponents bonus points, so don’t hand out any medals. With this caveat that strategy will lose them points in the long term. There’s no perfect way of solving it- a player might still try and equally distribute the points, perhaps giving more to the person with lowest MP to reduce competition for the final standings, but at least this system gets the optimizer considering whom to give which medal to)

Now we have a scoring structure with three aspects- match scores, conduct and creative play, with some overlap in the latter two scoring systems. I think this model does a fairly good job of partly answering all three of the problems noted earlier in the article

  • By widening the axes on which prizes are awarded and accomplishments are tracked, we increase player engagement by giving them additional goals to aim for when the primary ones are out of contention
  • By adding a formal behavioural check we cause players to be more aware of their conduct and consequently more capable of entering into the desired attitude for the event
  • By formalizing and incentivizing behaviour suited to the commander game but out of place elsewhere, we support players who wish to act in contextually appropriate ways and help avoid them being penalized by players who would otherwise be upset at those actions.

Now, a final question that must be addressed is prizing. Normally tournament prizing in magic is either cash or ‘product’- booster packs of random cards that have a fairly stable market value/demand and consequently are just as good as cash with a little more effort. This is problematic for this kind of tournament, since such prizing will attract players for the wrong reasons- this is definitely a tournament to be played for the tournament, not the prizes. However, a good TO wants to offer their players something exciting to look forward to for many reasons- to build hype and attendance, to keep motivation up on the day and to make the event feel more worthwhile afterwards. Consequently, prizes must be hard to convert into cash, but have high value for the kind of player interested in commander. Off the top of my head, here are some examples of how I might prize such a tournament

1)      A card alter commission of the player’s commander by a respected card alterer

2)      A commander specific deckbox, perhaps a custom cut and engraved wooden or metal one

3)      A big binder full of useful commander deck building cards with moderate-low market value.

4)      An original piece of art from one of the commander card sets

5)      A chance to play with some local community figures- podcasters, pro players, judges or similar.

6)      A prebuilt, themed commander deck full of tricky to find or quirky cards- something cool like a dinosaurs deck or a zombies deck

7)      A pick from a set of useful-for-commander medium-to-high value singles for minor prizes. It’s important that these not be the object of major competition, so a player should be eligible for no more than one, and none should be of monetary value much greater than the tournament entrance fee.

By choosing from these, or similar, you ensure players have something unique and exciting to compete for, but that those prizes are not too attractive to folks in it for the money. Alternatively, once things get rolling, you can also simply hand out certificates and trophies, since prestige is a heady gift in its own right.


So there you have it. My attempt to nut out the framework for a potentially successful commander tournament. If you’re a MtG TO or just an avid player who has comments, let me know and I’d love to chat about actually implementing such a tournament or the no doubt many problems it has.

Leave a comment

Posted by on January 4, 2014 in Uncategorized


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: