Latest Articles

How to build and structure a team

This is part two of a multi-part series covering all aspects of what it means to have, form, and manage a team in Hex. The first part covered the theoretical reasons why teams are inevitable and a necessity in Hex which you can find here. This is the second part that covers building a team while part three will cover maintenance of a team.

“‘Just put random people in a skype group and you’re pretty much good to go’ -Zubrin, 2015′” -Kroan pretending to quote my future article, 2015

The formation of a Hex team can be the vital step between a group of players who are doing well in draft and 8-person constructed tournaments and moving towards consistent success in larger tournaments. I have previously covered theoretically why teams are inevitable and several players, who are not in teams, remain interested in being part of a collective researching, testing, and competing effort. For many players, their short-term goals are to be picked up by a team and be a natural inclusion in what they do. However, teams have some natural limits in their membership; if they get too big, they start risking the inherent collective action problem I discussed before; also, the more players you have on a team, the more likely you will face each other in competition, which is not desirable except in a few marginal cases. Despite teams filling up quickly, there are always teams looking for a few new members, but there might be reasons that these teams are not for you: You could be much better than most of the team members, you may be much worse, you may not have the skill set for the position that the team is trying to fulfill (teams do not just pick out the best players), or you may just not mesh with the teams that have spots open currently. As such, this forces many players into a predicament: they can remain teamless, they can join a sub-optimal team, or they can start their own.

This post is for those thinking about starting their own teams. Creating and maintaining a team is not easy or costless; it requires effort and time to be successful. If you have prior experience in creating and leading a group, then it may be easier to transfer those skills to Hex; if not, then you may have quite a bit to learn. In my experience, I tend to take up the banner of leadership because others are either incapable or unwilling to do so. Ideally, I would rather not organize individuals or serve in leadership positions as it is a time- and resource-consuming process, but I suffer from being competent and, often, I am the person who is the least unwilling to make things happen (it is what lead to me creating and leading a raiding WoW guild back in Burning Crusade that continues to persist today). You very well may be in a similar position.

To be clear: for some of you, the best team that you may want to be in is the one that you create; you do not need good friends to create one, just a willingness to work with others to help bring your game to the next level. This article may help you in determining if team-building is right for you.

Initial Formation

Historically, the earliest trading card game teams form as a natural extension of playing the game locally. For some groups, it starts because friends at school played the game, they played enough together, and they started going to tournaments together and getting more serious about the game. For others, they may have gone to their hobby shop every week and played against the crowd leading to collaboration between the various players. These teams often did not form in full awareness that they were creating a team, but it became more obvious as they recruited some members, selected who they thought would be good additions to a play group, and avoided who would not be.

In the contemporary period of competitive gaming, some team formation still happens in the traditional way where a local hobby group slowly becomes more competitive and eventually becomes a regional, national, or globally competitive team; however, far more teams are built intentionally by one or two individuals seeking to create an optimal testing group. Sometimes, these teams come from established commercial avenues and are extensions of an already existing infrastructure. In the CCG Hearthstone, several groups from other games, such as League of Legends, have teams in the card game as a way to maintain competitive dominance in several online arenas. While we do not have commercial teams in Hex quite yet, they are inevitable as the game grows.

While you are not part of a gaming empire, you certainly want to create a team to compete. You may have some individuals in mind that you have been playing with who are also interested in getting a testing group going or you may be starting with no strong leads, but have hit the proving grounds and streams to find other like-minded players. That is great; find a select few people and build from there. An ideal place to start is with a handful of people to form a core of the team and expand outwards. Many of you who are reading this, and are currently not in a team, likely have a few people they play with on a regular or semi-regular basis. Those people you play with probably have a few people they play with as well. Generally, we can call this the Loose Network model of testing. Each individual has some level of trust with another individual and they share some of their technology and research with the other person; however, they may not share everything they know and any information they share may not be confidential. As such, there is a disincentive towards true collaboration and the network may not facilitate individuals reaching their full potential as they do not want to be too open about what they are doing. If I share my amazing decklist with friend B, friend B may share it with their testing partners C and D. Soon, my secret tech becomes public tech and I am worse off as a competitive player.

To make the group, you basically want to have a loose network and try to close off the interconnections that go outside a specific subset of players. The people you include from the network should be individuals that you trust and form the core of your group. It helps to have people you can trust to help build and recruit for the team so you do not have to do all the work. From this core, you need to constitute what the team does and how it makes decisions.

Before I get to team structures, I should make a few suggestions. First, team members should be expected to share their information freely and fully with the entire team. If they discover a one-card difference between the team’s deck and what they want, they should let the team know that just so there is full information available to the team and it does not appear that people are hiding information to the detriment of the team. It is awkward if the only member of the team to top 8 in a tournament is the one that made a slight tweak to the deck and did not tell anyone else.

In addition to this, the team should also decide what decks are public knowledge and which ones are team secrets. Especially at the beginning of the team’s life, you may not have enough people to actively test whenever someone wants to test. As such, you will inevitably have to test with people outside of the group. Decks that are public knowledge are the decks team members can use in the proving grounds or against friends outside of the team. Team secret decks are those that the team actively works on and they think have a high probability of doing well in the next tournament. Sometimes, this may just be a partial shift of a known archetype, other times, it may be a complete list. Clear designation can help prevent inadvertent leaks as well as encouraging the team to know the boundaries of the group.

Governance and Decision-making

Once you have the initial group of individuals, you need to decide how the group will make future decisions; you are no longer in the loose network model of testing and you have a closed group. Deciding decision-making rules is fundamental to the longevity of the group and resolving future disputes as it defines how much stake each member has in the group, the efficiency of the team, and the flow of information. I will outline three brief models that different teams employ and have various benefits and costs: the executive model, the democratic model, and the consensus model.

The Executive Model

The executive model is the most authoritarian version of group decision-making and is, perhaps, the one we are most used to when we think about our daily interactions with hierarchy in terms of our jobs or interactions with some political and social structures. The biggest benefits to this model of decision-making are its speed and efficiency: Either a single person, or a small subset of people, make decisions on behalf of the group. This includes issues such as team-versus-team matchups in major tournaments (does one person drop, do we play it out, or do we intentionally draw?) or how and when to recruit more members of the group. This type of structure can be pretty important if a team receives sponsorship and an individual needs to negotiate on behalf of the team and lay out the rules of the team’s behavior. Most players, unless joining an established team, will be hesitant to voluntarily form such a group due to the structure inherently disempowering them in helping decide the group’s fate. If you opt for this model, it is important for the leader to have some level of constraints on their ability to guide the team; making them bound to the group rules or limiting their ability to remove members (unless supported by other members) can make individuals more comfortable with this team model. We know with research in political systems that autocratic decision-making structures suffer from several flaws; two of these that could disrupt a team include information issues and time-inconsistent preferences. In authoritarian structures, it is hard for leaders to come by good information to make decisions as those who provide the information may have their fate tied to the information they provide. If they fear being removed from the group or face some kind of internal sanction at the whim of the leader, they are going to be more likely to present information that helps them as opposed to information that leads to the best possible decision. This, game theoretically, is the dictator’s dilemma and it is quite troublesome for effective governance. A clear example of this in recent history is the Iraqi information minister during the US invasion of Iraq in 2003 and his assurance that the Iraqi government was winning the ground war against US forces. Distributed power systems are more effective at providing reliable information as those who provide information (positive or negative) are less likely to receive retribution of it; this is why we get interesting historical trends such as democracies generally not experiencing famines; famines generally only occur in non-democratic systems. The other issue is time-inconsistent preferences. In economics, there are a series of problems where an actor will make one decision or declaration at an early time period to try to achieve the best outcome but, later, will reverse that decision given their preferences in the future. A dry example of this is the issue of political control over monetary policy: if monetary policy is controlled by those in political control of the government, those actors will promise not to manipulate their currency early on to encourage investors to invest and buyers and sellers to be confident in the currency. However, if that group is dependent upon elections to remain in power, they will be tempted later to increase the supply of money (expansionary monetary policy) to decrease unemployment and increase their overall approval and chances in the next election. To overcome this issue, many countries use an independent central bank so the political parties cannot be tempted to use electoral pressures to influence their decisions about monetary policy. A compelling literary example of this is in the Odyssey where Odysseus is warned about the dangers of the Sirens, a group of women whose beautiful singing convinces sailors to jump to their deaths to reach them and crash their ships along the rocky coast of the Sirens’ island. Odysseus tells his men to put wax in their ears so they will not hear the song and he orders them to tie him to the mast; in addition to this, he tells them to ignore his future pleas to be set free. Every time he asks to be set free, he orders them to tie the ropes tighter to ensure that he does not escape. He knows that his preferences are time-inconsistent: He presently does not want to lured to his death by the Sirens, but expects that in the future he will be unwilling to commit to what he wants now. As such, he builds a institution, a set of rules for his men to follow such that his time-inconsistency will not betray him or his men. The executive needs to do the same thing. The executive leader will certainly make promises to be fair and just in any decisions that they make and will not be swayed by bias or emotion, but there is the temptation to still do so when decisions become heated or tough. Simply kicking a member that ardently disagrees with you may be very tempting. As such, an executive should follow Odysseus and try to build rules that limit their power to prevent them from abusing it, making the team unhappy, and losing what they have built.

The Democratic Model

There are several different ways in which you can have democratic decision making and each of them have their own flaws. However, for the sake of brevity, we will assume a simple majority rules type decision making system. This is beneficial as it treats all the members of the group as equals and gives them a voice in the direction of the team. This is the type of team I expect to see most often when started by community members just getting together to form a competitive group.

The biggest word of caution I have about this model is that there is a severe risk of a tyrannical majority; that is, a subgroup that acts as a voting block against another, smaller subgroup within the team. This will lead to a system that is closer to the executive model and will ensure that some people feel left out of the group and be tempted to abandon the group. So, if you use this form of decision making, try to make safeguards against one group dominating another automatically and be careful that the group does not develop cliques within it.

The Consensus Model

This is my preferred version of how teams ought to run. Instead of every member getting a vote, every members gets a veto. As such, any change in the way the group operates can be stopped by a single person. This is potentially the slowest of the three structure types in terms of decision making as a single person can paralyze the progress of the group. However, it makes sure that everyone is on board with any decision, and if they are not, then you either need to convert them or find a different route to achieve what you want as a group. This makes the group the most inclusive and ensures everyone’s view, especially contrarian views, get seen.

This model is severely hampered by diverse groups. As such, a consensus group usually works when they are small teams and the members of the team are very like-minded. If the group is mostly comprised of real-life friends, this can be an optimal system. The team I am in has 15 formal members from diverse political and geographic backgrounds, but we ultimately still make decisions using this model. Of note, if a group gets too big, it may be impossible to make decisions and the group may have to revisit its decision-making procedures and opt for a high-threshold democracy (such as 80% or 67% of the group being on board for any decision).

Conclusion

This introductory guide is a start to help aid individuals in creating a formal, functioning team of players that are looking to become more successful at the game. I’ll end here at about the point where you have enough people to form a team and are really cutting your teeth on the advanced competitive scene. As such, I will pick up on part three of the series assuming you have roughly 5-6 players and are trying to figure out where to build the team from there and the types of players you should be looking for to increase your chances of long term success as well as a few ways in which you can formalize the recruitment process (i.e. should you have an application process? The answer is: maybe, but probably!). Let me know if there are any questions you would like me to address in subsequent sections on this topic.

Michael Allen is a competitive HexTCG player, co-host of the 2 Turns Ahead podcast, and founder and moderator of the Hex Subreddit.

2 Comments on How to build and structure a team

  1. I think it is interesting to point out and discuss that in a TCG team or guild have a very different dynamics than in MMOs and team games. In Hex (currently and the future) the competitive play is one on one, and you don’t need to be on a team to participate. Inside a game, being on a team or not doesn’t matter in the way it does in MMOs and team games.

    In MMOs, you need other people to be able to accomplish things and access content that a single player can not. In many team games, you are required to have X amount of players on each team to start a game. TCGs generally do not have built in content or mechanics that strongly encourages a player to be part of a team, and the benefits are not as spelled out as it is in some other game genres. And at this time, Hex Beta does not have any system that is built in to help form teams or provide functionality for teams to use (except mailing cards to other team mates).

    • Yeah, I agree and had that in mind; this was part of the goal of the first article in the series was to discuss the advantages of teams and why they will certainly exist despite not be as seemingly necessary due to the game mechanics.

Leave a Reply

%d bloggers like this: