Mechanics: Drafting

22 Mar

As with all my in-depth articles, there is a brief summary to begin.


Drafting is a widespread mechanic throughout various game genres and mediums. It’s a very powerful mechanic and has a lot of internal diversity and complexity, so it’s worth looking at in detail. Drafting is ultimately a resource distribution mechanic that splits a pool of distinct resources as evenly as possible between players. In general, drafting is more effective the more complex and unique each resource ‘unit’ in the draft is. There’s no point in drafting gold or lumber, drafting territories on a map or heroes for a battle is a more reasonable use for the mechanic. It allows for a high level of engagement, strategy , emergent gameplay and, in the right scenario, can even form the core of a game of its own right. In this article I’ll look at some of the various properties a draft can possess.

  • It can be clear if all the available resources are seen by all players, or obscured if not
  • It can be open if players can see the choices of their opponents, partly closed if they can deduce/guess them or fully closed if they have no way of knowing.
  • It can be complete if all resources in the pool are drafted to players (though they need not all be used), or incomplete if only a fraction of the available resources are drafted
  • It can be unrestricted if players have access to the total pool of resources to choose from at any point, or restricted if not.
  • It can be a common pool draft if players choose resources from a shared pool or a private pool draft if players have their own unique pool to choose from.

To see how combinations of these properties work out, this article looks at four different games with drafting mechanics- a sports team draft (clear, open, complete, unrestricted, common pool), an MTG booster draft (obscured, partly closed, complete, restricted, common pool), the game 7 Wonders (obscured, open, incomplete, restricted, common pool) and Moba videogame drafts (clear, open, incomplete, restricted, common pool)


Fundamentally, Drafting as a game mechanic is taking a pool of items and then dividing them up between a group of players by each player taking an item in turn. Naturally, the most desirable items will be chosen first, the least desirable chosen last. Complexities are added when, for example, choosing an item changes the values of the remaining items, if players can only see a fraction of the resource pool at a time or even if players draw from separate pools, basing what they choose on what other players choose.

Drafting ensures that the available pool of resources is distributed as evenly as possible between players. It’s most useful when the resources themselves are complex and synergistic, rather than having simple, independent values. Take for example an ‘I pick- you pick’ draft of one hundred tokens between two people, each value from 1-100 being represented once. Not only will the results always be the same (assuming both players seek to maximize their score), but the to pick first will always end up 50 points ahead. Therefore, for a draft to be useful the drafted resources must be complex and have their values depend somewhat upon the properties of other resources in the pool.


While I’m sure someone will be able to dredge up some antiquarian example of this process, the most ready origin of this mechanic is how teams are picked for a social game of team sports. Two captains are chosen and take turns to select players from a common pool. Both captains can see the whole pool at all times, what their opponent is picking and the whole resource pool is picked by one or other player. In my endless search to provide terminology for everything this gives a few opportunities

  • It is a clear draft. If the total pool of resources is somehow obscured (for example, only a few of the resources are available to see and pick from at any one time) it is an obscured draft
  • It is an open draft. A draft where the choices of the opponent are partly or fully unknown is a partly or fully closed draft
  • It is a complete draft. A draft where only some of the potential resource pool is chosen is an incomplete draft.
  • It is an unrestricted draft. A draft where only part of the overall resource pool can be chosen at any one time is a restricted draft.

As both players can see the whole pool and the selections of opponents are immediately and fully known, an open draft is the most suited of all drafts to pre-consideration and drafting strategy. Since, in an ideal case, both players will always get the optimal resources possible in the circumstance, it is important that there is a level of ambiguity within the resources themselves for this kind of draft. For this example, the current form of a chosen player, the inherent complexity of human capabilities and synergies between players serve to provide this ambiguity. An open draft can be considered similar to any other perfect information strategy game, chess being the most relevant example. players can make a branching tree of choices, all of which are knowable, but the complexities of the emergent scenario prevent the ‘solving’ of the game. This results in that most wonderful of all games- one that is easy to learn but difficult to master.


To compare, let’s look at Magic, the Gathering‘s booster draft. In the standard format, a number of players are each given a set of booster packs of random cards. They open one, pick a card of their choice to keep without showing it and pass the remainder off to their neighbour. Then they pick a card from the new booster they have been passed and so on, going around in a circle. An MTG draft is an obscured, partly closed, complete draft. If there are only two players, it is essentially an open draft as the resource pool is almost constantly visible and the choices of the opponent plain. If there are more than 14 players (the amount of useful cards per pack), the draft is fully closed as there is no way of guessing which opponents picked which cards. (There’s a variant of the booster draft known as a ‘rochester draft’ in which one booster at a time is laid face up on the table and players take turns taking cards from it. This would be a clear, open, complete draft just like the sports draft). While MTG cards are less ambiguous in their potential than human players, the booster draft throws up the complication that the overall resource pool is obscured and what can be picked at any one time is restricted. In addition, MTG cards have very clear synergistic relationships with one another, so the power of a card is always contingent on other cards. If a card is powerful only with another rare, highly sought card to support it, its power in a booster draft is mitigated due to the obscured resource pool-there’s a low chance you’ll have a chance to get both.

However, it still provides a player an interesting choice- if for example they open a pack and see this card first, they can take it before any other player sees it in the draft. The player might then get lucky and open another pack to find the needed card inside, or another player might pass on the needed card, not needing it themselves and not knowing the synergy is a potential threat. Both of these are quite low probabilities, but may be deemed worth the risk if the power of success is strong enough.

Another effect of the synergy relationships in MTG deck building is that the first few cards chosen by a player powerfully alter the value of further cards chosen. A player who picks a first card of a certain colour is going to more highly value cards of that colour, or specific cards which complement their first pick for their next picks. This creates a snowballing effect where, as more picks are made, values alter significantly on cards remaining in the pool for each player.

In regular booster drafts, there’s little opportunity for ‘counterpicking’- seeing the choice of your opponent and picking a resource that is specifically powerful against their choice. However, as the players get to see the overall resource pool in bits and pieces, they can create a general impression of what they might be up against. If they see a lot of cards with the ‘flying’ power in the draft, for example, when it comes time to build their deck out of their drafted cards they may weight anti-flying cards more highly than they would in a blind deck building exercise. This sort of subtle drafting strategy is what lends the MTG draft a lot of its charm and differentiates a skilled player from an unskilled one. As with all games of imperfect information, it provides an opportunity for memorisation, prediction and mental modelling.


The next variant of the draft is one derived directly from MTG’s drafting formats. It is a boardgame called 7 Wonders, whose principal mechanic is an extended draft. Players draft resources which are worth either a fixed value, or a value contingent on what other players draft. For example, a player who drafts a lot of ‘military’ resources will get a lot of points if neighbouring players don’t match them and those players will also lose points. If they only draft as many as their neighbours, neither gain or lose points.

Unlike an MTG draft, a 7 Wonders draft is made from a pool of cards which are always the same. However, these cards are randomly distributed into hands, which are then passed around just as in the MTG booster draft. So, players are aware of the total pool of resources, but are limited in what they can choose during each stage of the draft and in predicting what they will be able to choose in the future. Unlike an MTG draft, in almost all cases when a card is chosen it is then shown to all players before the next is picked, so players have a good idea of what other players are drafting. The result is that counterpicking and exploiting other players’ choices become a far more important strategy in the draft, but the obscured pool prevents choices being cut and dry. 7 Wonders achieves a very tight balance between drafting an optimum strategy (which varies from player to player because of unique bonuses each player receives) and counter-picking to weaken the optimum strategies of other players.

One of the most interesting and novel mechanics in 7 Wonders is that the draft is iterative. Players first draft from a set of weak cards, then once all these have been drafted, a new set of stronger cards are doled out, followed by one more even stronger set. There are complex synergies between the three pools- weaker cards in the first pool make acquiring more powerful cards in the later pools easier. This mitigates some of the problems drafts have with varying resource power, by making a strong early pickup only inherently more valuable in that it forces opponents into certain opposing strategies, since they must exploit the more powerful later game synergies their slow-to-build choices allow.


Finally we reach the realm of the digital, with perhaps the most widely known drafting mechanic in game design circles, the ‘captains’ mode’ of Dota and other ‘Moba’ genre games. Moba games pit two teams of players against one another, controlling a single unit each from a large pool of such units, each with a unique set of properties. The depth of the games comes from the depth of this resource pool, over a hundred such units. Since only a small number of units are selected in a given game, there is a practically infinite number of potential combinations.

Moba drafts also bring the consideration of mechanical skill to the drafting table. A given resource’s value changes not only based on what other players pick, but on the specific player to pick that resource. Since the pool is very deep, a player may be more or less comfortable or familiar with a given unit, very few players are comfortable using all or even most of the available units.

The depth of the resource pool in Mobas often means that certain combinations are unexpectedly powerful. This, combined with the desire to ensure players don’t settle on one optimised unit combination, created the ban mechanic. At the beginning of a Moba draft, each team selects a few units which neither team may pick at all, removing them from the pool. While this is essentially a stopgap measure, as it prevents only the first couple of levels of stagnation as play becomes optimised, the complexity of interaction between players and resources in Moba games mean the impact of bans is sufficient to keep the game interesting and variable over a fairly long period of time.

The ban mechanic varies between Moba games with interesting effects. For example, in Dota, each team bans two units, then picks three, then bans a further three, then picks a final two, for a total of five picks and five bans each. Since extremely powerful synergies between units are possible in Dota, the split picks and bans allow teams to see the initial choices of their opponents and prevent them achieving an optimal draft based on that information. In League of legends, all bans are carried out before picks, so the teams must ‘blind’ ban out powerful compositions the enemy team is known for playing, compositions that are strong against their intended composition or even just overall powerful compositions. Thus in League of legends, counterpicking has a stronger role, as the only way to break the enemy team’s composition is to draft the unit(s) they need yourself.


Having sampled a variety of varied draft mechanics, we can come back to those properties I referenced all the way back at the start.

Whether a draft is clear or obscured has a lot of bearing on drafts in which resource synergy is important. If you can see and understand all the synergies available, the size of the resource pool becomes important, as does ensuring the ability to counter-play, such as the ban mechanic in Moba drafts. Obscuring the pool is also very powerful if it is randomly generated, driving players to develop an understanding of the underlying mechanics of how the randomizer functions and fundamental values and synergies in the draft.

Whether a draft is open or closed is most important when the pool of resources is a common one (while all the games noted here use a common resource pool, it’s important to note that it’s possible to use a combination of these properties to design a draft where each player has a unique pool to draw from). In an open draft it is important that the vast majority of resources are not strictly ‘beaten’ by other resources in the pool, as this leads to situations in which counter-play becomes dominant may leave the leading player at a disadvantage.

Whether a draft is complete or incomplete is again important depending on the overall resources being drafted. An incomplete draft typically requires all elements in the draft have roughly equal values, with only slight variations that are amplified by synergy. If there are clearly weaker resources in an incomplete draft, these will simply not be chosen and are redundant. Since a complete draft requires all resources to be picked so it doesn’t have this issue. A complete draft can also have its total resource pool more easily tailored to always be within certain parameters, useful when the players are not well versed in a complex game and prone to making sub-optimal choices.

Whether a draft is restricted or unrestricted plays a role in determining how resource pools may be distributed in terms of the relative values of individual resources. The iterative draft of 7 Wonders demonstrates this principle, as does the banning mechanic from Mobas. By adding artificial limitations on the drafting pool, 7 Wonders allows for an overall pool containing resources of wildly differing strengths, while Mobas prevent stagnation and optimisation of the draft.


As I have hopefully demonstrated, drafting is a fascinating mechanic to mess around with and explore. It’s capable of creating a huge amount of engagement and emergent complexity, enough that a game can be successful on the strength of a well designed draft alone as 7 Wonders demonstrates. The fact it need only be the initial phase of another game structure entirely means it’s an incredibly versatile mechanic to keep in the toolbox for troubleshooting and problem solving. I hope this exploration has been interesting and eye-opening for you, dear reader. It’s certainly gotten me thinking a bit harder about how I might design a game around drafting myself, or how existing drafting mechanics might be altered to good effect. If you have any examples of somewhere drafting does something cool I haven’t covered, don’t hesitate to let me know!


4 responses to “Mechanics: Drafting

  1. shadowcentaur

    May 21, 2014 at 3:44 am

    Interesting read. I’ve been pondering what a minimal game-theoretic model of drafting would look like. How much could be stripped out and still maintain the fundamental essence of what makes a drafting game stable (if not fun).

  2. Charlie Cleveland

    September 24, 2016 at 3:25 am

    Thanks for the exhaustive read! Very helpful for the game I’m working on with Bruno Faidutti (a rule breaking game like Cosmic or Magic).


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: