Posted on

Defining Worker Placement

Would worker placement by any other name smell as sweet? Perhaps. But there’s power in names: they allow us to develop a common vocabulary, so that we know what other people mean, permitting us to set our own expectations.

That means that a big kerfluffle about naming conventions is significant, such as when a notable board game show says that a non-worker-placement game is one of the top games in that category of play. Because it muddles our meanings, it impairs our communication, and it sets incorrect expectations:  if you loved worker-placement games and picked up the game in question based on a recommendation, you might well be disappointed (or not: it’s a great game otherwise).

So this week I wanted to give my own definitions of worker placement, starting with a look at its history.

A History of Worker Placement

The worker-placement mechanic grew out of an older mechanic, role selection, which I previously wrote about while discussing “role civilization” games. It started with Verrater (1998) and continued on through Meuterer (2000) before it gained mainstream success with Citadels (2000) and Puerto Rico (2002).

Basically, role selection turned the even older mechanic of action selection on its head by transforming those actions into a limited resource. Players no longer got to freely determine what they would do. Instead, they had to contend for actions with their opponents. It innovated gaming by creating tension and uncertainty in what previous players would have considered a pretty essential element of gaming: whether a player got to do the action he wanted! It also dramatically ramped up interactivity: a player might now select an action not just because he wanted it, but also to deny it to his opponents; and players now kept careful watch over their opponents’ turns, as they fearfully waited to see if they’d lose the action they wanted to take.

The selectable actions tended to take the physical form of cards (or their equivalent, tiles). These were nice components because they offered a visual reminder of whether an action was available or not, and they could also be used to list out what could be done with the action. Various games offered variants on this core play: most supported open selection, but some such as Citadels instead allowed hidden selection; and most permitted singular actions, but some such as Puerto Rico instead caused a player to open up a new “phase” of play in which everyone took part.

Enter Caylus (2005), the game that created the genre of worker placement. (Was it the first worker-selection game? Probably not, but many of the claims for earlier “firsts” are spurious.) Worker placement started with the core of role selection: creating contention for player actions. But it moved the actions from a set of “role” cards to spaces on the board, and it moved the selection from free selection to a choice that was linked to a player resources: meeples. These two changes allowed the classic role-selection mechanic to expand within a game: not only could the number of actions easily grow, but the acquisition of resources for use in the worker placement became yet another subgame. Worker placement also managed to create a much more coherent and evocative play system than role selection by the simple fact that all the action spaces were integrated into a playing surface, rather than scattered about on cards, which is in my opinion why it succeeded.

Though Caylus (2005) wasn’t the first game to use worker-placement-like mechanics, it defined and popularized the category. (And what does it say that my only photograph of Caylus shows an empty board?)

Caylus was a phenomenon in the year or two after its release. I personally thought that it had fundamental flaws largely related to a play length that could telescope by a factor of two dependent solely on player choice. I only played it four times and quickly fired it when the better polished Caylus Magna Carta (2007) card game came out. But, I also spent those two years dodging Caylus games. The game was very popular, in spite of its flaws, and that was largely due to the popularity of the worker-placement mechanic, which took the simple idea of a player choosing an action, and turned that into a multi-leveled, competitive set of game systems … before you even got to the resource-management systems that defined the core of Caylus’ play.

In the next few years, the one-two punch of Agricola (2007) and Stone Age (2008) locked in worker-placement as a major eurogame mechanic. Since then, worker placement has probably become the most-common mechanic in eurogames, displacing the auction and majority-control mechanics that were commonplace in the genre’s youngest days and standing up to the deckbuilding that soon followed. It’s so popular that it may even appear in a majority of new eurogames.

Popular latter-day worker-placement games include Caverna (2013), A Feast for Odin (2016), Le Havre (2008), Lords of Waterdeep (2012)Tzolk’in (2012), and Viticulture (2015).

What Worker Placement Is

So what is a worker-placement game? I generally believe that it’s an action-selection game, which focuses on players choosing what they’ll do on a turn from a menu of possibilities, where that choice also has four additional characteristics:

  1. Actions Are Shared. The actions that are the heart of the game need to be publicly available, to be used by any of the players — like the building spaces in Caylus or the action spaces in Agricola or Viticulture.
  2. Actions Are Limited. Each action can only be taken a limited number of times. Most often, that means that an action can only be taken once in a round, but there might be multiple spaces for the same action, or it might be possible to take an action multiple times. As long as the limitations are meaningful and cause tension (and anguish!) during the game, you’ve still got worker placement. Some designs such as Stone Age have specific rules for how many workers can be placed on each space. Other designs are more far-flung, such as Keyflower (2012), which allows multiple players to take the same action, but at ever increasing costs, with a hard limit of three.
  3. Actions Are in Contention. This is the critical point where those other two characteristics come together: those shared and limited actions are in contention, so that when one player grabs one, it’s not available to his opponents. This is almost the defining characteristic of worker-placement games. Without it, you have some sort of standard action selection: with it, you have the tense and constrained gameplay that grew out of role-selection games.
  4. Selection is Limited by Resources. However, as the term “worker placement” suggests, there’s one more critical element: the workers. Players don’t just choose their limited, shared actions willy nilly, they have to use limited resources to grab them. Usually, these resources take the form of workers placed on worker-placement spaces. Sometimes, players must also pay additional costs. Generally, one could imagine many sorts of resources being expended and still having a game that’s a lot like worker placement.

A few other characteristics are optional, but are ubiquitous enough that they’re found in many worker-placement games.

  1. (Optional, but common) Actions Expand Over Time. One of Caylus’ big innovations was that it slowly expanded its available actions, increasing complexity as the game went on. Agricola and others followed this same formula, but there are some worker-placement games such as Viticulture that don’t do this. Not using this option largely removes the ability for players to ramp-up their workers, and so removes an orthogonal way to advance in the game — but if you’ve got a game such as Viticulture that’s quite multi-faceted already, that’s not a problem.
  2. (Optional, but rare) Actions Can Be Player OwnedThis was Caylus’ other big innovation that hasn’t become mandatory for worker-placement games. In fact, it’s pretty rare. But some games such as Keyflower do allow players to own some or most of the worker-placement spaces. This creates another orthogonal strategic question: do you aid your opponents by using their spaces or not?

Does a worker-placement game need to all all four of the “mandatory” points in my description? I would say yes, but there’s a lot of possibility for variation, such as in The Manhattan Project (2012), which uses worker placement as part of a larger whole. But, if you get far enough away, you might have moved entirely out of the category of play.

The Manhattan Project (2012) is a little outside of the norm because it combines pure worker placement with personal action selection, and uses the same worker-meeples for both.

As with so many things in life, you’ll know worker-placement when you see it (as long as you have a strong definition, like the one herein), but generally you should see something that focuses heavily on these four or more points.

What Worker Placement Is Not

Worker placement is sometimes confused with a few other categories of play:

Action Selection. As with worker placement, action selection depends on players deciding what they’re going to do based on some menu of possibilities. Unlike worker placement, actions in action selection aren’t shared and they aren’t in contention. I’d generally define worker placement as a subset (or an advancement) of the action-selection mechanic.

Area Control. This sort of play uses tokens to gain control of regions in a game through warfare: new markers come in, and old markers are either destroyed or move out. Perhaps those regions later produce resources or allow some special action, but players don’t actually use their area-control meeples to select the action (nor is that action available to others in any meaningful way).

Auction. Here, players bid resources to win prizes. They could theoretically be bidding on actions, which puts auctions into the realm of worker-placement games. However, bidding on an auction is a totally different mechanic from simply placing a meeple: it has uncertainty, the possibility of loss, and generally requires a different type of strategy. You could perhaps make the argument that worker placement is a subset of auctions where everyone bids “1” and the first “1” wins, but you could perhaps make the argument that for most mechanics, so I think it’s specious.

Majority Control. This category of play is very similar to area control, except that control of a region is determined by who has the most meeples in a territory, not by who last player to drive their opponents out. The rest of the discussion applies.

Village (2012) is hard to categorize. It definitely has shared action spaces, and you’re definitely placing limited resources on them, and they’re definitely in contention … but that contention is defined by the resource cubes that you remove from the regions, not by the number of meeples placed there. I’d call it worker-placement, but an usual variant.

Role Selection. The history section of this article already covered the difference between role selection and worker placement, but in short, worker placement built on the action contention of role selection, but integrated it more tightly into a central board and introduced the need to expend resources to use those actions.

Some Controversial Games

So, finally, we come to those games that people claim are worker selection, but where I’d argue otherwise. I’ve listed five of the most popular here, with my argument for why they’re not worker placement.

Carcassonne (2000). You place meeples on territories, and if you have the majority of meeples in the territory when it scores, then you earn the points. In other words, the meeples, even though they have worker-like names such as “farmer”, “knight”, “monk”, and “thief” are majority-control markers. They do not select actions of any sort; instead, they contend for points. Majority Control.

Keydom (1998) / Aladdin’s Dragons (2000). This is a game that some claim was the first worker-placement game, and it’s not, in large part due to its absence of discrete workers and the uncertainty of its action selection. In fact, it’s really a bidding game (call it auction or majority control, as you prefer). Players place tokens in action spaces, and then when they’re tallied up, they determine who gets to take the actions. It’s still a pretty unique action-selection auction that I’d love to see available in a new “key” edition from Richard Breese. Auction.

Orléans (2015). Players draw characters from bags and combine them into formulas to take actions. The fact that actions are activated through the combination of multiple “worker” pieces would probably disqualify this game from being worker selection on its own. But, the actions also aren’t shared, so there’s no contention. Action Selection.

Robinson Crusoe (2012). There’s a set of shared actions in this co-op, all plainly visible on the main board. Players take them by putting one or more of their action discs on the space. It’s not worker placement because there’s no particular contention: players can stack their “workers” wherever they want. In fact, it’s been pretty hard to marry co-op games and worker placement because the contention at the heart of worker placement runs exactly at odds with cooperative design. Action Selection.

Scythe (2016). And finally, this is the game that got the argument started, when a popular video show claimed it was a top worker-placement game. Many of us thought that the reviewer was talking about Scythe’s action-selection mechanic, where players choose their actions from their personal player boards, which of course has no contention (though it creates some clever limitations based upon its pairing of actions, where those pairs always have to be done together). However, in retrospect, it looks like the argument was weaker than that: Scythe calls its some of its meeples “workers”, and that seems to be the reason for the claim that Scythe is a worker-placement game. But, a name does not make it so. The “workers” are area-control units, and though they do generate resources, they have no part in the action-selection cycle of the game. Action SelectionArea Control.

Scythe (2016). Play this fun and innovative game, but please don’t call it worker placement. That’s a simple action-selection board shown above.


Worker placement depends on an action selection system where (i) the actions are shared, (ii) the actions are limited, (iii) the actions are in contention, and (iv) the actions require limited resources to activate. It is not area control, auction, or majority control. A game isn’t worker placement just because it calls one of its tokens a “worker” and you at some point place it somewhere.

Liked it? Take a second to support Shannon Appelcline on Patreon!

The original article can be found on the great Mechanics & Meeples