Posted on

A Model for Decision Making in Games, Part Four: Case Studies

Last fall I laid out a model for decision making in games that divided it into three parts: action selection; action execution; and action resolution. This article concludes the series by examining how it works using the concrete examples of three games: the DSP winners from three different decades — which also does a nice job of highlighting how the hobby has changed in the last twenty or more years.


Tigris & Euphrates (1997) by Reiner Knizia

Tigris & Euphrates is a game in the classic eurogame mold. It features some elements rare in modern eurogames, such as more chaotic resolution, and it has a simpler selection model. Put all that together and you have a game that does a great job of demonstrating how this decision making model can be used to map out any entire game. 

Action Selection: Rules Selection, limited Action Points

Tigris & Euphrates doesn’t feel like it has an action-selection model because it uses an old-school methodology. Each turn, players choose two actions from a list found in the rulebook (and on the back of the screens): position a leader, play a tile, play a catastrophe tile, or swap tiles. Despite this simplicity, there are just enough mechanics here to create interesting selection tactics, and that’s largely due to the fact that players can make two action-selection choices each turn, creating the opportunity for surprising combos.

Action Execution: Area Control (Spatial Choice, Token Choice)

Even after choosing an action, a player still has to decide how to execute that action, and here the choices are pretty wide open. For tile swaps, a player can exchange up to six tiles, and for token placements he has up to twelve token choices: four leaders, six civilization tiles, and two catastrophe tiles. Those can go on up to 176 board position. There are certainly real restrictions (such as the placement limitations on blue tiles and river spaces) and practical restrictions (such as the fact that players will primarily place tiles adjacent to their own kingdoms), and they reduce these 12×176 options . However, it wouldn’t be unfair to say that there are still hundreds or thousands of possible action-execution choices, far beyond the four action-selection choices.

Action Resolution: Automatic, Arbitrary

Most results in Tigris & Euphrates automatically occur, but when a player creates a conflict by placement of a leader or by connecting kingdoms, then an arbitrary action resolution occurs — something that’s much more rare in modern eurogames. And, it’s an intriguing one that combines the placement of extant tiles on the board and the play of arbitrary tiles from hands. This keeps the game in line with the strategy of long-term placement, but allows for some tactical surprises (and some pressing your luck). Though there are far more choices in in the action execution phase, the heart of Tigris & Euphrates is really in its resolution, because that’s where the results of the big moves are determined.

Case Study Conclusion

Tigris & Euphrates shows how a game can combine fully featured elements of all three parts of the decision making system. There’s just enough action selection to keep things interesting with a minimal set of two Action Points used for Rules Selection. (A fully old-school game would have just allowed one selection.) Most of the gameplay then occurs in the action selection, but its choices are pretty classic: picking tiles and deciding where to place them. Finally, there’s action resolution to make sure the choices aren’t too staid. It’s a really direct conflict, but one that prioritizes long-term planning over seat-of-the-pants play. Overall, Tigris & Euphrates looks like a bit of a hybrid in the modern day: it partially develops the action-selection phase more common for eurogames, but maintains both a complex action-execution phase and a partially random action-resolution phase, as iswasmore frequent in older Anglo-American releases.


Agricola (2007) by Uwe Rosenberg

Agricola certainly isn’t the first worker-placement game, but it’s a well-known example of the category that uses the mechanic in a fairly standard way; it demonstrates what people like or don’t like about the mechanic.

Action Selection: Worker Placement

Agricola is as pure as worker placement gets. At the start of the game, there are 16 actions. 14 more are added over the course of the game, for a total of 30. Whenever a player takes an action, no one else can take it again that round, though there is some duplication of functionality. The majority of the game’s contention thus happens in the action-selection phase; this is also where the players have to constantly adjust their play to maintain their efficiency. While the heart of Tigris & Euphrates was in its execution, in Agricola it’s right here in the selection, and there’s a reason for that: in each game, the heart is where the players come into conflict with each other.

Action Execution: Resource Management (minimal Spatial Choice, limited Construction & Token Choice)

At first glance, the Action Execution phase of Agricola might seem to be fully degenerated. After all, critics of worker placement say that you just get the choice of which action to take, and that’s it. That certainly could be true for some worker placement games, but in Agriciola, it’d be fairer to say that the action execution phase is limited, not fully degenerated. First, there’s minimal spatial choice: how you expand your house, build your fields, and fence them is meaningful, but slight variations don’t make a lot of strategic difference. Second, there’s a limited but crucial Construction Choice for which improvement and occupation cards a player chooses to build, and directly linked to that, which resources he spends to do so. These are much more important to long-term play because they create the foundation of a player’s gaming engine.

Together, these limited options roughly double the total breadth of choices in Agricola. No, the 30 or so choices available at the start of the game in Agriciola don’t equal the hundreds or thousands in Tigris & Euphrates, but it’s also hard to compare a choice like “place in the space above my kingdom or one space to the right of that or one space to the left of that” with “take two wood or play a wood cutter to improve my wood engine or play a joinery to get better output from my wood engine”. Each game finds complexity and strategy in different ways.

Action Resolution: Automatic.

The final, action-resolution phase, is wholly degenerated in Tigris & Euphrates. There are no surprises: the results of everything are automatic.

Case Study Conclusion

Agricola is an example of how the eurogame market came of its own a decade or more on. Rather than having the wide-open but abstract choices of Tigris & Euphrates, it has more constrained and also more evocative choices. And rather than having the in-your face conflict of Tigris & Euphrates, it instead makes its conflict more hands-off: players don’t play tiles to attack each other, but instead fight to get in line first. It’s certainly a different style of play, despite both games originating in the German community. It shows how dramatically rearranging these three decision making phases can result in dramatically different games.


Terraforming Mars (2016)

Although a modern board game, Terraforming Mars is some ways feels like a throwback, because it goes to the old Anglo-American model of being more random and more simulationist. However, it continues the trend of eurogames away from an action-resolution phase and toward an ever-more complex action-selection phase.

Action Selection: Card Selection and/or Card Drafting then Action Resources and/or Menu Selection

Terraforming Mars actually breaks its action selection into two separate parts, showing how action selection is becoming increasingly complex in eurogames, decade by decade.

First, players decide which cards they want to put into their hands. This is primarily done through “Card Selection”, a methodology that I didn’t include in my original Action Selection article: players are given a large handful of cards, and must decide which are worth purchasing for future use. However, there’s also a somewhat popular variant: players can draft their cards in later rounds of play, which increases game length, but also decreases randomness and increases interactivity.

Really, though, that’s just the prelude: players select actions through an Action Resources method that’s totally separate from the card selection and/or drafting. Players decide which cards to play, in large part dependent upon what they can afford. And, they may never play some of their cards. As a further alternative, players can choose set actions from a menu, whether they have appropriate cards or not, which once more decreases the randomness. Later in the game, players will also get to choose actions from their personal menu of played cards that have actions on them, allowing the creation of interdependent engines as the game goes on

More generally, Terraforming Mars shows how you can use cards to initiate actions, but also how you can both acquire those cards in multiple ways and how you can control randomness while doing so.

Action Execution: Resource Management (minimal Exchange Choice, Expenditure Choice, minimal Opponent Choice, Spatial Choice)

Fundamentally, the majority of the choice in Terraforming Mars comes from the action selection, where a player chooses which card to play, which project to build. or which action to activate. There are expenditure choices created by some of these selection choices, as players choose how to pay for their cards. In addition, a minority or cards, projects, and actions also create execution choices: which tokens do you expend (Expenditure), what do you swap them for (Exchange), what opponent do you steal resources or tokens from (Opponent), or where do you place a tile (Spatial). They’re critical elements of the game, but less so than the core choice of selecting and playing cards and choosing actions.

Action Resolution: Automatic.

Once again, this modern eurogame has a degenerated Action Resolution phase: there are no surprises; once you have your cards in hand, everything occurs as planned.

Case Study Conclusion

In many ways, Terraforming Mars “feels” like a much more classic board game, akin to the more primordial design of something like Tigris & Euphrates. But, when you look at its design, if anything it doubles down on the design decisions seen in the popular worker-placement category of play: most of the decision are made from a limited menu of no more than twenty choices in the selection phase, with the execution phase just offering minor details to those decisions, and the resolution phase contributing nothing.


Appendix: A Chart of Games

The following chart encompasses not just these three games, but also several more that were used as examples in the original articles.

Game Selection Execution Mechanic Choices Resolution
Agricola Worker Placement Resource Management Construction
Spatial
Token
Automatic
Dominion Deck Building Card Management Opponent Automatic
Near and Far Rules Choice Adventure/Exploration Exchange
Expenditure
Obstacle
Purchase
Spatial
Token
Random
Unknown
Pathfinder ACG Deck Building Card Management Expenditure
Region
Token
Random
Unknown
Puerto Rico Phase Selection Resource Management Construction
Exchange
Automatic
Settlers of Catan Action Resources for
Menu Selection
Resource Management Construction
Exchange
Expenditure
Spatial
Token
Automatic
Decision
Terraforming Mars Card Drafting or
Card Selection then
Action Resources or
Menu Selection
Resource Management Exchange
Expenditure
Opponent
Spatial
Automatic
Tigris & Euphrates Action Points for
Rules Selection
Area Control Spatial
Token
Arbitrary or
Automatic
Tikal Action Points for
Menu Selection
Majority Control Exchange
Spatial
Automatic

These primarily euro-driven games again show a tendency toward the selection phase and away from the resolution phase, but the more Anglo-American design of Near and Far shows the other possibilities out there.

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

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