July 18, 2016

This Thing Called Strategy



Geopolitical strategy, circa 1675


For the past few years, a question has been haunting my dreams: What is strategy? A narrower follow up question is: What makes a compelling strategy game?

One reason this question has been bothering me, particularly in terms of 4X or Civilization-style games, is that so often the gameplay does not feel like what strategy is or ought to be, at least for me. If the gameplay isn’t strategy, then what exactly is it? And if I’m not getting what I want out of a strategy game, then what in the heck do I really want?!

I have a number of pet theories floating around these troubling questions, which might help me work towards an answer. Fair warning though, much of this article will be spent in the realm of “pontification” or “theorycrafting.” Back in the old days, we called this “BSing.” You’ve been warned!

That said, the concepts I’m trying to discuss are hard to wrap the mind around (well, my mind anyway), so I’ve tried to break my thinking down into bite-sized morsels. These morsels are parts of a bigger thesis I’m working towards. Usually the thesis statement goes at the beginning, but I’m saving it until the end for dramatic effect.

Note: This article is mirrored at eXplorminate, posted July 15, 2016


Simulation, Game, Sandbox, Toy, or Puzzle? Or none of the above?


Games, Contests, Puzzles, and Toys, Oh My!

I’m going to start with something that might ruffle some feathers: many of the games we love to play aren’t really “games” at all. Game designer Keith Burgun, in his hierarchy of interactive forms, describes proper games as a “contest of decision making.” What does that mean? Let’s step back for a moment and consider Burgun’s hierarchy in full.

At the basic level there are toys. Toys are a system of interaction that may have any number of rules (from just a few to a great many) that describes how the system works or operates - but there are no prescribed goals. A big pile of LEGOs on the floor is a perfect example of a toy. It’s a sandbox where you can do whatever you want subject to the constraints (i.e. rules) of how the pieces lock together. Even then, you can break or bend the rules with few repercussions.

Now consider a puzzle. Puzzles are systems of interaction that generally have a single solution or prescribed goal state. A jigsaw puzzle has a correct final arrangement, just as we might follow the instructions to build a LEGO set and arrive at the “goal” of the finished castle/spaceship/hospital. Puzzles generally have optimal or perfect solutions - they are about solving for something.

At the next level are contests. Contests build on the notion of a puzzle by layering in a means of evaluating the result. With a jigsaw puzzle, it is either solved or it’s not. But in a contest, the end result can be measured in some objective way and compared across participants. A running race is a contest to see who can cross the finish line first. We could likewise start a stopwatch and see who can build a certain LEGO set the fastest. Generally however, there are few decisions to make in a contest. The optimal path is usually clear and it comes down to who can execute or solve it better or faster.

Finally we have games. Games introduce the notion of making decisions. The need to make decisions exists because the “optimal paths” to victory are unclear and interlinked with the decisions of other participants. You might not know what move your opponent is going to make, or what the results of a combat encounter will be, or what diplomatic arrangements your enemies are making behind your back. And so you have to make a decision about how to move forward without having perfect information and without knowing the optimal route to accomplish your goal. To round out the LEGO example, consider the game Mobile Frame Zero, which creates a miniature battle “game” out of constructed LEGO robots.

I need to pause for a moment and make an important distinction. Burgun’s use of the word “game” is very specific - and in this article I’m not intending it to replace the more common understanding of a game as a type of media (e.g. a video game or a board game). So, we can have a video game or a board game (or a sports game) that is structurally a puzzle, or a contest, or a toy, or a proper “game.” When referring to Burgun’s definition of a game, I will use the term “game” (in quotes) or the term proper game or strategy game to keep things clear.

Each step in the hierarchy builds on the prior, and so “games” are contests but with the additional element of making decisions. If we think about 4X games, it isn’t hard to imagine one manifesting as any of the four interactive forms. Imagine a 4X game with no opposing empires and no random events. Two players instead play separate instances of the exact setup and we see who gets the highest score at the end of a certain number of turns. We just made a 4X contest. Take out the ability to compare scores, leaving a singular, solved “win state” instead (e.g. transcend or colonize 50 planets!), along with no competing empires, and we just made a 4X puzzle. Strip out any sense of goals, and we have some sort of space colonization sandbox - a toy, or perhaps an empire simulation.


Civilization: A study in internal versus external game mechanics for over 25 years


Internal vs. External Systems

Now that we have a basic understanding of interactive forms, we can examine how different mechanical systems relate to each type of form. In particular, there is an important aspect to 4X game mechanics that drives what sort of interactive form it is: internal versus external systems.

Internal systems relate to gameplay mechanics that exist and operate primarily within and amongst the assets you control directly in the game. In a typical 4X game (Civilization, Alpha Centauri, Master of Orion, etc.), internal systems include city or colony management: production queues, population happiness, tax rates, economic balance, research priorities, etc. Consider this: if there were no other players or empires in the game, which mechanical systems would continue to function more or less as normal? Those are the internal systems.

The external systems are gameplay mechanics that create and/or depend on interactions with forces outside of your control. Most often these are the interactions you have with other players or empires through the likes of military conquest, espionage, diplomacy, trade, foreign relations, and so on. Beyond other players or empires, it could also include asymmetric forces like random events, endgame threats, space amoebas, or other sources of randomness that add chaos and unpredictability to the gameplay. The key aspect to keep in mind about external systems is that they are outside of the player’s control.

These differences are critically important. In order to have strategic gameplay there has to be an engagement with external systems. Why? Because these external systems and resulting interactions, per Burgun’s hierarchy, are what enable a game to be a proper “game” - and not a puzzle or a contest. External forces provide ambiguities, which obfuscate the optimal paths to victory, and in turn create room for strategic play where we can’t be certain whether our long-term decisions will pay off or not. Moreover, being able to navigate these ambiguities better than your opponent is where skill matters in determining the eventual winner. Games that have many levels of skill (e.g. Chess rankings) and more elaborate heuristics, tend to be deeper and more strategic games.

By contrast, the more a game leans on internal systems, the more puzzle- or contest-like it tends to be (e.g. Apollo4X). In most 4X games, for a given setup, there is an optimal path to expand and grow your empire that follows the rules of the game. This optimal solution can exist because there are few (or no) external systems that make the potential results of the decision process unclear. Of course, external pressures might shift or change what you are optimizing towards during the game - but once that shift in direction is decided, the actions that follow are largely self-evident.


In some games, survival is the only form of winning


The Goal of Succeeding versus Surviving

A curious quality to games is the difference between succeeding (e.g. meeting a victory condition) and surviving. Some games are structured around the notion that eventually you will fail to survive. Consider the game Tetris. Eventually, the blocks will fall so quickly that the game becomes mechanically unwinnable, and so the game ends and you get a final score. Burgun’s iOS game Empire is the Tetris of 4X games. Eventually your empire will be overrun by external forces - the challenge is to see how long you can survive and how big your final score will be.

Survival games can also be driven by more passive or internal forces. There are plenty of survival sandbox games these days (The Long Dark is a nice one), and here it is less about keeping ahead of some menacing threat actively trying to kill you and more about managing your own affairs and assets such that they don’t unravel and lead to your demise.

Similarly, Paradox’s grand strategy games tend not to have specific victory conditions. Games usually end when the time period covered by the game is over, and the main question is whether or not you survived to that end point. Players might also establish goals of their own choosing during the game. In this regard, these games function more like Burgun’s “toy” definition - although I’m inclined to call them “simulation sandboxes” given the level of complexity and the potential for “failing to survive.” So does the lack of a defined victory condition make it less of a proper “game?” I’m not sure - but maybe.

Most 4X games, however, concern themselves with the notion of victory and “succeeding” - being the first to reach a goal or victory condition. Granted, there may still be an aspect of survival at work, as other empires may decide to wipe you off the planet (or galaxy)! And so in many 4X games, there is a tension between the need to survive and the need to achieve victory; finding the balance is certainly a question of strategic decision making.

So what then are these strategic decisions?


Civilization. Civilization never changes (Heiko G√ľnther’s Advanced Civ Re-design)


The Balance of Actions

The next theory I want to lay out is an approach for categorizing the different types of actions or activities one might take in a strategy game. Personally, I want games that emphasize making interesting choices as opposed to making mindless non-decisions. Think of it this way: deciding whether to spend the afternoon at the park or going to see a matinee movie might be an interesting choice, but deciding to turn on the car in order to drive is a necessary (and boring) part of achieving either goal. We’ll get to what interesting means in game terms in a bit. For now, I tend to see actions in the following types:

Strategic Decisions: These are high-levels decisions that feed into how you are going to win the game. Most often, strategic decisions are influenced by external systems. Is my neighbor going to invade me (or not), and should I therefore strike first (or not)? How much should I invest in building military units versus funding empire growth? Who should I conduct espionage against or form an alliance with? What type of victory condition am I working towards, and how will I get there before everyone else? Do I need to shift strategies? Strategic decisions exist in our minds - they don’t play out in the physical game space. They are about establishing objectives that set you on a path to victory.

Tactical Decisions/Actions
: These are the important decision points and/or actions players take to actualize their strategic decisions or to respond to short-term issues and events. They relate to how you will accomplish an objective. If a long-term strategic plan calls for subjugating a neighboring empire, how are you going to do it? What type of fleet will you build and what route will it take? How will you deal with enemy forces or planetary defenses? Unlike strategic decisions, the result of making a tactical decision is usually reflected by a change to the game state - e.g. I move my fleets to another system, and thus the game state has changed.

Optimization Activities: These are actions that relate mostly to internal systems and consequently ask you to solve or optimize for a particular objective. Do I build my research lab and then my production facility, or vice versa? A lot of time can be spent in 4X games optimizing a particular decision point, and, depending on the complexity and math involved, can be very challenging or relatively trivial. Adjusting the allocation of workers on a colony between production, food, and research is an optimization task as there is often a best solution for a given strategic goal. 4X games are occasionally derided as being “spreadsheet managers,” and the need to optimize outputs (or military efficiency) strikes at the heart of that criticism.

Upkeep & Overhead Actions: These are the routine actions that relate, again, mostly to internal systems and are part of the maintenance or upkeep of your assets. Generally, there is little choice in these actions, they are things you just have to do to advance the game state. In board games these upkeep actions are quite common (reshuffle decks, refill tokens, pay upkeep costs, etc.). We see these in 4X video games, too: tweak the ship design to add the newest laser weapons, add the newly-researched building to your all your production queues, send constructed units to the rally point, clear notifications to advance the turn. These are “no brainer” decisions that rarely require much thinking.

I’ve often found myself critiquing strategy games by asking “what percentage of my time am I spending on what types of actions?” The optimal balance is, of course, a matter of personal preference. For me, I’d much rather spend my time making strategic and tactical decisions, rather than running optimization exercises. Overhead actions, ideally, are just automated and resolved by a competent AI or streamlined UI - or else removed entirely. As a result, I tend to prefer games that emphasize external systems (e.g. more wargame focused 4X titles) over those focused on internal systems and hence optimizations and puzzle-solving.

The notion of survival versus success is also relevant to this topic. Strategic or tactical decisions are easiest to see as they relate to external factors (e.g. other empires), which in turn relate to the choices you make to move closer to success. Less common, but certainly possible, are strategic and tactical decisions relating to survival and internal mechanisms. Grand strategy games often latch onto this idea - where various internal pressures (e.g. mismanagement) can result in a revolt or collapse (e.g. a coup or assassination). This transforms them into external factors, which could then destroy your empire. But I feel like more could be explored along these lines.

A stone here, a stone there… Where will you Go?


The Deception of Complexity

Consider for a moment the classic board game Go. Go has a ruleset that can be explained in a few sentences. And while it’s one of the simplest strategy games, it also has nearly unrivaled depth. This no doubt accounts for the game’s lasting appeal over the course of thousands of years (yes, thousands). The key point is that mechanical complexity does not equal depth, and Go is a testament to the notion that great depth can emerge from simple systems. And so, if we can achieve great strategic depth through simplicity, what role does complexity then play in strategy games?

Complexity can affect gameplay in two fundamental ways. First, complexity can affect the size of the decision space. Playing Go on a 9x9 grid is less complex than playing on a full 19x19 board, where there are vastly more possible moves and game states. Second, complexity can affect the number of factors or layers that go into making a decision. Imagine a simple, multilateral wargame with no option for diplomacy. Now insert diplomacy - suddenly there is a new system for interaction that can influence your decisions for who to defend or war against.

But does this added complexity always make for a deeper strategic game? Not necessarily.

Perhaps enabled by increased computing power, I feel that strategy games have become more complex over time. For many, this added complexity is welcome because it means the game has more longevity - it takes longer to tease apart all the inner workings and to build up skill. We see this frequently in modern board games as well, where learning the rules of the system is a major part of a game’s appeal. Players discuss the joys and thrills of learning how a new system operates and what all the levers and cogs do. But this can be a double-edged sword.

In many cases, complexity merely makes the math of solving optimization problems more convoluted and challenging - diverting attention away from the real strategic interactions in the game. For example, many 4X games have giant tooltips filled with positive and negative modifiers explaining all the factors affecting a colony’s happiness. Maximizing happiness, and in turn productivity outputs, requires identifying what options you have to mitigate each of the contributing factors and determining which has the best net return. You might even conduct this optimization task across all of your colonies to determine exactly which one yields the most bang for the buck. In this regard, the complexity is making the optimization harder, but it doesn’t really deepen the strategic landscape - you are still trying to solve for the same X.

Moreover, once you’ve cracked the code and learned these internal optimizations, you have solved the major puzzle of the game - and can then beat it relatively easily over and over again. There might be strategic or tactical decisions to be made - but they are no longer as interesting and gameplay depth has been diminished as a consequence. A question to ask yourself is this: does a given strategy game become more interesting or less interesting as you play it more?

That’s the same face I make when I strategerize...


The Quest for Deep, Interesting Decisions

My ideal strategy game is one where I spend most of my time making interesting strategic and tactical decisions - compared to optimization and upkeep actions. But what makes a choice interesting in the first place? Principally, an interesting strategic decision is one where you have to make a choice and you are uncertain about what the long-term payoff of that choice will be. But you are not shooting blindly in the dark, either. This balance of uncertainty - and the nature of it - is crucial because otherwise the “game” is reduced to a solvable, though potentially quite complex, puzzle.

Uncertainty itself can arise from a number of sources, each of which has an implication on the strategic depth of a game.

One source of uncertainty is chaos or randomness in the game system. If random events, die rolls, or the Wizard-Kings of Probability have a bearing on your long-term decisions, then clearly the outcome has uncertainty to it. However, this may not make a deeper or more strategic game; rather it may just make it more unpredictable and harder to predict. Would chess be considered as skillful and deep if there was only a 50/50 chance to capture a piece? The randomness would make it difficult to strategize and diminish the potential gains for careful planning. In other cases, for example in a game like poker, high degrees of uncertainty adds another level - one of probability and risk assessment - to the optimization activities. It makes decisions more uncertain and harder to calculate, but maybe not in a fundamentally more interesting way. What makes poker interesting is that the randomness of the deal is filtered through the skills and behaviors of other players in an interactive way.

So then, the other major source of uncertainty is related to the interactions between players - and here is where decisions become more interesting. If “games” are understood to be interactive systems that are contests of decision making, then having to account for and react to the actions of your opponents is crucial. Player interactions are external in nature and manifest across a number of 4X game systems: diplomacy, military positioning, espionage, etc. They can also take on a number of different forms: open negotiation, bluffing and feigning, double-think, maneuvering, etc. The crucial skill is being able to read your opponent based on understanding their position, personality, and playstyle, and in turn identify your likely moves (and countermoves). This is where you can leverage your own wit or cunning to achieve a strategic advantage. This is where skill and experience comes into play.

Ultimately, what makes choices interesting is whether or not the strategic landscape of the game - the multi-layered decision spaces that exists in your mind - allows unique and consequential decisions to emerge. In the board game world, games are often discussed as having either “pre-baked” strategic pathways that are created by the designer (and to be discovered by the player) versus games that are more player-driven and emergent in the game states and situations that arise. The pre-baked path approach relies heavily on “learning the system” and on complex internal mechanics.These are often paired with limited player interaction and less volatility as a result. The player-driven approach is more in line with the “simple to learn, lifetime to master” notion - where the depth and interest comes from unique situations where player personalities mix in an interactive and dynamic environment. The former is predominantly about optimizations, the latter is concerned with strategic or tactical interactions.

Diplomacy for another age


Implications for 4X Game Design

I’ve laid out a number of pet theories in this article:

- The definition of a game versus a puzzle, toy, contest, or simulation
- Internal versus external systems
- Surviving versus succeeding (victory, goals)
- Types of actions (strategic, tactical, optimization, upkeep)
- The roles of complexity
- Interesting decisions, uncertainty, and player- vs. system-driven games.

What does all of this mean for 4X games? If I have one big critique (here is the thesis!) of 4X games, it is that they often emphasize the exact wrong things in their design (given my preferences), and so I don’t find many of them to be all that strategic as a result. In many cases I’m not even sure they could be classified as proper “games” (per Burgun’s hierarchy) - they feel, to me, more like puzzles.

Complexity appears to be increasing in 4X games, but much of this complexity is directed towards internal game systems: ever more intricate systems of colony management, internal policies, worker optimizations, more complex development pathways, and so on. Little of this really affects how interesting the big long-term strategic decisions are. In fact, the focus on creating compelling or interesting victory conditions (essential for a proper “game”) seems to be in decline - making the choice of what you are optimizing for all the more obvious. In so many 4X games, I feel your race selections and starting position railroad you down a certain track towards a certain pre-ordained victory condition. You might start the game game knowing you are going for a technological win because your empire/species is all about boosting technology. The decisions that follow from there are all about optimizing and solving for X. It’s a puzzle, not a game.

One of the challenges with complexity also has to do with the AI’s capabilities and level of cunning. On one hand, a shift towards greater focus on internal system complexity could be seen as a way to sidestep a weak strategic AI. However, the AI still has to navigate these complex internal systems, and often it ends up receiving bigger and bigger bonuses to compensate for its inefficiencies. This isn’t a good foundation to build a competitive strategic game. On the other hand, simpler game systems might be able to better leverage a computer’s brute-force calculation power to legitimately out-optimize or out-wit the player. I have a Go app on my phone and the AI, sans bonuses, absolutely trounces me. Go figure...

Other types of 4X games (and especially grand strategy games) take a different approach. They are using increasing complexity as a basis for building more detailed simulation models. Within this type of simulation, players are at liberty to decide their own goals and what game systems to focus their choices around. It is a sandbox experience and, short of a failure to survive, is not usually oriented around goals or victory conditions at all. This is, of course, a perfectly valid approach, and simulations have a great capacity to allow for player-created narratives to emerge. But in a certain sense, these really are not “games” either - at least in the strict sense of active competition for victory.


Depth through simplicity?


So, 4X games appear stuck between a puzzle optimization pole on one end and a complex simulation pole on the other. And neither of these really results in a focus on making interesting strategic decisions based on external, player-driven interactions.

Personally, I’d love to see a 4X game take a different approach and embrace mechanical simplicity - using it to build a more interesting interactive player environment. What would a 4X game with practically zero empire management look like - with all the focus instead on diplomacy, military maneuvering, controlling shared markets, and cultural exchange? The skill of the game, and its potential depth, would be less contingent on knowing the optimal pathways and instead about making strategic decisions within an emergent and dynamic game space, including the personalities and eccentricities of your rivals.

Most titles seem to drift towards either the survival/sandbox simulator or the optimization/ steamroller to victory. There are a few games that strive to zero-in on interesting strategic decisions and that focus more on external interactions as a result. Age of Wonders III, for example, has relatively simple empire management and de-emphasizes optimization tasks. Instead it emphasizes military positioning, maneuvering, and the careful use of magic resources - all higher level strategic or tactical decisions. This bring it closer to a proper strategy “game” than many other 4X games, at least given my preferences. I would put Master of Orion (the first one) or Sword of the Stars (the first one) in the same category. They are relatively simple games mechanically that emphasize external interactive systems over complex internal mechanics. But fewer and fewer games seem to follow in their footsteps.

As a parting thought, consider these various pet theories and whether they have informed or changed your perspective of 4X games that you have played. How do your own interests and preferences align or not with these concepts? Do you see other styles of 4X or strategy games that do or could exist? Do you feel that the games you play are are “puzzles” or “contests” or “games?”

As always, the comment line is open.


1 comment:

  1. Great awesome bit you got there. Me and my kids love games like Minecraft and Roblox. I say try the new Roblox Helper and get free robux.

    ReplyDelete