Even though the following post was written over the course of the last 6 months and served mostly as a way to figure things out for myself, I believe it will be of interest for whoever follows my work and the future of the Citystate 'franchise'.
This post argues whether the deep economic and political systems should be preserved, simplified or dismissed in the upcoming new Citystate game. More particularly, I review here what did not work in Citystate II both in terms of game design and marketing. Since I primarily focus here on negative points, this post might come out quite bleak and pessimistic but, again, this is a discussion I had with myself, not a corporate-community-marketing post. Hopefully this will tell you about my train of thought and help you understand some of the design decisions behind both Citystate II and my next project.
After releasing two games I came to the realization that my target audience does not really exist. There are basically 2 groups of people interested in Citystate: people looking for a political simulation and people looking for a Cities Skylines alternative. The problem here is that these 2 groups are probably impossible to satisfy simultaneously, so when the few of each group get their hands on Citystate they often walk away disappointed.
Now don't get me wrong, I have read dozens and dozens of emails, posts and reviews acclaiming the games, some even saying that it changed the way they look at their city and country (something I am glad to hear since it was kind of the objective of the game). But in the end, the audience for what Citystate really is just happens to be too narrow or unreachable. This makes maintaining and updating Citystate a little unwise financially as we get further away from the initial release and as sales numbers continue to drop. Then there's the fact as well that, as a developer, I want players to enjoy what I make, even when the audience is relatively restricted.
Improving upon the first game
Even before the development of Citystate 1 ended I already had a clear picture of what I wanted to do with the sequel. The question of accessibility and difficulty was the most problematic part of the first Citystate, or so I thought at the time. To that end, policies in the sequel were to be more understandable and mechanics more transparent. Citystate II had to have a well thought-out, closed-loop economy that could scale easily. All the classic city building features should be available at launch and the overall experience was to be more coherent and streamlined.
I wanted the original Citystate to be a meaningful experience, one that make you reflect upon political ideals and maybe even broaden and nuance your own beliefs and opinions. But bringing up political topics to the table is a very tricky thing to do, especially when the topic is included and distributed in a mainstream platform such as Steam. The closest comparison I can think of would be like putting up a heavily politicized comic book in the children's aisle of your local supermarket. It's probably not the best place nor the best medium for such an enterprise.
In response to the initial backlash Citystate 1 received because of its political comments, I have considerably tuned down the importance of politics in the second game. The number of angry reviews pointing at the political aspects of the game were consequently much lower, but this also removed some of the fun out of the initial idea. Pivoting from a quirky political sim to a realistic macroeconomics sim certainly leads to a more streamlined, politically correct product but it did harm the initial concept as well.
SimCity 4 vs Cities Skylines
The idea of Citystate being somewhat educational was important to me. Cities Skylines indirectly portrays a very naïve representation of society where bike lanes and bus stops are enough to make a city strive. The player controls more directly everything the city does. There is no invisible hand and as long as you provide exactly the services the population requires the city will do just great. Don't get me wrong, it is one of the best sandboxes ever, but it is definitely not a thought-provoking experience. SimCity 4, in contrast, forces you to see the city not as a simple list of problems waiting to be solved but instead as a chaotic, autonomous organism that needs supervision. The player has to create the right conditions for development to happen, the city grows in a more chaotic and independent manner. The player more often finds himself fine-tuning and experimenting than just solving specific problems once at a time which forces him to build a mental model of the city at a macro level rather than only going through a list of micromanagement tasks.
I think that city builders today are missing complexity and that young players do not learn much from them. My desire to change that is what motivated me to develop Citystate in the first place. This approach is nonetheless risky because it goes against what every other city builder has looked like for the past 10 years.
On release day Citystate II was poorly balanced (the economy tended to spiral into recession every so often…) and riddled with bugs. Surprisingly though the reviews were not so terrible (On day one Citystate 2 had 71% positive reviews while Citystate 1 had 62%).
What is important though and that I did not understand at the time is that bugs and balancing issues were not the main problem with Citystate II.
There is something that we, developers do when we read feedback from players: we tend to take everything very literally. But in most cases, I have come to realize that a negative review pointing at some very specific inconveniences can actually mean something quite different. Even myself as a player would find it difficult to explain exactly why I did not like a certain game. Pointing at technical issues or comparing it to other games can be a way of rationalizing our feelings. But the root of the matter might be that the game is just not fun enough, the core game loops are not captivating us and so our attention shifts too easily to bugs and UI annoyances. I have realized that because reviews have only slightly improved even though most of bugs and annoyances have been fixed since then. None of the 10 major updates and hundreds of bugs fixes have been sufficient to distinctly tip the scales.
For many, it seems that Citystate II is a very counter intuitive experience.
The most challenging part of developing a simulation game is the necessity to have to guess how most people see the world so as to design systems that they can intuitively understand. But this presupposes that the developer consciously makes the effort of adapting the game to the players’ mind in the first place. Here I have to admit that, erroneously, this was not my priority when designing the game mechanics of Citystate II. A recent negative review crystallized this rationale for me (but there are many others very much like this one):
“I played on the hardest difficulty and felt no challenge […] With riots and slums all over, I never had to worry about my budget so I had no real incentive to learn the game mechanics and improve my strategy” (Joshi Kosei)
For a game that is specifically designed around social issues and politics, seeing that a player felt no challenge with slums and riots taking over his city means that the game fails to communicate what it is even all about.
“The game kind of loses its charm when you realize that for the most part the policies are just stat modifiers.” (Arc_1996)
The underlying social and economic system was the number one thing I thought Citystate was excelling at; this is the game’s hook and what sets it apart from every other game in the genre. This comment, a number of similar others, watching playthroughs on YouTube, and a look at players stats as well, convinced me that the core of the game has been generally misunderstood and overlooked. The very central ‘social ladder’ system was disregarded by most. This explains why so many players failed to see any parallel between the 'stats' and how the city looks and grows. My conclusion here is that something was quite wrong about how the game is presented to the player, both internally (UI, tutorial, progression) and externally (trailer, store page).
Another explanation however could simply be that turning macroeconomics into a game is not possible: too complex, intimidating or just plain boring for a broader audience.
City buildings games of the past 10 years have convinced players that a city (or a colony, village, town, …) relies on resource management, production lines or traffic congestion. The existing knowledge players bring with them when they play Citystate indirectly leads them to make a series of incorrect assumptions about how the game works which in turn makes their experience confusing or even frustrating.
To summarize here's a couple of hypotheses about what 'went wrong'.
Hypothesis 1: Game design and quality issues
A realistic city building game with economics and politics could become more mainstream but Citystate II failed to introduce interesting game mechanics. The lack of clear goals, rewards and immediate feedback (https://thinkgamedesign.com/flow-theory-game-design/) prevented players from staying engaged and focused. This is primarily a problem of (or lack thereof) game design.
On the other hand, quality was not as good as I wanted it to be and some of the most basic tools like road construction were too crude and not powerful enough. One would assume that the next Citystate could work perfectly well with the same hardcore economic mechanics but only if they are gamified, and, if the rest of the game meets higher standards. (“The gamification of games” https://newsletter.gamediscover.co/p/publishers-and-their-hidden-steam). However and given that I already had two, moderately successful attempts at adding realism to the city building genre I am now seriously considering changing my approach in a more fundamental way, and this brings us to hypothesis number 2.
Hypothesis 2: A niche too small
It is possible that politics and economics are simply not compatible with classic city building mechanics. Unique management features and deep simulation systems are not the 'revolutionary' approach most players expect or even look for. The vast majority of city building enthusiasts clearly prefer to design, build and customize their city directly rather than tweaking values in a spreadsheet. As far as I can tell and after 6 years in the industry developing city builders, I am now increasingly convinced that the sandbox, player-centered approach prevails over the realistic, old-fashion tycoon experience. Examples that reinforce this hypothesis are everywhere, this is just how games have evolved over the past 20 years: Rollercoaster II (tycoon-like, 2002) VS Planet Coaster (sandbox-like, 2015); Railroad Tycoon III (tycoon-like, 2003) VS Railway Empire (tycoon-like too but a fraction of the sales!, 2018); SimCity 4 (tycoon-like, 2003) VS Cities Skylines (sandbox-like, 2015).
Simulation vs Sandbox
When I had just started to work on Citystate II, I read this interview of Sid Meier (https://www.gamedeveloper.com/pc/analysis-sid-meier-s-key-design-lessons) in which he said: “The computer was the one having the fun, not the player.” At the time I disagreed with Sid’s approach which I thought was opposite to Will Wright’s mantra: “make a toy not a game.” Since then I changed my mind and here's why.
The simulation aspect of Citystate is not what most players are excited about. I surely had fun creating algorithms that mimic real world economics and that reach such a degree of complexity and authenticity. This is possibly the most realistic simulation I know of when it comes to economic development and social issues. Citystate is really the kind of game I would enjoy playing myself, something that compels you to experiment instead of presenting you with a premade to-do-list and straightforward objectives. But in the final analysis, who is having the most fun, the dev or the player?
According to Sid Meier, the more elaborate the simulation is the less interactive the game becomes, prioritizing realism means that much of the controls will be taken away from the player and handed over to the computer instead. I was aware of that risk when developing Citystate II and so I made sure players could set every variable of the game. Let’s take demand as an example: in all other city builders demand is independently generated by the game whereas in Citystate you can precisely set it with the immigration sliders. But the problem as I understand it now is that, in the eyes of the players these controls do not appear as powerful as they really are because they do not have an immediate effect. Opening your borders at 100% does not change the residential demand to 100% in the next second; every change the player makes always has cumulative, long-term effects, the game does not deliver immediate (and gratifying) feedback.
In some cases, players are not even aware they have access to these tools or they just do not understand what they are here for or even what they mean. Some of that is definitely related to UI and tutorialization issues but it is also because people do not want to (or do not know how to) read spreadsheets and flowcharts.
Marketability and long-term support
Another recurrent comment that has been made has to deal with the game’s hook or rather the absence of it. As I said earlier this can be explained by the fact that the game does not communicate well enough but I think this is also because "a city builder with realistic economics" does not make a compelling pitch nor does it succeed to differentiate the game from other similar titles. Even Cities Skylines II calls itself "the most realistic city builder ever […] with deep simulation and a living economy".
The term ‘economy’ itself is probably not even understood as I think it would have. For gamers, the economy refers to the exchange of resources and items. A city builder focused on economics then makes you think that you would micromanage resource extraction and production lines like in Anno or Tropico. But Citystate does quite the opposite of that wherein the economy here is a set of free market forces that the player can regulate in order to determine how wealth is allocated among the population (https://www.investopedia.com/terms/e/economy.asp). This is the contemporary definition of economy, not the Marxist or medieval version of it but that is not how most people see it and that is what matters here. Even when understood in context and for what it actually stands for in Citystate, I don't think this makes for an attractive enough premise.
I see 2 paths forward.
I make another attempt at the “Economic development city builder” idea, this time focusing first on the gamification side of things, with scores, levels, quests and objectives. Trim down the number of economic variables a little, simplify the UI further, come up with an incredible 100-steps-tutorial and write a complete in-game wiki. Then I develop a better road construction tool with the ability to build curved streets along with a more realistic looking terrain.
A lot of the technical issues I encountered with Citystate II happened because I had absolutely no experience with Unity and C# which pushed me early on to cut back features and simplify core systems (modular buildings, flexible zoning, transversal roads, intercity connections, mod support, …) but now I have more experience and a big library of existing code so a lot more things are possible. This approach is safer from a commercial standpoint since it will surely still be of interest for the people that bought the previous games.
I start by implementing construction and customization tools first, making sure everything is super easy to use and mod ready. I'd create a grid less zoning tool and procedurally generated buildings that can be edited and redesigned freely. I'd replace policies with a short list of sliders and remove most of the ‘exotic’ economic variables. This approach would be much more focused on the sandbox and creative features than on the simulation mechanics, making it more of a city painter than a society simulator. This would be much closer to what the majority of players expect to play when they get into a new city builder. The technical side of things here is a little more intimidating but the potential audience is much bigger.
As we've seen, these 2 approaches abide by 2 different theories of game design. Citystate III is a more alike Will Wright’s work, focusing on multi-layer simulation systems (https://www.youtube.com/watch?v=c91IWh4agzU&t=3141s), emergence and the importance of realism, a toy rather than a game. On the other hand, Citystate Metropolis is a more modern approach based on what works today in the gaming industry and aimed at a contemporary, broader audience.
Rather than taking one of these two strategies as is, one would carefully set the cursor somewhere in between and indeed, powerful arguments can be made for both approaches.
On one hand, the release of Cities Skylines 2 will efficiently satiate the market of sandbox-type city builders and there is not much a one-man team can do to compete with a corporation and its decades of experience. Here, Citystate 3 would appear to be a more unique alternative that can more clearly stand out from the competition.
On the other hand, the politics and economics behind Citystate 2 systems have been proven too difficult or even impossible to gamify. These are not ideas most are familiar with and they have at least partially failed to captivate the city building community. The 'hook' of the previous Citystate games have relied too heavily on player preconceptions which in turn created politicized feedback, confused people about the game's identity and eventually pushed it back into a small corner of the market.
So, if you've read until here, you probably have an idea about what direction the next Citystate will tend toward, that is, the Citystate Metropolis approach: sandbox first, simulation second.
A couple of final arguments have put the nail in the coffin: One, I have already created 2 macroeconomic simulators, as imperfect as they are, they are here and available to anyone that wants to experience something different. Two, focusing first on developing powerful construction and customization tools does not prevent deeper systems from being implemented on top later. But more on that very soon!