The connection problems of SimCity's botched launch may almost be behind us, but now that more players are actually playing, the critical bombardment has adjusted its aim to target the simulation itself. Players are reporting bugs, quirks, and mysterious behaviors, and discontent has swelled into accusations. Our review criticizes many of these problems, but is there a bigger story? Were we misled?
Posing that question without answering it would just be rhetoric, so I've run my own experiments to directly test the biggest claims. In addition to judging their accuracy, I've provided context to inform judgment on whether we were misinformed by our own assumptions or by Maxis itself, as well as my opinions on the individual controversies.
[Update] Just after this article was written, SimCity Lead Designer Stone Librande published a post which explains more about some of these issues and the design decisions behind them. He also promises a patch to car pathfinding to help alleviate traffic jams.
If you cut off your internet connection while playing SimCity, it doesn't dump you back to the launcher. It warns you, but continues to simulate your city normally until it decides you've been mucking around offline for too long and boots you. But now a modder claims that he enabled offline play for longer than the 20 minute limit by tweaking the game's package files. (The video shows debug mode, which I'll get to later—the claim appears in the description.)
I haven't been able to test that claim yet, but it isn't necessary—20 minutes is long enough to establish that SimCity can simulate a single city locally. So where are the "portions of the computing" that Maxis GM Lucy Bradshaw claimed happen on EA's servers? Most likely, Bradshaw was referring to the common pool of regional data cities draw from, which informs interactions like commuting workers and shoppers. From what I can tell on the front-end, it's essentially shared cloud saving.
There's also the Global Market (a shared commodities exchange), leaderboards, achievements, and other standard online game niceties, but there is no evidence that the core single-player experience can't be replicated offline for longer than the connection timer allows.
Maxis seems to be leaving out the caveat: you can't play SimCity offline as it was intended , and it's intended to be a connected experience. This is an unfortunate stance, because as I stated in my review, the bulk of my fun was a solo experience.
It is true that Sims can go to a different job each morning and return to a different house each night, and that they seem to seek the nearest instance of a building type to fulfill their desires. In a way, it's an elegant solution—with little processing per Sim, it encourages the creation of wealth-segmented neighborhoods which satisfy multiple desires. At higher populations, however, it can abruptly lead to a traffic gridlock and other unintended behaviors, something I criticized in our review.
Sleuths who have done their research already know there's no conspiracy here, because Maxis has said from the start that this is how Sims work. In fact, the May 2012 issue of PC Gamer US contains our first feature on SimCity, and in it I wrote:
"These Sims aren't as complex as the families we love to torture in The Sims. They start each day at the top of a flowchart, asking a series of simple questions such as: 'Am I sick? Do I have shopping money? Do I need to find a job? If there aren't any jobs, is there a park to sit in?'"
This was repeated and elaborated on in many subsequent previews and interviews with developers, both here and at other outlets. We were never misled about Sim behavior.
After 500 residents, every house of six adds about .5 people to a city's total population count. That ratio seems to increase with population, so that at 200,000 residents, somewhere around 20,000 appear to actually be simulated Sims.
To test this, I built a city containing exactly 90 low-wealth houses. Each house adds six Sims to the population, so 90 houses should give me a population of 540. The reported population, however, was 583—a discrepancy of 43, almost exactly half the number of houses. The community is calling these mysterious population padders "phantom Sims."
The population chart supports this observation. If you add up the total number of reported workers, shoppers, students, and homeless in a city, you get the correct number—in my 90 house town, the total was 140. That means the "total population" is likely flavor text which is meant to represent the fiction of the city—perhaps it accounts for children, who aren't otherwise factored in.
Because SimCity was sold largely on the power of its simulation engine, this should have been explicitly stated, but it's not actively hidden and doesn't change my overall appraisal of the game. SimCity is a game—its job is to create the illusion of a city, not to achieve a perfect 1-to-1 simulation.
Our assumption that the population count was part of that simulation and not a fictional game element was justified, but I haven't seen any marketing which explicitly states that. It should have been stated to avoid the apparent deception, but our assumptions aren't required features.
Most likely, things don't work so well with 100,000 individual Sims, and I think most would have accepted that if it were stated up front. After all, we accepted that our SimCity 4 cities didn't actually have 1,000,000 residents, or any at all, except as a number. It's a shame this wasn't communicated better, or it could have been a non-issue.
In the process of discovering the population number padding, experimenters discovered more quirks and made more claims. This Reddit poster identifies quirks in how death is handled by experimenting in a one-house town, but misidentifies the consequences.
It is true that destroying a building can cause some of the Sims in that building to disappear. In my own one-house town, I destroyed my power plant, "killing" three Sims and leaving a household with three residents while the game still reported a population of six. The poster feared that this could be a bug which seriously affects the integrity of the simulation, but it actually rights itself naturally, suggesting it's just a design shortcut.
SimCity bases population reporting on the number of residences available, not the number of Sims, because residences will always be refilled throughout the course of regular play. If I do absolutely nothing, my city will go on with only three out of the potential six Sims living in my city, but then I wouldn't be playing the game. Any time new Sims come to my town, they fill vacant housing, and it happens all the time. If Sims move into a different, new house, non-full households will also be filled. Even if you don't zone any more residential, non-full households will run out of money and eventually be abandoned, which also initiates repopulation.
At higher populations, Sims come and go all the time, making the effects of killing Sims hard to identify. If you want to avoid possible complications, turn off plopped buildings before demolishing them. I have not been able to replicate the claim (made in the same Reddit post) that Sims stay in powered-off buildings for any more than a couple seconds.
This claim has been made in several responses to the discovery of non-simulated population: balancing residential zoning to commercial and industrial zoning is impossible because the phantom Sims are mucking it all up. I'm calling this one false with an asterisk, *because I can't prove it without seeing the game's programming. Based on my experiments, however, there is no evidence to indicate that the fictional population count has any effect on the Sims in your city.
Instead, I've identified three main reasons the worker to jobs ratio is difficult to balance.
1. The "demand" bars are not informed by the workforce. Zoning more of what's in demand—usually residential and industrial, often simultaneously—is not a good way to proceed, and this is something I criticized in the review.
2. Balancing land value is the real trick. Knowing that a basic, low-wealth house adds six Sims to a city, it's easy to match the number of low-wealth residents to the number of low-wealth jobs. However, because you need medium-wealth Sims, you have to build parks and services to increase land value. Land value isn't precise, making containing high, medium, and low-wealth neighborhoods a challenge.
3. Density increases must be carefully managed. If you start off zoning high density roads everywhere, you're going to have a problem. Building density must be increased carefully so you don't end up with a bunch of apartment buildings feeding excess workers to small shops and factories.
I've criticized the always-online requirement for effectively preventing modding, but I may have spoken too soon. It's shocking that, despite the danger to public regions and the potential for exploits to affect the leaderboards, players aren't prevented from logging in with modified package files.
The first success of the SimCity modding community is DebugEnable.package, a file (may violate EA's TOS—use at your own risk) which enables debug mode, allowing players to draw regional highways. It seems that, even in public regions, these permanent debug changes are saved to the server. That can't be allowed, right? It probably isn't, but Creative Director Ocean Quigley did condone modding in a tweet last month:
@ simcitymodding @ moskow23 I say go for it, but it'll be a little while before we can provide any support.— Ocean Quigley (@oceanquigley) February 7, 2013
That doesn't mean you should assume it's no risk. Modding SimCity may still violate EA's TOS, and pointing to Quigley's permission probably won't help. He did say the game is "built to be moddable," so there's a chance it'll be officially supported in the future.
*Of course, I acknowledge that this is an opinion, but I have question to ask, and it might sting a bit. I'm having a lot of fun dissecting SimCity. I'm disappointed when Sims don't behave logically, but I take it as a challenge, and when my 200k pop city is flooded by a sudden deluge of traffic from the region, I spend hours trying to fix it. I want to know how it all works, why it works that way, and how I can exploit it.
I'm glad we're talking about it, but it's not what every SimCity player is talking about. I know other players who just like making fun, creative city designs. They don't care to dig into the simulation's gears and spin them around to see what they do, or whether or not they can achieve perfect stability while maximizing density. That kind of player probably doesn't notice or isn't bothered by many of these issues, except the always-online problem. So my question is: was SimCity made for us?
I don't think it was designed to be min-maxed or intentionally tested for flaws with contrived gameplay scenarios. I'm having fun doing all that anyway, but I wonder if we're missing the experience it was designed for by spending our time trying to break it. The part of it that's a game.
I've concluded that it's not broken—not wholly, at least. If it were wholly broken, I don't think I could have had the number of positive experiences I've had with it, and it's not an appliance that's supposed to function in one specific way. It's a game—an experience generator—and a lot of those experiences are good. Sometimes players play in totally unpredictable ways, and games can't always support every approach. SimCity supports a more diverse set of experience-seekers than many other games, and in that, I think it succeeded.