It's tangential to the article, but if you like(d) SimCity, you should definitely check out Cities: Skylines [0][1], it's a really well designed game.
Another semi-related thing, I'm curious if/when we'll live in a time when every decision is run on some world simulation before being "implemented". Does anybody know how often/if Expert Systems[2] are used in politics/business? How would you approach designing such "simulation", what would be the hardest problems?
Don't get me wrong, Skylines is basically a good game to me, but as someone that has been playing the "Sim" series of games since the original SimCity, I think they missed many of the lessons learned by those games and really could have released a much better game.
Most of my qualms with Skylines are sort of "nitpicking", but they are pretty glaring:
1) the annoying tweet bird with no option to disable (at
last check) because the devs felt it was central to their
design
2) terrible UI graphics (I know this is subjective,
but...) seriously; the UI graphics are not very
polished compared to SimCity 2000 or SimCity 3 even
and look like they were made by a programmer
3) funky road placement; the road placement tools have a
lot of the same "floaty" feeling that SimCity "5" did that
has lead to much frustration (see the skylines subreddit
for various complaints about this)
Personally, I wish someone would create a modern version of SimCity 2000 instead using a voxel engine. I miss the precision of placement that it had and the simplicity of various systems.
With that said, they did get a lot of things "right" as far as I can tell, such as traffic simulation, map size, and general performance (mostly thanks to Unity really....) and their friendliness to mods, no extra drm, etc.
I agree that they have a lot of things to improve, yet I'd argue that Cities: Skylines is the best city simulation so far (disclaimer: I loved SC3000, I hated SC4 and 2013). But maybe I'll change my mind, I'm playing Skylines just to relax once in a while so I haven't had a chance to hate it yet ;).
One thing that's great is the community (and Colossal/Paradox attitude towards people that mod the game). The game is pretty new and there is already tons of different mods, it'll be interesting to watch how this evolves over time. So, you say that you don't like Chirpy - you may be interested in this mod: https://steamcommunity.com/sharedfiles/filedetails/?id=40579... (fun name, isn't it?). AFAIK, there are already multiple communities that build mods, for example: https://www.reddit.com/r/CitiesSkylinesModding
I was of a mixed mind with SimCity 4; there were certain things about it that were much better and other things not so much.
I enjoyed the Rush Hour expansion more than the original SC4. One thing that SC4 did extremely well though was music and general presentation, although I'll admit to missing much of the original composer's music for SimCity 3000/2000.
SimCity 2000 had a lot of great mechanics but a fairly terrible interface -- somewhat understandable given its original incarnation as a dos game and certain UI mechanics hadn't come along yet (such as being able to use the scroll wheel or panning around).
But I'll always fondly remember when I upgraded to a SoundBlaster AWE32 just so I could get "real midi" music from SimCity 2000 :-)
Mods fix most of your complaints, but regardless, I completely agree with the gist of what you're saying.
As good as Skylines is, the game isn't quite Maxis caliber. SimCity (2013) had incredible UI and UX design. It's just a shame that it was marred by tiny city sizes, overemphasis on simplicity, and online services dependency.
That said, Colossal Order's budget probably wasn't Maxis-caliber either, so I'm hoping the revenue from Skylines will allow them to be more ambitious moving forward, and make something truly outstanding (as opposed to merely awesome).
Sorry, but I've never agreed with "mods are the fixes" approach. I evaluate products (which is what this is) based on how the creator has released them -- not based on modifications others provide.
In addition, mods aren't a great option since doing so marks you as a "cheater" or disables achievements, etc.
I generally agree with your conclusions about SimCity (2013) though; generally great UI, but tiny city sizes, online services dependency (which is greatly mitigated now as of the "final" patch), and some of the "streamlining" done to the game.
Eh, ultimately it's a game, and games are about fun -- not usually about fully accurate simulation.
When the two can go hand-in-hand great, but I'll pick fun over accuracy every time for a game.
Now with that said, if your purpose is to build an accurate simulation -- go for it! But SimCity has always been a game first and foremost to me with a quasi-interesting simulation.
Most of these kinds of simulations aren't expert systems but microsimulations (cf. http://en.wikipedia.org/wiki/Microsimulation) and they are actually used quite a lot for certain kinds of problems, like figuring out the indirect effects of changes in taxation. It is often possible to check baseline accuracy (does it at least work in current conditions?) but as for predictive accuracy, well...
Yeah, it's incredible. But AFAIK, it's not only cars, agents are everywhere in this game. If I recall correctly, latest SimCity also was focused on agents, but it seems Skylines devs are better ;). Good discussion: https://www.reddit.com/r/Games/comments/2rm5k2/cities_skylin...
SimCity used agents in places they're totally inappropriate though. Like power generation - that was simulated as agents which travelled along powerlines and jumped into the nearest empty building each tick.
Which of course utterly fails to capture any of the detail of an electrical grid.
The article seems phony. It claims things that aren't possible in SimCity, like:
> took money from SimProvidence’s budget and gave it to folks living in low-income housing to allow them to upgrade their living situation.
SimCity doesn't have any model of "low-income housing" "giving money to folks". Money is only spent on police, fire department, and transportation, all city-wide.
Strategy in SimCity & Cities: Skylines wouldn't survive politics.
The games greatly reward spending all of your money as quickly as possible, and deficit-spending on services until your city grows into positive income. In other words, spend as much money as quickly as possible building and supporting zones, since they make money; then, even though your budget is negative while you spend on the services to support those zones, you will have not wasted precious dollars on services you can pay for in deficit.
This is a really tricky part of the game, and extremely nerve wrecking. If you neglect to spend exactly right, you end up with $0 and no way to fix a mistake once you hit the play button. On the other hand, your city will grow as quickly as it can given your budget.
You could say this is a really accurate model of real life. If you have an efficient way to spend money, should you spend all of it? You could say that governments rarely have efficient ways to spend money, but they always have a most-efficient-among-the-options way to spend it.
If there's anything to take away from the simulation, it's that cash surpluses don't make sense. Not for governments, not for Apple, not for virtual mayors or StarCraft commanders, not for anyone.
A cash surplus is a symptom of playing the game wrong, not a problem with the game. And in our politics of austerity, that's really crazy.
If you have good investment opportunities, you should not sit on cash. If good investment opportunities are coming soon, you should sit on cash. The trouble is that many forms of wasteful consumption are disguised as investment, in politics.
The biggest unrealism factor in all city sim games is that building happens instantly. I suppose it has to be there just to make them playable but I'd love to see a city sim game where building proposals get stuck in committee for years, challenged via environmental review, cut down in size and scope and eventually built as a compromised version.
Part of the difficulty of urban planning is that the feedback loops are so long. It requires you to plan 5 - 30 years into the future and mistakes are hideously expensive. Finding a way to model that would add a new dimension of depth to city sims IMHO.
This is absolutely correct, but the underlying assumption is that the player can play the game perfectly.
This just isn't possible in a world of infinite variability. Human-controlled centralized management cannot outperform a decentralized, non-planned, ad-hoc market in these cases.
It's underlying assumption that the best way to turn your city from a trailer park to a leafy green paradise is an extremely regressive tax structure always struck me as the most politically questionable part of the simulation.
Something similar was done by journalists for the UK 2015 General Election. They took Democracy 3, an indie game with a very complex neural net simulation, and attempted to find out how the various main parties' policies held up: http://www.newstatesman.com/writers/116017
Another semi-related thing, I'm curious if/when we'll live in a time when every decision is run on some world simulation before being "implemented". Does anybody know how often/if Expert Systems[2] are used in politics/business? How would you approach designing such "simulation", what would be the hardest problems?
[0] http://www.citiesskylines.com/
[1] Review: https://www.youtube.com/watch?v=ZJb7C2OE9xM
[2] https://en.wikipedia.org/wiki/Expert_system