I figured I'd just create a new topic for this. Here's an interesting update. I'm not sure what their freebie is going to be (I think it would be both fitting and ironic if it were Sim City 4). http://www.simcity.com/en_US/blog/article/a-simcity-update-and-something-for-your-trouble
Spore was not bad... for a $15.00 game. But considering it wasn't a $15.00 game... I'm hoping for Sim City 4, as I almost bought it the other night, but then decided to wait and see what happens with this game.. Just hope it's not spore (or the creature creator). The Creature Creator was actually pretty cool for what it is. I did spend lots and lots of hours with that part of the game.
From what I read it sounded more like it wouldn't be a 'specific' game, but probably one from a choice.
I did get Sim City 4 as my free game (there was a choice between a bunch of different things, including Mass Effect 3, and several others which don't come to mind immediately). But by the time I got it, I kind of lost my desire to play it lol. Honestly, Sim City 2013, as it's been dubbed, had some interesting ideas that were simply implemented extremely poorly. There was more thought put into keeping people from pirating it, than into actually making it a good game. The biggest things were: 1. On-line should have been optional 2. It should have been tested a whole lot more, especially, but not only, stress testing, and they should have been better prepared for what to do should the servers be overtaxed (Trion had the right idea with Rift, in that they already had a contingency plan for what happened, and so although there were problems, they told us the plan right off, and implemented it, and things were fine in a couple of days. 2b. Communication with Maxis/EA is horrible -- they lied, they misled, and they simply destroyed any kind of trust that people might have had in them by doing that.. 3. The Sims AI is horrible, which causes many other problems, such as interminable traffic issues once your city has any kind of significant population. Minor problems include the fact that although they added curved roads, there's no contingency for orienting your buildings to line up with those curved roads, so it's even more inefficient to use them than it normally would be. If you are going to allow curved roads, then you have to also include ways of rotating the buildings and other facilities such that they face that road head-on. While I acknowledge that the size of the city is in keeping with the philosophy of building co-dependent cities in a single region, the size that was chosen is still quite a bit limiting, especially when you consider that some of the city plots already are limited in useful land based on terrain. I didn't think that at first, until I saw how quickly I could fill in the available land. Eventually, I'm sure, some of the problems will be fixed (hopefully traffic and AI issues, in particular, will be addressed). But most, if not all of the limitations are not going to go away. I don't think the game is totally without its potential to have fun with. In spite of the problems, I did have fun with it for a few days. And I may pick it up again in the future, maybe after a few more patches.
If you never played the 4rth, you should seriously give it a spin. Of course it has old isometric 3D and no curves, but it is pretty much everything a city builder should be; A note on SC2013's AI agents : it was a grand idea, but they basically just showed everyone that the idea couldn't really scale. I think with our current technology and hardware, it's difficult to keep any kind of interesting simulation with than several hundreds agents (that's what games like CE and DF aim for). With the glassbox engine, they tried to scale up by reducing the AI to the dumbest thing and fugding the numbers to make it look there's more agents going around your city than there really is. IMO games like sim city should not deal with ai agents until we're capable of scaling it to big numbers and make it behave in an appropriate way. Having a sim in which every agent don't have a fixed home and change job every day, is not really acceptable for a city builder, and the price for having this half assed agent AI is ridiculously small cities. At this point it's preferable not to have agents at all, and just keep everything abstract. It's a shame because I would have paid for a sequel to sim city 4 with the same underlying principles took some steps further
Well, the whole thing can actually scale. It's just that as the scale gets bigger some things change, just like people in villages behave differently than people in towns. We can make a comparison to playing "The Sims" alone versus sharing the game with other players - as the number of potential controllers for one neighbourhood increases, some things have to be changes so that it wouldn't descend into . Even implementing simple multiple-tier threshold values for functions which control the "citizens" would work there, but from what your posts I can take a (likely very accurate) guess that no such thing is there. The AI agents themselves aren't really all that much of a problem even on larger scale if you knew what was happening - it works in "Dwarf Fortress" (admittedly not on a scale as large as is possible in "Sim City 2013", and only if you aren't averse to creating a boring-but-safe settlement [because the point here is lack of chaos, which is nigh-impossible to accomplish if you make your "Dwarf Fortress" settlement "not boring"], but it's close enough), and it works in "Tropico" series (once again, on a smaller scale, and without as many variables and things, but it only means that "complicated" and "in-depth" is not always "better"), and it worked well enough in "Wiggles" (once again, on a smaller scale and with occasional AI hiccups, but the whole thing doesn't start falling into ruin even if you somehow manage to get a more sizeable populace, which is contrary to what we have in "Sim City 2013"). No, the problem appears to be something that indicates their hastiness with releasing the game and thus likely also hastiness with finishing the beta tests without performing them well-enough. Although I do admit that at least part of what I said there comes from a bias - I tend to dislike games with year number put on the end of them (with the exception of the "Anno" series, for a rather obvious reason) because it just screams "okay, so we release one game of the series a year, no matter if it's completed or not" in my opinion. Really, I'd likely like "Sim City 10: Crazy Title Numbering" more than I like "Sim City 2013" as a name for a game (even though I'm not exactly fond of people making sequels with sky-high sequel numbers).
Actually, it's simply called "Sim City" -- people are just adding the 2013 themselves to distinguish it from the original which had the exact same name, as well as from the series in general. Some people are also calling it Sim City 5. It was a dumb idea to call it Sim City. Let's leave it at that. There were bad decisions made all the way down the line, and they are continuing, like the continued insistence that the game operates like an MMO.
Yay for me for not only being biased but also not noticing that the basis on which I formed my bias was a faulty one. I guess that's what comes with not buying non-indie games and not watching TV. But yeah, I agree it is not something which ought to be called "Sim City". Not that it doesn't have the elements the series is known for, but really, with the way it works it's more of a beacon for "social" (I put this word in quotes because I perceive real "social interactions" in games as playing a game with someone due to a conscious decision and not because it is required; sort of like playing an MMO game in itself is not something I perceive as "social interactions" unless one were to form a party with his friends) interactions rather than a video game.
Social interactions in any game are obviously optional. In MMOs, there are even people who join groups, unfortunately, who can be pretty anti-social lol. In this case, there is certainly an optional social component in that you can communicate and interact with other people with cities in the same region. To be fair, there is an optional multi-player component. 1. You can give money 2. you can sell services by sending spare police cars, garbage trucks, ambulances, etc. 3. You can unlock technology for your region by researching it at a university -- unlock it and it's unlocked for everyone 4. You can unlock city services for your region by building the proper improvement to your city hall (for example, build a department of education, and it benefits everyone, not just you) 5. Build a 'Wonder' and it benefits everyone (eg. an International Airport). Wonders are build on special sites, so they don't take up space in your city. But they are also incredibly expensive. 6. There is some interaction and travel between cities in the same region (such as tourism, etc.) if you create sufficient transportation (don't ask me if it works properly). The major reasons why this does NOT make it an MMO is that 1. There's a very limited number of people that you can share a region with (it varies by what region you select, but I think the maximum is like 12 or 16) 2. Mostly, you aren't actually interacting with live people because there's no requirement that while any of this is happening, that any other person is actually on-line. 3. And, in fact, unless it's been fixed (haven't actually been playing lately), none of this happens in real-time. At least at the start, it could take hours or more for the other person to actually receive the money that you give them, as an example. We (meaning every single person in the whole wide world other than EA) don't call games MMOs if you only share a virtual space with a maximum of 16 people. We expect, when we play an MMO, that there's someone else on-line that we can interact with in real-time, and when I do something you see me doing that something almost immediately.
If you're going to give many example of games where AI agents don't actually scale i'm going to have trouble believing that kind of system can actually scale Most of those games get a few hundred agents at best, which is far from the hundreds thousand agent a functionning city simulation would require. It can't scale because simulating agents require tons of processing power we don't currently have in our everyday computers. My not-so-recent core2duo gets hammered by dwarf fortress if I have 200 dwarves running around. So as you imply you cut down on agent complexity to release processing power and thus having more agents, but in simcity they did to the point where simulation didn't have any sense anymore (most prominently, sims looking for jobs and a new home every single sodding day)
I don't know how the agents in Sim City work, but I do believe that they aren't working in a sensible way, because instead of taking traffic into account when plotting a route, it seems like they simply take the most direct route to their destination, leaving one or two main streets with bumper to bumper traffic, and other, equally wide streets fairly empty. What ends up happening is that, because everyone takes the same routes, police, ambulances, fire engines, buses, etc. all get stuck in that same traffic jam (usually). I thought it might have been particularly bad in my city because it was a tourist haven (had a big stadium, and a couple of tourist location) but when I looked on-line to see what I could do, it turned out that almost everyone was having the exact same problem as I was. I actually studied fluid dynamics/mechanics in college, and one of the things we had to model at one point was traffic flow (traffic flow is similar in a lot of ways to fluid flow, but because you have discrete vehicles, there's actually some significant differences, but there' still enough similarities that you can predict that traffic should spread out to go around disruptions, such as traffic jams. (Just an aside -- playing with open channel flow in the lab is real fun, especially on a hot day in southern California).
My point was that these games had one thing in common - as the number of "agents"/"characters"/however you want to call them increased, some things in their behaviour changed. This is harder to notice than I would have wanted it to be, but yes, it really happens. And "Sim City 2013" is different in that regard because there's nothing checking if there aren't any (rather ridiculous) overflows of traffic in certain places. That is, assuming you are capable of making a city for millions of citizens, or tens of millions of them - only then will the real agent count reach tens of thousands. But you do have a point - even normally, when you play you have thousands of agents, perhaps even tens of thousands, so it is not the same as it is for the games I mentioned. Erm, no. It is more difficult to scale because we have to take into account the burden on the CPU, but it is by no means impossible. Dwarf Fortress ravages your CPU when you have hundreds of dwarves not because it's impossible to scale things efficiently, but rather because the game keeps track of everything and then some more, and thus its system requires a lot of computing power in order to keep the game from lagging. But then again, the system used in Dwarf Fortress checks every possible dwarf and every possible thing that is being interacted with by any dwarf or any other thing that can be interacted with, during every step of "gameplay". That is different in "Sim City 2013" due to the fact that you don't need to check every possible thing every frame - no, you merely have to render agents every frame, but calculations pertaining to their behaviour can be called as needed because there aren't any Dwarf Fortress-esque parameters that the game would have to check. And a simple "can I go this street or do I try another route because this one is too busy" calculation (I know it is not "simple", but it is arguably simpler than many calculations performed in games) takes much less processing power than a "what sort of material, out of many available types, do I take for this sculpture, assuming the material is there, and which way should I go to take it, assuming that I do care about working instead of taking a break to drink some more booze or eat a dwarven bread, a check of which is also to be made, and how will the booze affect the dwarf unless there is no booze and how will walking affect the dwarf unless there is no walking unless there is and how will carrying affect the dwarf unless what is being carried weights nothing and how will the material affect the sculpture, unless the sculpture doesn't exist which is to be checked" one (which is what we get in Dwarf Fortress, every frame, for every dwarf). Now, that is not to say their model can't be saved. Because it can. But in order for that to happen, they have to admit they botched the whole thing and start working on fixing it seriously.
I'll just leave that here : http://imgur.com/a/gW7F9 'Simulation' must means something else entirely at Maxis-EA.
The man managed to do this because for some reasons, 0 tax = infinite happiness. I guess It's in the realm of "fixable" issues, but show just how shallow the underlying systems are in that game.
I know, Kamisma. My point is (obviously) that it doesn't (or at least shouldn't) work that way, and it shouldn't be happening because that's one of the "basic" situations for testing (with "basic" meaning that instead of having to do some complicated combinations, they merely had to turn one value to its extreme and watch what was happening) and thus it working like that means that either it was a deliberate decisions (erm, "WHAT?") or their beta-testing process needs alpha-testing.
