Sim City is back in 2013!

[Deleted member 13524]

EA won't say a thing because Maxis have been very clear that they want all the guilt to fall upon them.

I don't get why people keep slamming EA despite the fact that Maxis has said over, over and over again that this was all their own choice.

Colonial Marines sucks and the guilt is universally on the developer (Gearbox) and not the publisher (Sega), even though none of them comment on the issue.
Simcity sucks and everyone blames the publisher (EA) despite the developer (Maxis) keeps saying they're responsible for the always-online DRM.
What gives?


Mass Effect 3, Crysis 3, Dead Space 3 are all games without always-online DRM and were all recently published by EA.
How come everyone jumps to the conclusion that EA keeps being the big bad monster that forced Maxis into releasing an always-on DRM infected title and somehow didn't force Bioware, Crytek and Visceral to do the same with their games?

What if.. (shock) the always-online DRM was actually an option made by Maxis?



Nonetheless, with review sites actually lowering their scores after the launch-day review, SimCity is getting a much harsher treatment than Diablo 3 did.

Maybe it's because people inherently like (or used to like..) Blizzard a lot more than Electronic Arts. Maybe it's because Diablo 3 f"#$ked up so much and so many people that everyone (reviewers and buyers alike) became much more cautious over always-online DRM.

To me, there's a clear hipocrisy over EA's faults or lack of them. I'll just call this EA's Hatred Boost (EA'HB).


Nonetheless, I think that for the consumer EA'HB actually came as a good thing.
Blizzard got away with Diablo 3. They actually convinced lots of people that the always-online was an integral part of the game and "it wouldn't be the same game if it was online". This was proved to be a blatant lie (duh) when Diablo 3 was announced for PS3 and PS4 with offline modes.
However, it's clear that EA/Maxis are not getting away with it. The world+dog is slamming EA for Simcity and there's a clear intention on the global gamer community that always-on isn't welcome.

I can't even imagine what will happen to the next game with always-on DRM. Flat zeros on review sites? Crashing pre-orders? 80% sale discounts a week after release?
Just let them try. I dare them!


Either EA deserves all the focused criticism or not, I'm glad that SimCity is being a flop on public opinion, and I'll be even happier if it turns out a commercial flop as well.
The commercial failure of a game that threatens the way I like to play and keep videogames is good news for me.
I'm just sad that this didn't happen to all games that abolished LAN multiplayer in the PC.

 

[Mize]

When the offline patch released by community, i wonder what EA PR people face will look like

Maybe they'll be happy because then I can buy it.

 

[Mize]

[snip...yes it's true we hate EA]

Always on DRM needs to die and if EA has to be the sacrifice, so be it. I can live with DRM...make me login every X days or something, but don't take away the ability to play offline. Laptops are *finally* at the point where they're halfway decent for gaming and more and more planes have AC power and online DRM just destroys what could be a very good thing. In a couple weeks I'm going to Poland and a few weeks after Japan and then a few weeks after China. I would love to be playing a new SimCity for some of those hours en route. As it is now it'll be Civ V, XCOM and Starcraft2...

 

[Grall]

What if.. (shock) the always-online DRM was actually an option made by Maxis?

Speculation of course, without any evidence either way it could just as well be a matter of EA demanding this from maxis. Anyhow, I'm under the impression that EA actually owns maxis, making the issue essentially moot. You can't blame something on someone else, when that someone is part of you.

Also, what's wrong with hating on EA for sketchy-at-best reasons? They effing deserve it and you know it.

Btw Mize, Starcraft 2 can't be played without logging in to battle.net AFAIK, essentially making it always-online as well. Maybe the game tolerates you logging in via wifi at the airport and then dropping the connection while playing skirmish or campaign battles while in the air, I dunno. Just be prepared for disappointment beforehand if I were you...

 

[Mize]

Btw Mize, Starcraft 2 can't be played without logging in to battle.net AFAIK, essentially making it always-online as well. Maybe the game tolerates you logging in via wifi at the airport and then dropping the connection while playing skirmish or campaign battles while in the air, I dunno. Just be prepared for disappointment beforehand if I were you...

:S it's supposed to be playable in guest mode...guess I'd better check.
[grumble grumble...get off my lawn!]


Seems it (offline guest mode) was working and then mysteriously broke after the 1.5 patch with Blizzard saying it's a bug and they're working to fix it...like a month or two ago.

 

[Gerry]

I think that's quite clear though. I don't understand why some may think the server is doing some simulation works.

I guess because that's what Maxis have been telling people.

"We offload a significant amount of the calculations to our servers so that the computations are off the local PCs and are moved into the cloud," Maxis general manager Lucy Bradshaw explained to Polygon. "It wouldn't be possible to make the game offline without a significant amount of engineering work by our team."

Now they're either bullshitting or they're being deliberately vague as to what "calculations" are moved to the server. But the quote above is certainly implying that some core simulation logic is carried out on the EA servers.

 

[Arwin]

Agents seem to 'cheat' in interesting ways.

http://www.neogaf.com/forum/showthread.php?t=523158

 

[Malo]

Reading all the discovered snippets about SC 2013 and watching all those videos on YT is rather depressing. It really looks like a very poorly done simulation. Seems like they had grandiose plans for agent based simulation then realized that they couldn't really do it and made do with what they could get working on a very primitive level.

 

[Grall]

Now they're either bullshitting or they're being deliberately vague as to what "calculations" are moved to the server. But the quote above is certainly implying that some core simulation logic is carried out on the EA servers.

Nothing really seems to be carried out on EA servers, the other day a Kotaku article described how the writer was able to continue playing for 19 minutes with the internet cord disconnected. The only thing that did not work was trading between cities.

 

[pcchen]

Agents seem to 'cheat' in interesting ways.

http://www.neogaf.com/forum/showthread.php?t=523158

I noticed the "population" problem too. In one of my previous post, I complained about the problem of balancing between residential, commercial, and industrial area. One significant issue is the lack of workers despite a large population. For example, one of my city with more than 100k population, has less than 20k workers. Yet, there are more than 20K "job vacancies" unfilled. At the same time, shoppers are complaining about "not enough shops!" while some shops couldn't find any workers.

This could be a bug, but either way, this game requires some serious balancing effort to make it work.

 

[Mize]

I think I'm going to buy SimCity 4 Deluxe.

 

[3dilettante]

The thing about the inflated population numbers is that I would have figured a more scalable SimCity simulation would have done something like that, instead of what the marketing blurbs about simulating every sim said.

A statistically modeled or abstracted simulation could use the physically modeled sim agents to guide the outcome. Over time, the smaller number of active agents would allow results to accumulate over multiple cycles to a good approximation of the desired result, and thus saving processing while also allowing for interesting behavior.

That would have allowed for bigger cities, and I would have thought some of the twitchiness of an agent-based city would be smoothed over by some sanity checking from the stat simulation.
On the other hand, maybe they tried that and found that small twitches in the AI portion would blow up, as it does seem that players are not seeing results that seem commensurate with the agent side's numbers being extrapolated upwards.

edit:
One problem, though, is that the agents have no persistent context between workdays. There's no constant home or work address, and behaviors you'd expect to happen don't because there's no history effect or way to look back on bad outcomes to weight choices differently.

 

[Mize]

Perhaps there aren't resources to have each of hundreds of thousands of agents have constant home and workplaces, etc., but they should at least have assigned areas of the city to travel to and from and only then apply the fill the first vacancy method. It could even be a city area entry side for transport to make things more natural. Adds two data points (2-4 bytes max) to each agent. Then again, 2-4 bytes more than zero is a lot.

 

[3dilettante]

Even for a pretty old system, I don't think the space numbers are that bad with the RAM and HDD capacities listed. In terms of active agents, it's not even required that every travelling agent needs to run every cycle, particularly if the game can determine things like whether nothing signficant has changed for buildings on either side of a commute.
Even without that, just telling the simulation to run with prior results every other cycle probably wouldn't be that bad once a city is large enough to swamp individual changes.

Larger city context might be worse if you decide that client-side storage couldn't be used.

 

[pcchen]

Yeah, I think agents do not take a lot of memory. If you have 64 bytes for each agent, in a city with 1 million inhabitants, still requires only 64MB. Even a mobile phone has enough memory for that.

Also, in real world, most people don't change job daily. They don't move daily. They don't even change how they go to work daily. Some behavior are more random, such as shopping, but that's generally not that frequently. The routes for buses, garbage trucks, school buses, trams, etc. do not have to change unless there are new stops or roads. There's no need to re-calculate routes for each bus every time it leaves the bus terminal.

In its current state, I think this game has potential, but they really need to solve these problems.

 

[Mize]

It's not the memory, but the overhead to move all those bytes and track them whilst doing the rest. As it is a zero byte moves to the nearest vacancy. If I assign four bytes to it (home region and entry side, work region and entry side) then I have to process these for every agent.

 

[Jedi2016]

Which appears to bring us back to the "bigger computers = bigger cities" argument. Allow more powerful systems to create better simulations. Give the game system requirements that are actually current for a mid-range modern game.

I'm not playing on my Dad's computer. I'm playing on my computer. And my computer is a bit of a beast. It could handle significantly more simulation than this game is pushing.

In regards to traffic, it makes me wonder what they changed when they "optimized" it a few patches ago. If anything, whatever they did apparently made it a lot worse, that it didn't always work as "dumb" as it does now.

 

[Arwin]

Yeah, I think agents do not take a lot of memory. If you have 64 bytes for each agent, in a city with 1 million inhabitants, still requires only 64MB. Even a mobile phone has enough memory for that.

Also, in real world, most people don't change job daily. They don't move daily. They don't even change how they go to work daily. Some behavior are more random, such as shopping, but that's generally not that frequently. The routes for buses, garbage trucks, school buses, trams, etc. do not have to change unless there are new stops or roads. There's no need to re-calculate routes for each bus every time it leaves the bus terminal.

In its current state, I think this game has potential, but they really need to solve these problems.

I agree. The route only need to be recalculated when there is a clear issue, and in that case, a standard temporary route could be calculated and implemented by traffic police, to make it even more realistic, or feed it into 'gps' traffic systems.

It must be pretty hard to program something like this, but it still has me itching a little to try and find out for myself.

Also, looks like any remaining feeble arguments that the game needs online have been debunked. Someone already managed to unplug from the network and then continue playing for 15 minutes, and now someone discovered a debug mode, and this allows you to play offline permanently, and just about everything works, and Highways also become editable.

http://www.neogaf.com/forum/showthread.php?t=523776

 

[Grall]

Lying to your customers = egg on your own face...always.

 

[Mize]

This game would be great with real AI and offline play.
sigh.