SimCity is back, gaming is dead, RIP gaming.

[tgr]

So I've just come back from almost a week of work-related travel suck, and the first thread I see when I get back here is this thread.

All I can say is: Aahahahahahahhahahahahahahahaha

[Abelian75]

To me the AI problems smack of making a decision, getting really invested in it, and refusing to reconsider it once it proves to be unworkable.

Yeah, absolutely.  Despite what some are saying about how it's lazy programming and such, the problem of simulating the pathing for an entire population, with collision, with an environment that can change at any time, is incredibly hard.  Like, fuck, I wouldn't want to touch that shit.  And that should have been pretty obvious pretty quickly.  I'm sure the decision was made at a high level due to the popularity of The Sims and such, but if you have to sacrifice citizens actually having a static home and place of work, then I think it would be obvious to anyone that it isn't worth it.  Pretty blatant disconnect between the people making the decrees and the people able to talk about what is actual possible/working (which is obviously a common problem, but seems extremely bad here).

[Samwise]

I don't really buy the argument that the "simulate everything" approach falls down because of computational limits.  Computers are fucking fast these days, and simulation of a bunch of dudes spread over a large area is a perfect application for multiple cores/CPUs, which are usually underutilized despite being fairly ubiquitous in modern PCs.

I did a quick Google for "crowd pathfinding" and found this guy who says that his system works with a million NPCs on an arbitrarily complex map, and that it does things that Sim City apparently doesn't, like have the NPCs avoid congestion (duuuuuhhhh).  And that's something one guy just knocked together to play with after reading a paper on pathfinding that he thought was cool; there's probably plenty of room to optimize and expand it if you threw a development team at it.  Most of the other stuff a Sim does is very simple from a computational perspective; managing a few desire sliders and a queue of actions and whatever is not like cracking the Enigma code.

This whole thing reeks of dev/design clusterfuck that was rushed out the door.  It's not that they were too ambitious, or too lazy, it's that they had a bunch of conflicting requirements and internal fuckery combined with an unrealistic deadline.

[Abelian75]

Those NPCs are all going to the same place, though.

[Samwise]

As far as congestion avoidance goes that doesn't matter at all, because the density of each section of road is still common to all of the NPCs regardless of whether they're all going to the same place.

For the specific application of SimCity there are lots of ways that you could optimize -- at different times of day sims are going to different types of buildings, so you build path maps to those buildings (which is going to be a relatively small subset of your entire city) and all the sims going to a particular building share the same map.  You also don't need to update the traffic maps every millisecond, because in real life people don't react instantly to traffic conditions; doing a pass every few seconds is going to be more than good enough to get very lifelike behavior.  And again, you can parallelize the shit out of all that.

This is all armchair commentary of course and I'm sure they had all kinds of hassles and requirements to deal with; I'm just saying it's not some giant fundamentally uncrackable nut.

[tmp]

I assume that the Tropico guy is driven by the simulation rather than driving it. Or put another way, his actions are an expression of the underlying simulation rather than the definition.

It's been a while I played Tropico, but it's the opposite from what I remember -- the overall state of the city/country is determined by performance/actions of individual agents. Mind you, the populations don't exceed few hundred 'people', but that's in game made literally over ten years ago and running comfortably on machines of that time.

[Morat20]

Yeah, I call bullshit on it -- it's lazy agent programming. (It's possible -- probable, in fact -- that they wanted to avoid emergent behavior and keep the Sims doing what they 'should' as simulation agents). That first empty slot shit is absolute BS.

You don't solve the problem that way unless you're out of time to solve it at all (and given how core it is, they must have put it the fuck off) or you ran into insurmountable problems -- which should have been identified early enough to find a much better solution than a hack like that.

Hell, store 'favorite routes' and only reroute if congestion is particularly high for a length of time -- then you only have to pathfind a fraction of the time. (otherwise it follows it's fastest/shortest/most scenic/whatever route). And if it has to recalculate, trigger a work/home assesment -- bad traffic is a often a reason people move or change jobs, and they can either move or pick an alternate route.

Pathfinding isn't that hard, you're not looking for optimal paths --- just good ones.

[Samwise]

The fact that everyone and their brother is able to come up with better designs off the top of their head goes back to them clearly just running out of time and having to rush something the fuck out the door.  Which, again, I'm guessing is just horrible management more than anything.

[Sjofn]

I assume that the Tropico guy is driven by the simulation rather than driving it. Or put another way, his actions are an expression of the underlying simulation rather than the definition.

It's been a while I played Tropico, but it's the opposite from what I remember -- the overall state of the city/country is determined by performance/actions of individual agents. Mind you, the populations don't exceed few hundred 'people', but that's in game made literally over ten years ago and running comfortably on machines of that time.

It's like that in CotN too. Your city depends on your workers getting to where they need to be, including getting to places where needs (goods and worship, mostly) are filled.

If doing something like CotN or Tropico would melt down mighty supercomputers if done on a SC4-scale, then they should've ditched the idea rather than city size, since apparently people that aren't me play SimCity to build gigantic fucking cities (I like smaller cities, always have!). Instead, they cut back on city-size ... and then still failed to do the Tropico-thing. It's like the worst of both worlds.

[Paelos]

I do like being able to follow around people in tropico, figure out if they like me, and then have them arrested if they are bitching about my environmental policy.

[Samprimary]

jesus fucking CHRIST this game is fucked.

when work gets out in the evening, every sim leaving work automatically starts to head towards the nearest house with capacity. Even if that house gets filled to capacity in the interim, they'll still drive all the way there before going all "hurr oops guess i better walk over here now"

they have no place of residence. no memory. no goals. no established residence or place of work.

traffic is fucked. left turn lanes don't exist in simworld. entire cities have been ruined by the inability of a major intersection to permit left turns.

game's fucked. no sympathy for people who preordered it.

[Sky]

The fact that everyone and their brother is able to come up with better designs off the top of their head goes back to them clearly just running out of time and having to rush something the fuck out the door.  Which, again, I'm guessing is just horrible management more than anything.

EA: Too big to succeed.

[Ingmar]

The fact that everyone and their brother is able to come up with better designs off the top of their head goes back to them clearly just running out of time and having to rush something the fuck out the door.  Which, again, I'm guessing is just horrible management more than anything.

EA: Too big to succeed.

Maxis has to have a lot of the blame here as well. Most maybe.

[Soln]

I got nothin except to chime in with Margalis nailed it.  Mo popcorn.

[Maledict]

There are already games like Tropico and CotN where there are little people doing their little people things. You can follow a dude in Tropico 4's life, watch him grow up, decide to go to high school, graduate, get a high-school-requiring-job, look at his family (who all have histories and skills of their own), etc.

I assume that the Tropico guy is driven by the simulation rather than driving it. Or put another way, his actions are an expression of the underlying simulation rather than the definition.

If you have a working simulation based on systemic models you can populate your world with actors that jive with the results of the simulation. The problem with Sim City is that there are no underlying systems separate from the actions of the individual actors.

Not from my understanding of the game, no. Immigrants turn up at the docks with a set of skills, likes &'dislikes and possible family. If there's an opening that matches those they will take it, but if not the become unemployed and then criminals. The type of  job they will take depends on their education level and skills, and that then determines the type of housing they can afford. It is possible through education to improve their education level so they can take on different jobs, but this can cause you problems as well - provide education to everyone on your island and you'll have no-one to run your farms, mines and transport agencies that require unskilled labor.

Tl;dr - no, the individual actors in Tropico are generated and you then have to build your island around them to make it successful. Whilst they can change over time based on your islands facilities, they are not just a reflection of the state of the island but independant actors with needs.

[tgr]

Last I looked, which admittedly was in a let's play tropico 4 deal, there weren't that many "people" in your island, so there's not that much to simulate there.

[Sjofn]

How many do you consider "many?"

You could do something like SimCity Societies, where you have your little actors but the population at large is abstracted. So a house will spawn two Sims to go do their actor shit, but it also represents forty people in population or whatever.

Basically I'm not sure people really expected to have populations of thousands and thousands of people simulated at once, but once they saw how small cities were going to be, I think they assumed they'd at least be getting what existed in Tropico 1, a game from freaking 2001, rather than a bunch of hobos that are SO stupid it breaks the game.

[Maledict]

No, there aren't. Tropico absolutely deals on a much smaller, micro level than Sim city.

But the issue is that SimCity promised Tropico style sims - and completely failed on that. At the same time, they broke the macro level stuff as well that SimCity is based around so ultimately you get the worse of both worlds - you neither have the intricate, small scale detail to follow like in Tropico and as was promised for SimCity, and you also don't have the grand macro level city design that is the essence of SimCity.

Add on top of that the fact the fundamental systems in the game are completely broken and you're left with easily the biggest turkey of the last few years (since Elemental in fact!). How this was missed interviews is utterly beyond me - even more so than usual this is inexcusable.

[tgr]

What I would consider many would certainly exceed a few hundred, or even a few thousand. As I said, I've just watched a tropico 4 let's play to see what I was missing out on due to the account requirement, so I'm not sure how high up that goes, but I don't remember noticing the population number going too far above 150-200.

But just going back a bit on the reviews a bit, I find it interesting that there are so many things becoming blatantly apparent which apparently a lot (or all?) of the reviewers missed, beyond just the "hurr always online" requirement.

[Maledict]

I could understand reviewers missing PS3 Skyrim's faults initially - they only became apparent 25+ hours in, and so initial scores didn't detect them. They should have gone back and adjusted the scores once that issue was found, but the numbers made sense.

I could understand reviewers not getting the ME3 ending - either because they didn't play it to completion (again, 25+ hours required) or because they hadn't played the other games and so didnt understand what a fuck up it was.

But what I cannot understand is reviewers missing the fundamental broken aspects of this game. furthermore, to take Polygon's review in poin:

If addiction is a freight train, then SimCity is the roaring locomotive pulling you into the night.
You may play a hundred hours of this game without noticing, but behind the curtain, SimCity is working furiously to hook you — and keep you hooked. And in my case, it was enormously successful at both.

From the pleasing sounds of every various button press, to the satisfying way various parts of your city connect, then come to life (then die and come back from the dead), every element of this game has been perfectly and patiently engineered to engender an endorphin rush of accomplishment.

ONCE I SAW UNDER THE HOOD OF ALL THAT SIMCITY ALLOWS YOU TO DO, IT WAS HARD TO RESTRAIN THE URGE TO BECOME A MAD DICTATOR

In comparing my experiences in SimCity to those of my colleagues, I noticed a strange thing: The cities we'd constructed (and ruined) sounded remarkably like cities in our real world.

This is where SimCity's "always on" connected play shows its strength. Cities in the same region can share services like police and dump trucks back and forth and even gift resources and simoleons.

SimCity is a near-perfect fusion of the classic simulation game with modern social and online play elements. It is in every way the fully realized evolution of the franchise and a much welcome iteration, perfectly engineered to dispense the maximum amount of fun in the most efficient way possible. It is highly addicting, but there are worse things to be addicted to. Just be sure to set an alarm.

Russ Pitts, the guy who wrote that review, is a corrupt con artist who lies to the public for cash. That's the only reasonable explanation behind that review.

[koro]

Russ Pitts, the guy who wrote that review...

Russ Pitts

This is my surprised face: 

[Samprimary]

whsdlkgh

So with a little bit of package editing within SimCity, and a little playing about in the code, it's possible to enable debug mode. I linked the activation to the "Help Center" button in the main menu for ease. Most debug features are disabled without having an actual developer's build (they have terraforming tools etc. available in the full developer build!), but a few things do still work - including editing the main highways.

Not only that - but you can edit the highways ANYWHERE - even outside of your city boundary... and even if you quit the game and log back in later, it's all saved safely on the server.

This shows that highway editing will be easily possible, AND that editing outside of the artificially small city boundaries should be very viable too.

Other things I have modded out with a quick change: Unlimited time to remain disconnected (won't get booted at 20 minutes, can now be disconnected "forever"). Population count now shows REAL figure, not the "artificially inflated" figure. My large cities have a population of about 15k now, not 100k :P

http://www.youtube.com/watch?feature=player_embedded&v=Bmce9oIxJag

[tgr]

Um. Is it just me or is that city size 1/5th or something of the biggest SC4 city?

[Samprimary]

ain't you. sim city 5 limits you to a miserably small parcel of land.

[Fabricated]

Also, not only is the game entirely capable of working offline and letting you have large city sizes, but here's how the guy who modded the game supposedly did it.

"The way I did it involved taking a hex editor to SimCity_Game.package and replacing every instance of DEBUG=!1 with DEBUG=!0."

lol.

[Jeff Kelly]

It's not just sites of questionable reputation like Polygon.

I listened to Giant Bomb's podcast yesterday and they devoted a half hour to the Sim City 5 thing.

Mind you this was after the story broke that the game was basically unplayable and most of the info you all mentioned in this thread was already out.

The crtique was pretty mellow. It focused mostly on the bad communication by Maxis after the game shit itself and died and less on the fact that the game is broken beyond repair and still largely unplayable for most customers. At some point during the discussion they even discussed if the customer is to blame for that because everybody is buying shit on release day and acually expects to play it when everybody (read: every gamer nerd) knows servers will crap out at launch. They even denied that this has anything to do with EA's shitty management style ('people should get over the whole "EA is evil thing"') because in their reasoning 'this benefits nobody not even EA'.

This was the third or fourth time in a year that those guys discussed a crappy launch or EA fuckup and every time they basically sided with the developers (not the publisher) over the customers. They did it when they discussed Mass Effect 3 (even after they all had played it through). They did it with Sim City and on a few other occassions.

I like those guys and they are probably beyond reproach which makes this whole situation even more aggravating than if it only were hired mouths spouting off marketing bullshit. This is endemic in gaming journalism today.

With the whole ME3 deal people who complained about the end were 'entitled' and 'spoiled' and the whole deal was framed as an 'internet nerd fight' by 'disgruntled fans'. If EA would change the end it would be the end for authorship.

With Maxis the majority opinion by gaming journalists seems to be 'why do people expect to play on launch week, everybody knows the servers will be down'. Nobody focuses on the broken game mechanics, they all seem to expect that the servers will be fixed at some point and that this will be a great game then. As if anybody cares (except the people who spent their hard earned cash on a broken game) four weeks from now.

Nobody gave Bethesda flak because Skyrim on PS3 is so broken that they can't even release all add ons and that they can't fix it.

Nobody focuses on the culture of blatant lies and deception that seems to be endemic in the industry today. I could probably name several occasions where the PR or executive people by Bioware or Maxis have lied to the press and the customers to show their games in a better light. The whole 'let the game journos play the game on private servers before launch' thing reeks of a PR campaign that wants to sell the game as a great experience before everybody realizes the servers can't handle it.

The sites seem to be entirely sympathetic to the developers and won't even acknowledge that EA's management might have anything to do with those situations (since it mostly seem to be EA development houses and franchises that have serious problems these days). They also deny that this is basically a DRM strategy gone bad.
 
Through all this people's confidence in the quality of game releases and the quality of reporting about those releases falters. The basic recommendation even here seems to be 'everything a developer or publisher tells you about a game is bullshit and patently untrue, don't believe it' and also as corollary 'everything most game sites and game mags tell you is bullshit so don't believe them either' and as a result 'wait for at least two weeks after release to see just how much they fucked it up, better yet buy it half a year from now at reduced price'

This is toxic on so many levels that this should basically raise all the red flags everywhere, yet everybody seems to be utterly nonchalant about the whole set of issues. People have been burned so many times by so many different publishers that this has to have repercussions for future sales and it does. That so few people in the industry (either developers, publishers or journalists) seem to actually care and that there seems to be a subconscious disdain for the people buying those games is at least baffling to me.

Journalists: 'Even though we don't tell them because we're utterly dependent on the publishers' advertising money, gamers should already know that most games are buggy as hell at launch so if they are so dumb and buy them anyway then they don't deserve any better'.

Publishers: 'Let's push this game out of the door even though it's only half finished. By the time people notice we will have sold millions already and there is no lemon law that forces us to refund broken shit anyway'

Developers: 'The publishers threatened to shoot those kittens if we didn't agree to ship the game now, we didn't have any choice'

A viable business strategy this is not

The consequence for me is clear. I simply won't buy anything at release anymore. Instead I'll wait for at least a month or until it's at half price or game of the year edition with all add ons. With the overabundance of entertainment I won't even miss anything except the heartache and shit at launch. I alos recommend this strategy to everybody I know.

THis can't be in the best interest of the indutry though.

[tgr]

ain't you. sim city 5 limits you to a miserably small parcel of land.

Makes me wonder how the reviewers forgot to bitch about that as well. Did they get a version with a much larger piece of land or something?

[Goreschach]

http://www.youtube.com/watch?v=mA37cb10WMU

[Merusk]

whsdlkgh

So with a little bit of package editing within SimCity, and a little playing about in the code, it's possible to enable debug mode. I linked the activation to the "Help Center" button in the main menu for ease. Most debug features are disabled without having an actual developer's build (they have terraforming tools etc. available in the full developer build!), but a few things do still work - including editing the main highways.

Not only that - but you can edit the highways ANYWHERE - even outside of your city boundary... and even if you quit the game and log back in later, it's all saved safely on the server.

This shows that highway editing will be easily possible, AND that editing outside of the artificially small city boundaries should be very viable too.

Other things I have modded out with a quick change: Unlimited time to remain disconnected (won't get booted at 20 minutes, can now be disconnected "forever"). Population count now shows REAL figure, not the "artificially inflated" figure. My large cities have a population of about 15k now, not 100k :P

http://www.youtube.com/watch?feature=player_embedded&v=Bmce9oIxJag

Since he's connecting to EA's server, I expect to see a lawsuit and arrest under the DMCA for cybercrime and hacking in the near future.

Also, several of those make sense if you look at it through the lens of EA's approach to The Sims.  Release a basic product and then an x-pac that adds features every 6 months.  Bets that one of those features will be "larger city size!" along with a "Roadways and Mass Transit!" x-pac.

[apocrypha]

Edit: Nevermind, I've linked it to someone.

Good post Jeff.

[tgr]

With the whole ME3 deal people who complained about the end were 'entitled' and 'spoiled' and the whole deal was framed as an 'internet nerd fight' by 'disgruntled fans'. If EA would change the end it would be the end for authorship.

With Maxis the majority opinion by gaming journalists seems to be 'why do people expect to play on launch week, everybody knows the servers will be down'. Nobody focuses on the broken game mechanics, they all seem to expect that the servers will be fixed at some point and that this will be a great game then. As if anybody cares (except the people who spent their hard earned cash on a broken game) four weeks from now.

I'm not so much up in arms about the ME3 ending, considering the fact there's been tons of endings which, if you think about it, just makes you go "wha?". But if gaming journalists are actually going "everybody knows the servers will be down" and "they're all fools for actually buying the game at launch date and have the audacity of expecting the game to actually work", instead of basically kicking them in the nuts for being absolute shitheels for stuffing in shit everyone who's not a retard knows will fuck up more often than not on launch day/week, that is what worries me.

Then again, this has been a process which has been ongoing and escalating the last what, 5 years?

The consequence for me is clear. I simply won't buy anything at release anymore. Instead I'll wait for at least a month or until it's at half price or game of the year edition with all add ons. With the overabundance of entertainment I won't even miss anything except the heartache and shit at launch. I alos recommend this strategy to everybody I know.

THis can't be in the best interest of the indutry though.

I began implementing the "don't buy anything at release" procedure back around 2008/9, and that's just been reinforced time and time again as a good procedure to have. It's saved me tons of money.

[Paelos]

I agree with Jeff. I also think the day is coming when "publishers" aren't needed in the gaming industry, which will fix many of the fuckups. Developers won't be beholden to an overbearing group that has no interest other than providing an equity cut for their cash, and thus want a product as quick and dirty as possible. The publisher doesn't care if the product bombs and ruins the reputation of the developer, because it won't affect their deals with other developers.

It all comes down to financing operations. The more money that twirls around these projects, and the more we get out of hock in our own economy, the more likely you are to see traditional funders step up instead of just publishers. Right now, EA is nothing but a venture capitalist with an advertising budget for most of it's projects.

[Mithas]

stuff

I don't have anything to add other than to say this was a great post. If only some developers and/or publishers would read it and take note.

[Jeff Kelly]

I don't share your optimism.

Games like Sim City, Mass Effect, God of War and so on have huge development budgets. In fact they are so huge that they won't even tell you how much it cost to make one of those games.

Look at the end game credits of ME 3 for example. That's literally hundreds of people working on a product for several years. That's summer blockbuster money. I don't see how any of those games could have been or will be funded by a crowdfunding campaign so those games will always require publishers or financing and those people want to see a return on their investment.

Even an indie game on a mobile platform today most probably has a team of maybe one or two dozen people behind it, so even there you very quickly break the million dollar barrier.

The majority of the big kickstarter campaigns as far as games are concerned were never about 100% funding. They were there to collect enough funds so that they could show to an investor that there is a big enough interest in the game and find somebody who'll cough up the rest of the money. Project eternity for example was started on kickstarter to get additional funding in the range of several millon dollars by Bethesda, the company those guys work for.

Basically we're at the same stage movies were at the end of the eighties/mid nineties. The huge blockbusters of the eighties had already spawned an indie movement and people had already made a lot of independently funded smaller movies with a diffferent point of view and established an 'indie style'. Also that indie movement was already on its last legs because the funding required to even make an indie movie was breaking into seven figure budgets which brought everybody back into the studio system with all the crap it entailed.

Case in point: Wolfgang Petersen whose big break was 'Das Boot' and who went on to become a big hollywood director for a certain time once told  an interviewer that he had made 'Das Boot' for close to a million dollars and if he had to make exactly the same movie (same FX etc.) today (this was in 2009) the movie would cost more like 40 or 50 million dollars to make. The same fucking movie.

The sums of money required basically mandate investors or production companies especially since you need the budget up front to even start development on a game.

If you have time, listen one of the last few SMODcasts, the one where Kevin Smith talks to Aisha Tyler. He describes why he takes a break from filmmaking there. The gist is that he and all people involved had the most fun on small budget movies yet the budget required to even make an 'indie style' movie today is high enough that it basically makes sure that you can't escape the studio system and its influence over your effort.

I feel like this is exactly where the game industry is headed when even an old school adventure game or RPG needs to collect several million dollars to reach its funding goal.

[Paelos]

I don't share your optimism.

Games like Sim City, Mass Effect, God of War and so on have huge development budgets. In fact they are so huge that they won't even tell you how much it cost to make one of those games.

My point is that people will see those types of games as dead money. You don't need a massive budget to make a game. In fact, so much of that cost hasn't translated at all into a viable product at the end of the day. It's become an arms race of stupidity and bad business decisions because developers don't have control of their project, control of their costs, or control of their direction many times. The costs have spiraled into the stratosphere over the last five years without any tangible advances in gaming.

I understand a publisher getting frustrated with developers not understanding how a business is supposed to look, and wasting a shitton of money. The developers need business people in the organization to understand how to cost control a project, and the publishers need to trust those people and not interfere with the timing of the project. That's the ideal work. Now in a traditional financing arrangement, you wouldn't give equity to a publisher, you'd have a 30% equity, 70% bank financed agreement. The equity money gets your startup costs, the loan money covers your costs to completion, and as long as you pay interest on the loan, the bank doesn't give a fuck when you release the game.