Videos show path finding inherently "broken"

[Freiluft 6]

How difficult would it be simply to have traffic prefer a route based on the hierarchy of roadways SC5 has built into the game?

 

On the other hand, are you absolutely sure that the direct route using the dirt path isn't faster than the circuitous route on the avenue? If at a real life intersection I saw my destination at the end of a dirt road and a big avenue going off the the right, I might take the dirt road, thinking that not only was it the most direct, but also the fastest, especially if I didn't know the avenue made a big arc. Once home, I'd email the responsible jurisdiction demanding they upgrade the dirt road.

 

Part of the challenge of the game, however, is to make those roads you don't want heavily travelled not be the most direct route to anywhere else beyond the sims who live or work there.

 

The garbage trucks are similar to those "robot" vacuum cleaners or swimming pool cleaners that simply roam about randomly. They're not efficient, but they eventually get the job done with no additional programming. If, on the other hand, the task of the garbage trucks is "go find the next can and empty it," then it's failing miserably.

[Frode789 70]

wow this is actually terrible.. Did they learn nothing from SC4?

[Infinitum80 12]

Firstly let me say my opinion is of someone who has loved Sim City since the original but has YET to buy this game. My take on it is from spending plenty of hours watching people like Quill18 and others showcase the game in YouTube gameplay.

 

It is also suplemented by spending time on various forums and articles reading up on the problems.

 

The pathfinding is a joke.

Firstly, if this is intentional (ie - people would do this in real life) then I call BS on that. What next? The budget shows a defecit because the financial advisor called in sick and forgot to file the annual returns? The crime rate goes up because the prosecuting lawyer was found to be negligible and his prior arrests were thrown out?

 

I don't think it is at all intentional. The game is designed to simulate a city but the simultation has to do most of the work otherwise the micro involved is tedious. If this is the case then the game needs to allow us to set routes, traffic light priorities and similar. In real life people have traffic reports (*cough* traffic copter seen above busy intersections *cough*) SatNav and common sense to make the decision.

 

If you are on a road and simply by taking a 5 minute detour you can save 2 hours of gridlock, you will.

Someone on another forum (i dont have the link, sorry) said that if they are tourists perhaps they wont know the best route being from out of town... this is optimistic metagame that really wont be the truth.

 

The fact that cars arent even turning right in some (if not all) cases is just stupid - they will go 5 miles in the opposite direction when a simple right hand turn would have done the trick!

 

 

Pathing is CORE to Sim City.

This isnt a game where we want to place pretty buildings... isnt that something like Farmville?

 

They might as well have released a text adventure Sim City game for all it is worth

 

 

 

 

We had a beta to play

The beta was admitted to just test the servers (and even that failed)

 

 

If this was tested with Beta or even paid-EA testers and this went through as a design of the game... i weep

If this was tested but it slipped through, i question the validity of their testing practice

If this was tested and a different pathing process was invisioned, its a rush release that shows a bug

 

I think it was the second one - the designers wanted the game to work in a certain way and there wasnt enough testing on pathing!

 

If EA's plan was to rush a release, wait for enough people to complain about a certain few things and then focus on fixing those - they are in my eyes stealing from their customers in releasing a game that isnt ready and in fact they want to force their customers to work as testers for them! they should be paying the customers not vice versa!

 

 

1000s of 1 star amazon reviews

The biggest petition (or one of them) on Change.org for the change in DRM

Screams from websites saying that the calculations are done client not server side

An inside informer saying that calculations are not done server side (except for some marginal region)

 

 

 

 

If the truth regarding calculations is that we were lied to, EA may spend 100x the cost they lost in piracy to recover from the PR nightmare that ensues!

 

After all pirates would get into their code, see that the client is doing the work and workaround it - this would get out!

 

 

 

I am seriously hoping all of the negativity subsides, EA fixes the serious flaws and I will ultimately forgive them even if they did lie about the DRM calculations concern and if they did rush a poorly flawed game to market - as long as they can fix it and do it to a point that I can put my money up to pay for it.

 

 

Worst game release ever? possibly but there is still time for EA to redeem themselves - admit the screwed up with the pathfinding (and possibly the DRM) and reach out to the community to see how they can sort it.

 

Forget giving players $20 for a free game

 

.... Mcdonalds stuck a nail in your burger? a $20 gift voucher to buy another burger isnt going to fix the problem - Mcdonalds saying they got it wrong, saying sorry and suggesting how to resolve it for the future is the right way to handle it

 

 

 

We arent kids

Distracting us from the problem is not the fix, most of us grew up with Sim City from a young age and expect more.

EA it seems is full of execs who know nothing about the community and they really need to ask the question "How can we fix the problem" and they need to stop asking from INSIDE their bubble and reach out to us and ask us!

 

 

[GenXisT 127]

^ Fabulous!

[Infinitum80 12]

You may have seen this already - but just in case you havent

 

http://www.amazon.com/Electronic-Arts-41018ted-Edition2-SimCity/dp/B007VTVRFA/ref=sr_1_1?ie=UTF8&qid=1363183243&sr=8-1&keywords=simcity

 

Skirt past the near to 2000 1 star (I am sure if there was a 0 star option it would be lower) reviews

 

Most Helpful Customer Reviews section

 

http://www.amazon.com/review/R3MHPUOS7SBOHK/ref=cm_cr_dp_title?ie=UTF8&ASIN=B007VTVRFA&nodeID=468642&store=videogames

 

2,866 of 2,948 people found the following review helpful

 

 

This means that of the multitudes of people who have either bought the game, considering to buy the game or just browsing Amazon, near to 3000 people have actually spent the time to "vote" the review and of those, 97% of them found it helpful

 

Considering the review is basically saying bad things, that is a 97% consensus and near to 3000 people who are telling the Internet that they agree with this.

 

This is the equivilant of having a really bad YouTube video and instead of having the usual large green bar with the small touch of red where you had a few naysayers disagreeing with the video, its 100% reversed and if this was a YouTube video rating I reckon you couldnt even make out there was any green in the like/dislike bar!

 

EA may not check forums like this for feedback but the reviews for their product in the biggest marketplace for that product is something they will see.

 

They really need to eat some humble pie, admit they messed up and beg for forgiveness.

Gamers are forgiving as a whole but we have to see genuine reason to do so - a $20 giftcard and excuses and posts ignoring the problems wont fly.

 

I'd accept it for a few days but if their tone doesnt change soon for the better, id be happy to say id rather go without the game fullstop!

 

LOL, and the more cycnical ones of us might say "So, they gave everyone a $20 giftcard, if they all went out and bought Madden, they can use that data to figure that Sim City fans are American Football fans and gain a marketing win there as well!!"

[tSlater 134]

I am in no way endorsing the use of them, but I've been told by close friends that there's already a cracked version out there which works offline and doesn't connect to EA servers -- and functions completely with exception to multiplayer elements, further supporting the idea that all of the simulation and calculations are done client-side.

 

I cannot confirm this, however, since I do not have such copies and the new Sim City can't even run on my computer.

 

1000s of 1 star amazon reviews

The biggest petition (or one of them) on Change.org for the change in DRM

Screams from websites saying that the calculations are done client not server side

An inside informer saying that calculations are not done server side (except for some marginal region)

 

If the truth regarding calculations is that we were lied to, EA may spend 100x the cost they lost in piracy to recover from the PR nightmare that ensues!

 

After all pirates would get into their code, see that the client is doing the work and workaround it - this would get out!

[Infinitum80 12]

The sad thing about modern games now is with the development of most games being split into three camps

 

1 : Free to play, we will milk you over time

2 : Indie games, cost very little but are hit/miss regarding how good they are

3 : Expensive games, cost a bomb and you pretty much expect the world for what you get

 

Now, F2P and Indie games can allow people to get involved for nothing, or so little that they don't feel like their cost : enjoyment ratio is hurt

 

When you have an expensive game, like Sim City turn out to be screwed from Day-0, the ratio is beyond painful!

 

 

Why didn't they just make SimCity F2P and make the money on the DLC?

 

: F2P is for small cities, non-priority server access, less starting capital, no sandbox mode, only public regions, pay-2-play content (like fancier buildings, better tourist attractions etc)... and even tiered access pay-2-play so you can have access to certain features as a pay-2-play unlock (pay $10, get bigger cities)

: P2P monthly for subscription, get everything

 

I am being sarcastic... making Simcity like this (even though with paid DLC in the works, it will end up being as such anyways) would be tragic.

A single player game where you have to subscribe x10 the cost of what it would be retail (one purchase FOREVER) over a year would be criminal

 

 

I think they made this online for one reason - persistent income.

Unlike Sims 3 (or Sims 4) where they can expect people to purchase add-ons because of the nature of the game, Sim City really isn't the type of game where add-ons will be in demand. The core of the game is there at the start, the only add-ons you can get are fluffy missions and extra buildings - thats pretty much it.

 

There isn't much in the way of DLC development for SimCity because it isn't a Story-driven game or a level/map driven game. Its quite static and thats what makes it great.

 

So, how does EA make money out of SimCity? Micro transactions of course!!! - but the only way to realisticly do this is create a crowd of people they can wave a pretty banner in front of and get interest.

 

If SC was offline, the DLC would need to be marketed via Amazon, Google Adwords, Online / TV marketing, PC magazine marketing and email marketing... it would hit less than the intended people they want.

 

Now, imagine you have all of your players TRAPPED in a game that they will see a rotating banner whilst they wait to log-in pushing the upcomming DLC - possibly even in game adverts or popups and whilst most hardened gamers would filter that out, the increase in revenue for DLC transactions will go up.

 

 

 

Whilst hurting internet pirates is a thing, ultimately driving revenue past the release is their plan.

[elektrichead22 10]

 it's an eerie reflection of the initial sc4 alg model that pissed everyone off at intial release.

[Cool_Z 1,547]

 it's an eerie reflection of the initial sc4 alg model that pissed everyone off at intial release.

The SC4 engine (well the original one) was bad, but not completely broken as it is in SC2013

[morriswalters 16]

I'm not at all sure that pathfinding is broken as badly as people think it is.  I have a city running at about 200,000 population and I'm having no problems with traffic jams or lost garbage trucks.  Controlling intersections appear to be the key. 

[Infinitum80 12]

I havent got the game so really cant test (so tempted to buy it, but since the main reason i want to is search for flaws, i reckon i am not in the right mindset to get enjoyment from it!!)

 

Morris, did you try the test as per the youtube clips?

IE : Create a traffic target (expo centre) and 2 routes, one short but congested and one technically longer but quicker?

 

If so and it works for you, perhaps it is buggy and is working less than intended for all - otherwise if its global, no excuse from EA i reckon!

 

 

Call me silly but if i have traffic outside my door going RIGHT that has been there for 5 days or by just going left and taking a 1 hour detour i can get to work - guess which route i will take!

 

Also - if i am on a dual carriageway (avenue to you americans) then if i come out of my house and the road is moving right to left and I need to turn right, i need to go onto the road and make a uturn at the end of the dual carriageway to get where I am going

 

On a street, its far different. There are natural gaps in traffic you can edge out into and make the turn, even when 100% congested... so why is SimCity still telling the user to make a Dual Carriageway / Avenue style one way only route when going over the road is perfectly natural and needed.

 

Furthermore if we are working on a 1-way system as suggested above, why not let us make all the roads 1-way... then we could utilize roundabouts.

 

 

A roundabout for those outside UK/EU is a one way road that loops in a circle with off/on points. its like a clockface. So if you need to go from NORTH to WEST you get on at 12 oclock and travel around the road until you hit 9 oclock and exit.

 

Roundabouts are the single most smartest piece of traffic management and to see them removed from every Sim City game is a travesty

 

And then lastly, what about under/overpasses. If you are telling people they have no highways and need to use avenues like highways and have to travel 5 miles in one direction to turn around, why can't we have over or underpasses to allow people to come off the road, turn around and get back onto the junction and head the other way?

 

 

 

What really winds me up is that all of these things were learned back in the previous games

 

Highways worked in SC2000+ as citywide ring-roads since they enabled people to get on and off at set junctions and although this was simulated traffic and not actual agents, it still functioned the same fashion where the route is controlled by correct junctions.

 

 

At present you have poor pathfinding leading to badlly planned junction management (traffic lights at best) with no way for us to bypass the junction with an overpass.

 

If you have road A that is going for residential zone to industrial, north to south and road B that is going from the airport to the casino, west to east.

If (for arguments sake) we say that no one from either road has reason to travel to/from any of the points (so no resi's going to the casino and no workers going to the airport) then road A and B should be 100% seperate.

 

However for the airport traffic to reach the casino, they need to intersect at the centre of the city where the workers are heading to the industry zone.

 

They will butt heads and congestion will ensue. So if the workday traffic is congested, all of a sudden the airport>casino traffic is as well.

 

HOWEVER if i could create an overpass bridge to skip over the road - voila, no traffic problem

 

 

This was in SC2000, which TBH is from a performance/graphical point of view a shadow of games that can run on iPhones.

So... why on earth is the premier game from EA this year not able to perform better or as good as SC2000 in this respect

 

 

Its honestly like EA want to build a game that works for them but not for us.

And i have no idea why

 

 

 

 

 

And lastly... where did this "STICK ZONES ON EVERY PIECE OF ROAD" come from? Why cant a road be used to connected 2 zones without it being filled with decision points for people to stop on? I know zones are free but come on?

[DabblerNL 3]

I remember from SC4 that plunking just one factory and one shop in a massive residential area was enough to satisfy the need for access to work and shops for all residents traffic wise. At least that trick doesn't  work now. I do not really mind the simplification of "the first spot that suits my needs will do" (this game is not Cities in Motion) as long as they implement pathfinding and using multiple forms of public transport correctly. Just by adapting the animation of hundred people trying to enter a single trailer to just the ones that will fit and having the rest walk along  will make things look much more natural.

[morriswalters 16]

I havent got the game so really cant test (so tempted to buy it, but since the main reason i want to is search for flaws, i reckon i am not in the right mindset to get enjoyment from it!!)

 

Morris, did you try the test as per the youtube clips?

IE : Create a traffic target (expo centre) and 2 routes, one short but congested and one technically longer but quicker?

 

If so and it works for you, perhaps it is buggy and is working less than intended for all - otherwise if its global, no excuse from EA i reckon!

No., Why would I?  I believe the poster. But that still doesn't mean it is broken.  At least as badly as it appears.  Something can not work in some specific cases and yet work well overall,  Mornings are busy.  My city is broken in two by a streetcar avenue.  Close to the city entrance I let nothing cross that avenue.  Further away I let avenues cross and still further away lesser density roads.  At no time to this point has it ever stalled.  It gets busy, but it moves..  I'm running 2 recycling centers and a garbage dump and school buses.  All located on one side of the map. Things happen on a schedule.  I feed them into high density streets before they have to cross.  I also have city buses and street cars.  The behavior of the waste haulers is complex.  What they do appears to vary with the number of lanes on a road and whether there are buildings on both sides.  Once into a neighborhood they appear to loop until they are full or until nothing is left to pick up. Which means they will circle at times.  Buses the same.  But there is no fixed route, which makes sense since I change my city at a whim.  And it works.  My population is current fluttering around 190,000 and my recycling lets me run a 5000 simollion deficit by bringing in more money daily than I lose with my deficit.  The point of this is that if you control the intersections, you can control traffic.    If you wish I can post some screen grabs.  The traffic in the morning after an event is a bear outbound.  And the added hotels suck power.

[T Wrecks 10,209]

I remember from SC4 that plunking just one factory and one shop in a massive residential area was enough to satisfy the need for access to work and shops for all residents traffic wise.

Interesting to read. My experience is contrary, has always been. Not enough jobs => unemployment => abandonment. Or did I misunderstand you?

[Keigen Shu 17]

A lot of the service path-finding issues could've been fixed if they just added this logic:

"If I've been doing the roughly the same thing over and over again, i.e. collect about the same/lowering amounts of garbage for 2 minutes, go to the place where I'm needed the most RIGHT NOW."

 

[gryphus 0]

I am a software developer....

 

Imagine you use the variable of density or road speed to calculate a better route.... (ideal)

 

Once a single instance/agent/vehicle chooses an alternative road, that itself causes the variables to change their value, so you need to calculate again for the next car and again, and again......

 

So basicaly a good simulator based on single agents requires continuous calculation, and that means high processment capacity is required to run the simulation smoothly. (and also memory but thats not the big deal)

 

This doesnt happen in cities XL or SC4 because theres no single agents: and traffic is simplified and calculated periodicaly as a whole.

 

What im trying to say (with my poor english skills) is that... what we are seeing in these videos is a CHEAP SOLUTION for a problem of efficiency/performance... i'm sure they had it but at some point of the project the simulation has beed dumbed down for commercial interests....

 

Conclusion: low quality product.... don't even need to see the rest...

 

 

[CoolColJ 42]

 

[Board_Marshal 104]

 

 

That's a thing of beauty.

[Buggi 684]

The traffic system I have modeled for Boomtown doesn't care whatsoever about the distance traveled. What it uses instead is the TIME to cross any given area of transit. It is a very robust system that allows me to pre-calculate paths then periodically update the transit network to modify the time cost based on how many paths use that particular link.

 

So roads will have a set capacity that allows for full speed travel, once that is hit, more paths can use the road, but the time cost will increase as traffic will become slower. As the cost gets very high, new paths won't use the road and instead look for faster ways to their destination.

 

It is a system I designed and modeled 3 years ago and I am shocked that Maxis couldn't have come up with the same system.

[morriswalters 16]

It doesn't matter how you do it, it only matters if it works.  The best mechanism in the world won't correct for poor planning.  Obviously the shortest distance is not the only thing that counts.  Since neither of the two routes in the video are the shortest route, assuming it is legitimate, and I have no reason to believe otherwise..

[bixel 1,224]

I am a software developer....

 

Imagine you use the variable of density or road speed to calculate a better route.... (ideal)

 

Once a single instance/agent/vehicle chooses an alternative road, that itself causes the variables to change their value, so you need to calculate again for the next car and again, and again......

 

So basicaly a good simulator based on single agents requires continuous calculation, and that means high processment capacity is required to run the simulation smoothly. (and also memory but thats not the big deal)

 

This doesnt happen in cities XL or SC4 because theres no single agents: and traffic is simplified and calculated periodicaly as a whole.

 

What im trying to say (with my poor english skills) is that... what we are seeing in these videos is a CHEAP SOLUTION for a problem of efficiency/performance... i'm sure they had it but at some point of the project the simulation has beed dumbed down for commercial interests....

 

Conclusion: low quality product.... don't even need to see the rest...

 

I disagree a litte bit - here but mostly because there is a simple solution and it does not require 'calculations' or at least on a level that is difficult to do

 

Everyone who has played the game knows that when you lay down a road it plops down in specific length quantities, you can say it like a grid lets say 16. Buildings attached to the road may be a certain width like, 16, 32, 48 etc... So how do you know where the sim person is supposed to go? Lets say he is an owner of building lot 4435. the road has a building ownership property as an array.

 

Road.LotsL[] and Road.LotsR[], where the 1st [] is the left side, and the 2nd [] is the right side. So if the road is 6 tiles deep, the array is this length Road.LotsL[6] & Road.LotsR[6].

 

If house #4435 is 1 tile big it occupy one space in the array, and lets say its on the left side of the road, 3rd down, so Road.LotsL[3] = #4435 or Road.LotsL{0, 0, 4435, 0, 0, 0);

if house #4435 is quite large and occupys two spots, he goes into the 1st array index of that building number, example... Road.LotsL{0, 0, 4435, 4435, 0, 0);

 

this is a simply easy way of showing where the sim needs to go

[DabblerNL 3]

I remember from SC4 that plunking just one factory and one shop in a massive residential area was enough to satisfy the need for access to work and shops for all residents traffic wise.

Interesting to read. My experience is contrary, has always been. Not enough jobs => unemployment => abandonment. Or did I misunderstand you?

You had of course to provide enough jobs and shops, but it seemed that having your sims find a way to just one of them would satisfy the requirement that a job and a shop could be reached. It may be that I am quoting SC3000 of SC2000 behaviour here, it is that long ago :-)

[Infinitum80 12]

I get the people looking to cling to the hope that Maxis developed the game well and the pathfinding is just a glitch but i think it has been pretty much hashed out that it isnt a glitch, its a shortcut in programming and am sure if the developers were put to the microphone without fear of reprisal - they would admit it to and no doubt it was all down to something needing to be cut to save on resources.

 

What i reckon (not fact, just musings) is that Maxis told EA "pathfinding will cost X amount of resources that means you will need Y amount of servers and cost Z amount of dollars" - EA then said "Well, we reckon that if JUST the forecasted people play the game, we have the presales to pay for the resources we have however unless more come, we cant afford it - we will look at changing pathfinding when we have more business from it"

 

The floating people, the buggy/flawed roads, the bad servers - all these things were I believe cost savings by EA and Maxis had to cut corners to assure reliability and speed of release.

 

As more money gets thrown at it, the money will be spent on fixing the problems I hope but all EA cares about is money and they dont want to speculate their own when they can utilize customers money.

 

 

 

Synical? Perhaps but has EA really given us much reason to believe otherwise? We guess/fear/sumize the worst because that is what they have done to our confidence in them.

 

 

I guess I will need to wait 5 years for a good kickstarter SC clone to come out that addresses these problems!

[Marla_Singer 46]

I don't have the game, but from what I've read and seen, the worst trafic jams are mainly due to crowds of agents all getting the same instruction to go at the same place at the same time.

Considering this, I believe the simplification consisting in asking to cars to get to the nearest house/job is counter-productive. Not only people all rush to the same house, but also it makes public transportation unattractive. As a result, the best outcome is to mix altogether residential, commercial and industrial zones, so that the "nearest house/job" is actually different from a point of the map to another.

Considering the small size of maps, it would actually make more sense to me to randomize houses and jobs sims are going to (it would generate more realistic traffic patterns too). This could work well as long as a job slot is considered as "occupied" at the moment an agent is leaving house (and not at the moment he arrives at it).

Such a move shouldn't really take more memory than a function asking them to go to the nearest job/house. It wouldn't solve all (the single regional entrance is another big issue) but it would already be a big improvement.

http://www.youtube.com/watch?v=WH-CdkSqDSs

[Infinitum80 12]

The stupid one i heard was that agents dont reside in permenant locations, specifically with home and work.

 

Yes, i get that you could justify that everyone is a day worker and works in a different place each day but when they go home to a completely different lot is just crazy.

 

IMHO, its either graphical representation of traffic (so if its busy it refers to the relevant opportunities to reduce traffic, bus, rail, nearby extra routes etc) or its actual agents in traffic but you can't argue both because well you get what we have now.

 

As gamers we are used to seeing NPC's vanishing into thin air or going into rabbit holes, its fine and our suspension of disbelief works fine in this regard. What we don't like is when a game says to us that we are getting a real agent based engine and it falls down at the first hurdle suggesting either poor initial design (bad maxis!) or bad testing of possible flaws (bad maxis testers!)

 

Come on... testers figure out everything in a game.. what if i plop down a fire station whilst there is a fire nearby and hold my breath whilst tapping the escape and enter key in unison 100 times? - surely they could have tested "what if a traffic jam builds up when there is a quicker but longer route nearby?"

 

Either

a) This came up and was ignored (for financial or idiocy reasons)

b) This never came up because it worked in test but something changed and now it is bugged

c) This never came up because this wasnt tested

[trar 336]

While I am disappointed with the bad pathfinding, I should think that realistically, not everyone chooses the best path all the time. However, it doesn't happen too often, and it certainly doesn't lead to traffic snarls on small-capacity streets. Perhaps an algorithm that, like the NAM, determines pathfinding by capacity yet also sends a small, varying amount of traffic down a quicker yet lower-capacity route or routes (not that much, though) would be the best solution. I know a couple of people who prefer driving on the back roads to the main roads at any rate.

 

And while I'm on the subject, individually simulating sims doesn't work for me if they lose all identity once they step into their chosen workplace/residence. I don't find the whole 'treating electricity and sewage as individual agents' thing to be good design either, or the removal of water systems altogether for that matter. I mean, where are you going to plop your beautiful treatment plants and facilities now?!

 

Then again, perhaps this is all merely a clever allegory to the research being done with remote-control cars.

[marlonbraga 0]

im not sure if anyone have pointed it out already, but maxis have commented on those issues already on SimCity Blog Update 8

http://www.simcity.com/en_US/blog/article/simcity-update-8

So it seems they will fix it adding weight to redirect traffic based on road capacity reached.

[Marla_Singer 46]

While I am disappointed with the bad pathfinding, I should think that realistically, not everyone chooses the best path all the time.

I have something for you, it's called direction signs:

It doesn't guarantee the fastest route, but at least it makes you avoid dirt roads.

[trar 336]

While I am disappointed with the bad pathfinding, I should think that realistically, not everyone chooses the best path all the time.

I have something for you, it's called direction signs:

It doesn't guarantee the fastest route, but at least it makes you avoid dirt roads.

Yes, I know what signposts are, thank you very much. I'm talking about people who know this, yet still prefer back roads (because they're more scenic, or have less traffic, or they usually stop somewhere along the way), or people who are new to the area and don't see the signs for whatever reason (these people exist, i am quite sure of it), or people who simply want to take a nice drive. For that last reason alone, we should have randomizing transit algorithms. These random trips won't number nearly as much as the sims who actually have places to go and know how to get there of course, but they will exist, and still be a factor. Not a very big one, but still.

[Board_Marshal 104]

 

While I am disappointed with the bad pathfinding, I should think that realistically, not everyone chooses the best path all the time.

I have something for you, it's called direction signs:

It doesn't guarantee the fastest route, but at least it makes you avoid dirt roads.

Yes, I know what signposts are, thank you very much. I'm talking about people who know this, yet still prefer back roads (because they're more scenic, or have less traffic, or they usually stop somewhere along the way), or people who are new to the area and don't see the signs for whatever reason (these people exist, i am quite sure of it), or people who simply want to take a nice drive. For that last reason alone, we should have randomizing transit algorithms. These random trips won't number nearly as much as the sims who actually have places to go and know how to get there of course, but they will exist, and still be a factor. Not a very big one, but still.

 

I pointed out in another thread that somebody had made a recommendation that the education of the agent could determine determine the probability that it takes the best route.  That if Maxis is weighting distance, congestion, and road throughput (size) for pathfinding then after the optimal weights are determined, i.e., balanced, that the education of the sim agent would be a modifier impacting the weights making them possibly take a less than optimal path as the educated the Sim decreased.  For example, optimally educated Sims have a 1 modifier so they use the optimal weights while other education levels use 0.8 or 1.2 as modifiers (just proposed random #) which would slightly jack with the path selection.  This would be a nice way to tie education to behavior thus "simulating" peoples behavior in a city.  I think its an excellent idea.